J OI N T DIsC U ssION PA PER T H E W O R L D B A N K , B M Z , FA O , G I Z , I FA D , A N D U N C D F 70730 sUBsIDIEs As AN INsTRUMENT IN AGRICULTURE FINANCE: A REvIEW RICHARD L. MEYER JUNE 2011 Joint Discussion Paper – Issue GERMAN FEDERAL MINISTRY OF ECONOMIC COOPERATION AND DEVELOPMENT (BMZ), FOOD AND AGRICULTURE ORGANIZATION OF THE UNITED NATIONS (FAO), GERMAN AGENCY FOR INTERNATIONAL COOPERATION (GIZ), INTERNA- TIONAL FUND FOR AGRICULTURE DEVELOPMENT (IFAD), THE WORLD BANK, AND UNITED NA- TIONS CAPITAL DEVELOPMENT FUND (UNCDF) sUBsIDIEs As AN INsTRUMENT IN AGRICULTURE FINANCE: A REvIEW Richard L. Meyer © 2011 The International Bank for Reconstruction and Development / The World Bank 1818 H street, NW Washington, DC 20433 Telephone 202-473-1000 Internet www.worldbank.org E-mail feedback@worldbank.org All rights reserved. This work was undertaken under the Joint Donor CABFIN Initiative, whose core members are the German Federal Ministry of Economic Cooperation and Development (BMZ), Food and Agriculture Organization of the United Nations (FAO), German Agency for International Cooperation (GIZ), International Fund for Agricultural Development (IFAD), The International Bank for Reconstruction and Development/World Bank, and the United Nations Capital Development Fund (UNCDF). The findings, interpretations, and conclusions expressed in this volume do not necessarily reflect the views of the Executive Directors of The International Bank for Reconstruction and Development/The World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. Rights and Permissions The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. The International Bank for Reconstruction and Development / The World Bank encour- ages dissemination of its work and will normally grant permission to reproduce portions of the work promptly. For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center Inc., 222 Rosewood Drive, Danvers, MA 01923, UsA; telephone: 978-750-8400; fax: 978-750- 4470; Internet: www.copyright.com. All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H street NW, Washington, DC 20433, UsA; fax: 202-522-2422; e-mail: pubrights@worldbank.org. Author Prof. Emeritus Richard L. Meyer, Ph.D., Ohio state University Technical Editors Rauno Zander, Ph.D., and Heidi Fritschel Cover Photo: Renate Kloeppinger CONTENTS i Table of Contents ABBReVIAtIons AnD ACRonYMs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iii ACKnoWLeDGMents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iv eXeCUtIVe sUMMARY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v CHAPteR 1 Introduction 1 1 1 Organization of the Paper 1 1 2 Terminology 2 CHAPteR 2 Financial services and the three “F� Crises 4 CHAPteR 3 the Role of Finance in Agricultural Development and Rural Poverty Alleviation 6 CHAPteR 4 the Challenge of Developing Credit Markets 8 4 1 Private Markets and Market Failure 8 4 2 Information Asymmetries and Contract Enforcement Problems in Credit Markets 8 4 3 Special Challenges for Credit Markets to Serve Rural Areas and Farmers 9 CHAPteR 5 the Paradigm shift in Developing Credit Markets 11 5 1 The Old-Paradigm Approach 11 5 2 Emergence of a New Paradigm 12 CHAPteR 6 the Microfinance Industry: successful Application of the new Paradigm 14 6 1 Guidelines for Developing Microfinance 14 6 2 Microfinance Penetration into Rural Areas and Agriculture 15 6 3 Innovations and Prospects for Future Agricultural Lending 16 6 4 Insights Gained from Microfinance about the Impact of Finance 18 CHAPteR 7 Credit Demand, Rates of Return, and Interest Rate sensitivity 20 7 1 Demand for Credit 20 7 2 Rates of Return in Agriculture 20 7 3 Rates of Return in Microenterprises 22 7 4 Sensitivity of Loan Demand to Interest Rates 22 CHAPteR 8 strengthening Agricultural Credit: the Way Forward 24 8 1 The Use of Subsidies and Grants in Financial Sector Development 24 ii C ONTENTS Table of Contents 8 2 Forms of Subsidies 24 8 3 Criticisms of Subsidies 25 8 4 Analysis of the Use of Grants to Stimulate Investments 25 8 5 Use of Grants and Subsides for Rural Finance 26 8 6 The Concept of Smart Subsidies 28 8 7 Smart Subsidies for Microcredit and Agricultural Credit 28 CHAPteR 9 Five Major Approaches to support Agricultural Finance and Design of subsidy Inflows 30 9 1 Microinsurance and Weather Index-Based Insurance 30 9.1.1 Microinsurance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 9.1.2 Indexed crop and livestock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 9 2 Credit Guarantee Funds 33 9.2.1 Rationale for guarantee schemes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 9.2.2 Review of guarantee experiences in the 1990s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 9.2.3 CGAP study of guarantees for MFIs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 9.2.4 UsAID guarantees . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 9 3 Warehouse Receipts 37 9.3.1 Prerequisites for warehouse receipts lending . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 9.3.2 Warehouse receipts financing in Africa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 9 4 Specialized Agricultural Development Banks 40 9.4.1 Principles for reforming agricultural development banks . . . . . . . . . . . . . . . . . . . . . . . . . 40 9.4.2 Ownership and governance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 9.4.3 Adopting microfinance technologies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 9.4.4 Risk management techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 9 5 Agricultural Investment Funds 42 9.5.1 Microfinance investment funds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 9.5.2 Investment funds for agriculture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 CHAPteR 10 Conclusions and emerging Recommendations 45 10 1 Conclusions 45 10 2 Emerging Recommendations 46 ReFeRenCes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49 LIst oF FIGURes Figure 1: the Financial Market 2 Figure 2: Food Commodity Price Indices 5 LIst oF BoXes Box 1: examples of subsidies that Contribute to sustainable Rural Finance 26 Box 2: subsidies to the Poor for Asset Acquisition 27 S U B S I D I E S A S A N INST RUME NT IN AGRICULT URE FINA NCE: A REV IEW iii ABBREVIATIONS AND ACRONYMS ADB Asian Development Bank AgDB Agricultural development bank AGRA Alliance for a Green Revolution in Africa BAAC Thai Bank for Agriculture and Agricultural Cooperatives BMZ German Federal Ministry of Economic Cooperation and Development CABFIN Improving Capacity Building in Rural Finance CECAM Caisses d’Épargne et de Crédit Agricole Mutuels de Madagascar CGAP Consultative Group to Assist the Poor CvECA Community-managed village savings and credit organization DCA Development Credit Authority FAO Food and Agriculture Organization of the United Nations FJMC José Maria Covelo Foundation in Honduras GCv Grenier commun villageois GIZ Germany Agency for International Cooperation IFAD International Fund for Agricultural Development KACOFA Kapchorwa Commercial Farmers Association (Uganda) MF Microfinance MFI Microfinance institution MIs Management information systems MIv Microfinance investment vehicle MsMEs Micro, small, and medium enterprises NBC National Bank of Commerce (Tanzania) NGO Nongovernmental organization NMB National Microfinance Bank (Tanzania) Rural sPEED Rural savings Promotion and Enhancement of Enterprise Development sEAF small Enterprise Assistance Funds sFI special financial institution sMEs small and medium enterprises UNCDF United Nations Capital Development Fund UsAID U.s. Agency for International Development WFP World Food Programme All dollar amounts are U.s. dollars unless otherwise indicated. iv A CK NOWLE D GEM ENTS ACKnoWLeDGMents This paper is a joint undertaking of the CABFIN Initiative. CABFIN is a donor initiative focused on improving rural and agricul- ture finance, and its members are FAO, GIZ, IFAD, UNCDF, and the World Bank. Besides the main author, Dr. Richard Meyer, all CABFIN members were involved in providing technical feedback and reviewing various sections and, finally, the full text at different stages. This is, thus, truly a joint publication. The following CABFIN members were involved in the intense process and deserve special mention: Calvin Miller (FAO); Marietta Feddersen, Dr. Brigitte Klein, Nana Klietsch, and susanne schellhardt (GIZ); Michael Hamp (IFAD); Eric Dietz, Henri Dommel, Fode Ndiaye, Beth Porter, and Hanadi Tutunji (UNCDF); and Renate Kloeppinger, Ajai Nair, and Maria Pagura (World Bank). A special thanks goes to the following external reviewers for their insightful comments: Alexia Latortue (CGAP), Richard Roberts (Consultant), M.s. sriram (Indian Institute of Management), and Mark Wenner (IDB). Thanks goes also to Jamie Anderson (IFAD), Paul Armbruster (DGRv), Eric Duflos (CGAP), Michael Jainzik (KfW), Antonique Koning (CGAP), Johannes Majewski (GIZ), Ake Olofsson (FAO), Francesco Rispoli (IFAD), and Ilonka Rühle (sparkassenstiftung) for their insights and valuable feedback. SUBS ID IES A S A N INSTR UMENT IN A GRICULTUR E FINA NC E: A R EV IEW S U B S I D I E S A S A N INST RUME NT IN AGRICULT URE FINA NCE: A REV IEW v eXeCUtIVe sUMMARY sUBsIDIes As An InstRUMent In AGRICULtURe FInAnCe: A ReVIeW1 Providing sustainable financial services for rural areas and agriculture in developing countries has proven to be difficult in spite of recent reforms and billions of dollars spent in subsidizing programs to develop financial institutions. This paper presents a literature review of the lessons learned in the use of subsidies and investments as instruments of agricultural development finance. The emphasis is largely on agricultural credit, primarily for small farmers. Because of time limitations, this paper does not include the rapidly expanding literature on value chain finance. The paper is intended for decision makers in developing countries as well as staff in international agencies, nongovernmental organizations, and other organizations interested in supplying financial services for poor farm and nonfarm households in rural areas. DeVeLoPInG AGRICULtURAL AnD RURAL CReDIt MARKets Credit markets diverge from an idealized market because information is imperfect and loan contracts are difficult to enforce. Market failure is said to occur when the market fails to allocate resources efficiently. The complicated environmental, mate- rial, and production features of agriculture inhibit the demand for and supply of credit and insurance, making it especially difficult to create sustainable financial institutions to serve the sector. Not surprisingly, therefore, efforts to increase formal credit supplies have had a spotty record, and quick fixes have not worked. Most successes have been the result of careful long-term institutional development. In the period 1960–80, old-paradigm, subsidized, directed agricultural credit programs were common in top-down government and donor policies and programs. Unfortunately, attempts to resolve supposed market failure often resulted in government failure. In the 1980s, a new financial systems paradigm emerged that shifted the emphasis from dispersing cheap credit to creating sustainable institutions, treating borrowers and savers as clients rather than beneficiaries, developing products that clients demand, and pricing products and services to cover costs and risks. Donor agencies reduced the use of credit lines in favor of grants, loans, and technical assistance to help in the design of appropriate products, institutions, and policies. Microfinance also thrived by following this market-oriented approach. Microfinance institutions (MFIs) have made inroads into agriculture and rural areas, but more efforts are needed to design products and methodologies to fit the seasonal cash flow patterns of farm households. Managing the costs and risks of agricultural lending has been challenging. There is a need to better understand the demand for and use of agricultural credit to develop effective products, institutions, projects, and policies. The rapid growth of microfinance suggests that there may be large unmet demand for agricultural loans, but two issues need consideration. First, there may be a tendency to overestimate demand, as has occurred with microfinance. second, an empirical question concerns borrower sensitivity to interest rates relative to other factors affecting demand. Farmers’ demand for loans may be limited if the interest rates charged are as high as MFIs require to provide small microenterprise loans sustainably. 1 This is a publication of the joint donor CABFIN initiative. CABFIN stands for “Improving Capacity Building in Rural Finance.� Current core members of CABFIN are the German Federal Ministry of Economic Cooperation and Development (BMZ), Food and Agriculture Organiza- tion of the United Nations (FAO), International Fund for Agricultural Development (IFAD), Germany Agency for International Cooperation (GIZ), The World Bank, and United Nations Capital Development Fund (UNCDF). The document was prepared by Prof. Emeritus Richard L. Meyer, Ph.D., of the Ohio state University Department of Agricultural, Environmental, and Developmental Economics. Rauno Zander, Ph.D., did the technical editing. vi Ex ECUTIV E S UM M A RY Use oF sUBsIDIes Analysis of the use of subsidies in donor programs has led to guidelines for “smart� or “market-friendly� subsidies. These guidelines include the following: subsidize the institution but not the borrowers to reduce distortions; avoid subsidies to insti- tutions that undermine competition; subsidize the creation of public goods that benefit the entire financial sector; subsidize individual financial institutions where there is natural spillover to nonsubsidized institutions; identify quantitative performance measures so subsidies to financial institutions do not dull incentives for high performance; conduct comparative cost-benefit studies to identify subsidies that generate the greatest payoff; require grant recipients to demonstrate commitment through matching contributions; and design grants to financial institutions so recipients clearly understand the difference between grants and loans. FIVe MAjoR InteRVentIons This paper also evaluated five major interventions involving subsidies used to kick-start private sector activity in support of financial services. Microinsurance and weather-index-based insurance Many experiments are being conducted with micro- and weather-index-based insurance products to mitigate risks both for households and for lenders supplying financial services to the poor and to farmers engaged in crop and livestock enterprises. Weather-index insurance offers the promise of reducing the administrative, adverse selection, and moral hazard problems of traditional insurance. Bundling insurance with loans and savings is emerging as a logical step to reduce costs and speed adoption, but additional experiments in diverse environments are needed to develop best practices. Donors can play a useful role in conducting or financing careful evaluations to improve understanding about basic insurance questions. Robust evalua- tions are needed to assess whether insurance investments actually produce the desired effects and to determine the role for public subsidies in private insurance markets and catastrophe insurance. A logical role for governments and donors is to focus on long-term public goods investments, such as in weather-reporting stations and basic data collection and analysis, which are needed to create the conditions and infrastructure for robust insurance markets. Complementary investments are also needed in basic methods of mitigating risk through low-cost irrigation, drought-resistant seed varieties, improved sanitation, and preventive health care. Credit guarantee funds Donors and governments expect credit guarantee funds to reduce default risks and induce lenders to serve specific target groups or institutions. It is believed that guarantee subsidies accelerate learning so lenders will improve credit analysis and lend their liquid funds rather than investing them in government securities or lending only to highly collateralized borrowers. However, the methodology used in evaluating guarantees has been weak, so questions about additionality and sustainability remain. Guarantees may provide additional comfort for financial institutions interested in testing the feasibility of lending to a new clientele, but a guarantee alone is unlikely to induce additional lending if lenders lack such interest. International agencies can perform a valuable service by conducting evaluations to determine if and under what conditions guarantees produce the expected results and how the details of guarantee designs affect performance. It is also critical to evaluate whether they distort markets and discourage private credit market development. It may be that the training and technical assistance com- ponents of guarantee schemes are more important than the guarantees themselves to stimulate lending to a new clientele. This situation would suggest that “guarantee plus� programs are critical and that guarantees may be the frosting on the cake, not the cake itself. Warehouse receipts The basic rationale for warehouse receipts is that they reduce lenders’ risk by serving as a collateralized commodity that can be liquidated in the event of loan default. Commodities are stored in licensed and bonded warehouses that issue receipts certifying the amount and quality stored. The owners of the commodity (such as farmers and traders) provide the receipts to S U B S I D I E S A S A N INST RUME NT IN AGRICULT URE FINA NCE: A REV IEW vii lenders in exchange for loans. Except in the case of double or triple cropping, credit obtained after harvest does not directly solve the seasonal need for working capital to plant a new crop. There are too few careful analyses to conclude when and where warehouse receipts systems contribute to improved access to agricultural credit, especially for small farmers. They may improve commodity storage and marketing functions in value chains with trickle-down benefits through prices paid to farmers. The expenses of creating, operating, and monitoring these systems imply that scale is a serious challenge, so simple, small-scale village-level systems may be most appropriate for small farmers. Moreover, the critical need for small farmers may be production loans to meet seasonal cash outflows at the beginning of planting rather than marketing loans after harvest. More detailed analyses of farm-level commodity prices are needed to determine which crops normally experi- ence seasonal price variations large enough to compensate for storage and borrowing costs. The fact that warehousing is common for export crops suggests that economic barriers may constrain expansion into grains and other commodities produced primarily for local markets. several long-term public goods investments have been identified to make warehouse receipts financing work, and many may be appropriate for donor funding. Specialized agricultural development banks The subsidized, directed-credit paradigm led to the creation of many state-owned agricultural development banks (AgDBs). These banks have generally performed poorly, and there are debates about what to do with the failing institutions. successful reforms are possible only if governments make fundamental changes in ownership, governance, products, and perhaps even the clientele served. some reformed banks have successfully adapted microfinance procedures for agriculture. sophisticated risk management techniques are needed, however, for financial institutions that expect to make large loans to farmers and nonfarm businesses. Donors face the challenge of formulating a response when local leaders create new AgDBs, given the generally negative performance of such projects. To be successful, start-ups must have an institutional design that solves governance and management problems, a successful firewall against political interference, and a commitment to charging full cost recovery interest rates. One way to help avoid political capture may be to make small loans using microfinance technolo- gies and slowly graduate to larger loans as institutional capacity grows and access to commercial funding sources is achieved. Agricultural investment funds Agricultural investment funds are a form of financial mutual fund for pooling investor capital and investing it for investor prof- its. They offer opportunities to pool risk through diversified investments and employ professional fund managers to conduct risk assessments of investment opportunities and administer the investment portfolio. The estimates of huge requirements for agricultural investment in developing countries provide a logical argument for more external investing. Analysis is needed to determine if these funds induce more local lending by financial institutions. It is likely that investments will tend to benefit more affluent and entrepreneurial farmers and agribusinesses, so wealth and income distribution implications may be im- portant. Benefits in the form of better access to inputs, markets, and jobs may trickle down to small farmers and the poor, but high information, transaction, and contract enforcement costs mean that special measures are required to integrate poor farmers into value chains that benefit from these investments. Moreover, unless these funds invest in financial institutions that serve agriculture, they will not contribute to broadening the supply of financial services important to farmers and rural people. Intensive monitoring and analyses of fund activities is an appropriate and productive role for international agencies. subsidies for the technical assistance components of investments may help strengthen local capacity, make investments pro- ductive, mitigate risks, and cover some of the costs of helping small farmers participate in value chains in which investments are made. The funds may contribute to broadening the supply of rural financial services if technical assistance facilitates investment in financial institutions. ConCLUsIon Overall, this review concludes that there are no simple solutions for creating sustainable agricultural credit systems. With some noteworthy exceptions, the old-paradigm approach did not generally lead to sustainable agricultural credit in- stitutions. More recently, careful development of products, policies, institutions, and supportive infrastructure has led to greater success. Renewed interest in development economics has raised fundamental questions about financial services. viii Ex ECUTIV E S UM M A RY Researchers’ new, more rigorous methods hold promise for deepening our understanding of human behavior and how it influences credit market operations. This research, plus lessons learned from many innovations being tested around the world, needs to be broadly disseminated to benefit the entire financial industry. supporting and learning from re- search and innovations will provide international agencies with many opportunities to assist in pushing out the frontier of agricultural credit in developing countries and to use selective subsidies and investments for the greatest impact. S U B S I D I E S A S A N INST RUME NT IN AGRICULT URE FINA NCE: A REV IEW 1 Chapter 1: IntRoDUCtIon Providing sustainable financial services for rural areas and 1 for members of the CABFIN partnership). This review also agriculture in developing countries has proven to be dif- clearly defines the different types of subsidies currently used ficult. Billions of dollars have been spent subsidizing pro- in agricultural development and proposes “smart�—that is, grams and policies to develop financial institutions to serve appropriate—use of subsidies to support viable agricultural this neglected market. In most countries, decision makers, development. ministers of food and agriculture, and farmers are dissatis- fied with the results. Critics of the market-oriented financial The emphasis here is largely on agricultural credit, with less reforms implemented following the collapse of the directed- attention given to other financial services. This paper focuses credit paradigm claim these reforms have failed, given that on credit primarily for small farmers rather than for large agriculture continues to receive only a small share of total farmers and agribusinesses, which normally have better formal credit and that most farmers must rely on savings or access to commercial credit sources. It does not delve into informal credit supplies to finance their operating costs and the rapidly expanding literature on agricultural value chain fi- long-term investments. These critics argue for a return to nance, which offers a viable alternative for reducing risk and more active government intervention, including the creation costs in some types of agricultural finance while increasing of state-owned agricultural development banks. access to funding for some smallholders who may otherwise not be reached unless there are subsidies or directives for The urgency of the problem has increased in recent years be- financing them. cause of the triple shocks—food, fuel, and financial—that hit developing countries especially hard. A desire to “do some- The paper is intended for decision makers in developing thing quickly� pushed governments and donors to provide countries as well as staff in international agencies, financial resources and subsidies in response to shortages and ris- institutions, nongovernmental organizations (NGOs), and ing prices. These pressures logically raised questions about other organizations who make decisions about programs and the role of financial institutions in helping resolve short-term policies affecting financial services, especially credit, for poor problems as well as in supporting long-term agricultural farm and nonfarm households in rural areas. It is expected growth and development. Additional questions were raised to serve as an input into the preparation of an official policy about international agencies’ appropriate role in aiding gov- statement of the CABFIN donor consortium on the role and ernments that faced demands to adopt suboptimal policies status of subsidies in agricultural finance worldwide. to calm social tensions. This paper presents a literature review of issues related to 1 1 oRGAnIzAtIon oF tHe PAPeR recent subsidies and investments in the financial sector that section 1.2 introduces the basic terminology used in the pa- have been designed to address the immediate effects of per. section 2 of the paper discusses the impact of recent the crises and to develop the financial institutions necessary food, fuel, and financial crises on developing countries and to modernize agriculture. The literature consulted reflects the emergency actions taken by countries and international a combination of academic sources and reports from the agencies to reduce the suffering inflicted on poor people. It agencies and organizations involved in supplying financial also discusses the challenge of finding a balance between services in developing countries. Against this background, pragmatic immediate responses and longer-term objectives. this paper presents a review of subsidies as an instrument of agricultural development finance. This review is one of the The third section discusses the role of finance in agricultural major recent initiatives of the CABFIN partnership (see note development and poverty alleviation. section 4 deals with the challenge of creating credit markets in developing countries. 2 C H A PTER 1 — INTR OD UC TION It covers the economic concept of market failure, informa- subsidies—especially in the food, fertilizer, and credit mar- tion asymmetries and contract enforcement problems in kets—and the rationale for smart subsidies. It then describes credit markets, and the special challenges of serving rural experiences in five major areas of international agency ac- areas and farmers. tivities: microinsurance and weather-index-based insurance, credit guarantee funds, warehouse receipts, specialized The fifth section covers shifts in the paradigm used to inter- agricultural development banks, and agricultural investment vene in credit markets and summarizes the main features of funds. the old directed-credit and the new financial systems para- digms. This is followed by a sixth section that summarizes section 9 summarizes the main conclusions based on lit- highlights in the development of the microfinance industry. erature consulted for this review. It identifies major lessons It covers guidelines created for developing microfinance, learned with suggestions for priorities that CABFIN members microfinance penetration into rural areas and agriculture, might consider supporting in their projects and programs. innovations and prospects for future agricultural lending, and insights gained about the impact of finance on poor households. 1 2 teRMInoLoGY The terminology used in the paper follows the diagram in The seventh section addresses topics related to the demand Figure 1.2 The terminology covers the economic sector, for credit, including rates of return earned in agriculture and geographic location, and size of financial transactions. The in microenterprises, and research results analyzing sensitiv- financial market refers to all financial services for all purposes ity of loan demand to interest rates. from all sources used in both urban and rural areas, including section 8 describes major interventions by international credit, savings, insurance, remittances, and money transfers. agencies and points the way forward for agricultural The providers encompass all types of formal and semiformal credit. It reviews the debates about the use of grants and 2 This figure and a glossary of basic terms are found in IFAD 2010. FIGURe 1 Source: IFAD 2010, 12. S U B S I D I E S A S A N INST RUME NT IN AGRICULT URE FINA NCE: A REV IEW 3 institutions, including banks, credit unions, NGOs, and micro- rural areas but is now moving into more rural locations to finance institutions (MFIs). serve farm and nonfarm firms and households. Agricultural microfinance refers to small-size transactions for poor farm The terms rural and urban refer to location. In most coun- households and farm-related businesses. Rural microfinance tries, rural refers to nonurban geographic areas (villages, covers small-size financial transactions for both agricultural towns, and small cities) with fewer inhabitants than found and nonagricultural firms and households in rural areas. in larger cities and urban areas. Agricultural finance refers to financial services used by the agricultural sector, mean- Agricultural value chain finance refers to financial products ing farming and farm-related activities including input sup- and services that flow into or through an agricultural value or ply, processing, wholesaling, and marketing. Most of these supply chain. Providing credit into or through a value chain activities are conducted in rural areas, but large processing for an agricultural commodity is viewed as a complementary facilities and agribusinesses are also located in urban areas. approach to the financial systems approach that is the pri- Agricultural credit is normally provided in cash, but some pro- mary focus of this paper. grams provide in-kind loans for seed, fertilizer, and other farm production inputs. Subsidy refers to pecuniary aid usually furnished by a govern- ment to a private business, a charity, or an organization in the Rural finance is a broader category including all financial form of a cash grant, in-kind goods or services, or exemption services used by farm and nonfarm firms and households lo- from some requirement, such as a tariff or tax, that is nor- cated in rural areas. Many nonfarm enterprises in rural areas mally assessed on similar businesses or organizations. are directly related to agriculture, such as input supply or pro- cessing firms, but restaurants, hotels, retail shops, and other A grant is a gift of money or goods provided to a private rural businesses also require financial services. Financial in- business, charity, organization, or government to be used for stitutions that provide credit to farmers are often encouraged some specified purpose; in contrast to a loan, a grant is not to serve nonfarm customers as a way to diversify risks and expected to be repaid. expand their operations. An investment normally refers to an outlay of money or capi- Microfinance (MF) refers to financial services usually involv- tal designed to gain profitable returns in the form of interest, ing small transactions and products specifically designed for income, or appreciation in value. International donor agen- low-income households and small-scale businesses in both cies often use the term to mean funds spent in a project rural and urban areas. In many countries MF has concen- to accomplish some development objective in the country trated in urban and peri-urban areas or in densely populated where invested but not with the expectation of a direct finan- cial return to be earned by the agency. 4 CH A PTER 2 — FINA NCIA L S ERVICES A ND TH E TH R EE “F� C R IS ES Chapter 2: FInAnCIAL seRVICes AnD tHe tHRee “F� CRIses The beginning of the 21st century was promising for de- veloping countries, but the food, fuel, and fi nancial crises in 2007–08 severely disrupted per capita income growth and interrupted private capital fl ows. Commodity-importing countries faced deteriorating external balances, rising prices, and weaker income and household spending because of the high prices. The impact was most severe on the poor, who often spend 50 percent or more of their total budgets on food and fuel (World Bank 2009b). One study projected that the financial crisis would add 53 million people to the number of people living on less than $1.25 a day and 64 million to the number of people living on less than $2 a day (Chen and Ravallion 2009). The crises threatened to reverse progress Food crisis in the Horn of Africa (Photo: Alex Wynter/IFRC) toward improving financial access because of their effects on fi nancial institutions. If borrowers in rural and microfi- nance institutions were not able to repay their loans, MFIs as well as address longer-term issues.4 The short-term might be unable to refinance their loans when due or obtain emergency components of these programs logically focused additional financing to continue their rapid growth in lending.3 on safety nets to ensure household food security and re- duce poverty and vulnerability among the poorest groups. The crises revealed disturbing long-term trends in develop- Interagency assessments also identified medium-term ing countries that aggravated short-term problems. First, measures that stressed the provision of inputs, market de- investments in agriculture had lagged as both foreign as- velopment, price stabilization, and investments in transport, sistance and national government investments in agriculture storage, processing, marketing facilities, market information, declined. As a result, the growth rate of agricultural produc- and capacity building (FAO 2009a). Most programs did not tivity dropped from about 3.5 percent in the 1980s to 1.5 per- focus specifically on financial systems, but many country cent today. second, the negative effects of climate change assessments identified the need to develop credit and finan- will likely raise food prices and may put an extra 49 million cial services, emphasizing finance and microcredit for small people at risk of hunger by 2030 (IFAD 2009b). Third, natural farmers. several placed emphasis on scaling up microcredit resource degradation over the next 25 years is projected to and, increasingly, microinsurance programs aimed at small reduce global food production by as much as 12 percent, farmers and vulnerable groups and on strengthening exist- pushing world food prices as much as 30 percent higher ing rural financial institutions, such as farmers’ associations (IFPRI/Terrafrica/GTZ 2009). and credit unions. Warehouse receipt systems and voucher As the crises spread, governments rushed to enact mone- schemes were recommended in projects to deliver credit to tary and fiscal stimulus measures, and international agencies the poorest farmers through NGOs and MFIs (FAO 2009a). expanded their programs to alleviate short-term problems 3 Llanto and Badiola (n.d.) analyzed these issues for Asia, and 4 The World Bank (2008) summarized these initiatives for the UN CGAP (2010b) developed a large spreadsheet with information agencies, multilateral banks, and major bilateral development that evaluated MFI problems by region. agencies. S U B S I D I E S A S A N INST RUME NT IN AGRICULT URE FINA NCE: A REV IEW 5 A framework explaining the pathways through which credit Fortunately, the spike in international food prices quickly and savings contribute to food security was used to guide passed, as shown in Figure 2, but prices continue to be several field studies in the 1990s that demonstrated these higher than 2002–04 levels, and the recent closing of wheat relationships (Zeller et al. 1997). A more recent framework exports from Russia will once again cause price increases. for enhanced food security proposed a graduation model in These price changes and continuing food security problems which the poorest households are initially aided through asset pose serious challenges. It is important, therefore, to con- transfers (that is, grants and food aid), village savings groups, sider how developing sustainable rural and agricultural credit and skills training. As their capacity increases, participants systems contributes to long-term agricultural development eventually become creditworthy and receive market-oriented and poverty alleviation. small loans (UsAID 2010). This type of conceptualization may provide guidelines for future strategies to link short-term FIGURe 2 emergency assistance with long-term financial services.5 Responses to the crises revealed the practical need to bal- ance pragmatic immediate solutions with longer-term objec- tives.6 In times of crisis, governments face overwhelming pressures to adopt suboptimal policies to maintain social stability, but such policies may have negative long-term con- sequences. International agencies can help ensure that such policies are temporary, contribute to the design of clear exit strategies, and promote transparent administrative mecha- nisms, while emphasizing more sustainable and better- targeted mechanisms over the medium and long term. By adopting some flexibility, international agencies hope they will be welcome partners in moving from short-term policies to address sociopolitical unrest toward more sustainable long-term policies (World Bank 2008). A limitation of pragmatism, however, is that it may force decision makers to choose short-term activities that under- mine long-term development. For example, will supplying free or subsidized food provide relief to the urban poor but drive down food prices and discourage production by small farmers? Will government-subsidized fertilizer stymie the development of private fertilizer markets? Will debt repay- ment moratoriums or loan forgiveness programs thwart the Source: FAO 2009c, 14. development of sustainable financial institutions? While political necessities force local decision makers to focus on Recent crises have showed that different types of subsi- immediate needs, international agencies cannot lose sight of dies—whether direct interest rate subsidies or other indirect long-term development objectives. subsidies—are still very much part of the quick-fix toolbox of governments under pressure. These pragmatic and short- term responses must be accompanied by a clear path to sus- tainable agricultural finance in particular and to sustainable 5 This graduation approach is referred to in section 8.1.1.4. agricultural development in general. Donor agencies should 6 Principle 3 of the Declaration of the World summit on Food se- maintain this focus in their policy dialogues and identify sub- curity recognizes the need for a mixed approach: “strive for a comprehensive twin-track approach to food security that con- sidies that are not fiscally sustainable or have detrimental sists of: (1) direct action to immediately tackle hunger for the effects on overall resource allocation within the agricultural most vulnerable and (2) medium and long-term sustainable agri- sector. cultural, food security, nutrition and rural development programs to eliminate the root causes of hunger and poverty, including through the progressive realization of the right to adequate food� (FAO 2009b). 6 CHAP T E R 3 — T HE ROL E OF F INA NCE IN A GRICULTUR A L D EVELOPMENT A ND RURA L POV ERTY A L LEV IATION Chapter 3: tHe RoLe oF FInAnCe In AGRICULtURAL DeVeLoPMent AnD RURAL PoVeRtY ALLeVIAtIon Historically, economists have held strikingly different views the productivity, profitability, and sustainability of smallhold- about the importance of the financial system for economic ers. The report recognized the challenges and argued for growth (Levine 1997). some believed the financial sector financial innovations that put smallholders on a ladder of merely responds to economic development, adjusting to ascending financial market access flowing up from MFIs to changing demands from the real sector. Others believed commercial lenders. that financial systems play a crucial role in alleviating mar- ket frictions and hence influence savings rates, investment Financial constraints would be expected to be most con- decisions, technological innovation, and long-run growth straining in regions trying to achieve rapid agricultural growth. rates. Recent research has concluded that finance matters, The agricultural potential of the African Guinea savannah and an important contribution of finance is the provision of was compared to similar areas in Brazil and Thailand that credit to the most promising firms (Demirguc-Kunt, Beck, developed rapidly (World Bank 2009a). These regions share and Honohan 2008). However, research has not provided a medium to high agricultural potential but also face significant simple strategy or single road map for developing the finan- constraints in the form of infertile soils and variable rain- cial sector. fall. The study concluded that making the African region’s agriculture competitive depends on getting policies right, What about credit and agricultural and rural development? strengthening institutions, and increasing and improving Advocates of the old supply-leading agricultural credit viewed investments in the sector. Among other things, the region credit as an integral component of input packages designed requires macroeconomic policy reforms, land policy reforms for Green Revolution crop varieties. Actually establishing that enable smallholders to obtain access to land, scaled-up a causal link between credit and agricultural development, public investments in agriculture, greater private investment however, has proven difficult. For example, one comprehen- through public-private partnerships, and institutional reforms sive study looked at the investment decisions of government, to make markets work better and create self-sustaining rural financial institutions, and farmers and the effects on agricul- financial systems. tural investments and output in India (Binswanger, Khandker, and Rosenzweig 1993). The study covered the 1960s and This long list of recommendations demonstrates the com- 1970s, a period when India aggressively expanded its finan- plexity of implementing agricultural development. Modern cial system into rural areas. The authors concluded that the agriculture requires large amounts of purchased inputs and availability of credit was more important than subsidized investments in on-farm and off-farm storage, refrigeration, interest rates, and the expansion of banking had a larger im- processing, and transportation. Financial services are only pact on output through expanding fertilizer use than through one condition for success, and the successful provision of fi- increased investments. Bank expansion was greatly aided by nance requires supportive policies and infrastructure. As dis- government road investments and reduced transaction costs covered in the old agricultural credit paradigm, cheap loans for banks and farmers. cannot substitute for appropriate technology, input supplies, and access to remunerative markets. The authors of the World Development Report 2008 (World Bank 2007) argued that financial constraints are more per- Implications. Clearly the mechanisms that contribute to em- vasive in agriculture than in other sectors. Providing broader ployment growth and increased production and income, and access to financial services—credit, savings, insurance, and the role that finance plays in these changes, require greater transfer services for remittances—and reducing exposure to understanding. Financial services to help the poor to manage uninsured risks requires financial instruments that improve their money may produce considerable benefits. The demand for operating credit is likely to be much greater for larger, S U B S I D I E S A S A N INST RUME NT IN AGRICULT URE FINA NCE: A REV IEW 7 more commercially oriented farms that use large amounts of purchased inputs than it is for small subsistence farmers using fewer inputs and little hired labor. Therefore, large farms’ access to credit may be most important for aggregate food production. Larger farms are normally considered more creditworthy and have more ready access to credit from commercial and development banks. subsistence farmers, on the other hand, are more likely to face difficulty in obtain- ing formal credit and, therefore, may realize relatively large benefits from small loans, provided they can get access to land, inputs, and markets. A logical hypothesis, therefore, is that improved access to large loans with attractive terms and conditions will be more important for commercial food production than for livelihoods and poverty alleviation. The experience of microfinance sug- gests that microloans are highly useful to smooth household consumption, improve food security, and meet household emergencies and education expenses. The challenge is to develop sustainable financial systems that help the poor manage their money as well as provide larger loans to fi- nance value chains that produce greater impacts on output, employment, and income, as discussed in the next section. 8 C H A PTER 4 — TH E CH A LLENGE OF D EVELOPING CR ED IT M A R K ETS Chapter 4: tHe CHALLenGe oF DeVeLoPInG CReDIt MARKets Allegations of market failure are often used to justify inter- Externalities refer to the spillover costs or benefits, unin- ventions in and subsidization of credit markets. Decision tended consequences, or unintended side effects associated makers in low-income countries also advocate credit sub- with market transactions. An example occurs when mining sides to offset subsidies provided to famers in advanced pollutes water supplies, making it unusable for human con- nations. Advocates for the poor propose credit subsidies sumption or irrigating crops. to improve income distribution. It is important to under- stand the standard market failure rationale in order to Governments can intervene in many ways once they deter- evaluate if public subsidies and investments for develop- mine that market failure actually exists. For example, they ing agricultural credit markets are economically justified. can use moral suasion or rules and regulations to influence producers and consumers to act in socially desirable ways. Financial incentives can be used to induce desired behavior, 4 1 PRIVAte MARKets AnD MARKet FAILURe such as adoption of new technology. Through financial sup- port, governments encourage activities, such as research Economic theory demonstrates that when marginal private and extension, that farmers cannot finance themselves. costs are equated to marginal private benefits in the mar- Finally, governments intervene more directly in markets ketplace, marginal benefits and costs for all of society will when they replace or directly compete with private produc- also be equated. Market failure is said to occur when the ers in the production and distribution of goods and services. market fails to allocate resources efficiently. When the indi- state-owned development banks, savings banks, and insur- vidual’s pursuit of pure self-interest leads to results that are ance companies are common examples in financial markets. inefficient, improvements are theoretically possible from the societal point of view. How does market failure arise? Economic theory recognizes 4 2 InFoRMAtIon AsYMMetRIes AnD several categories of market failure, including imperfect com- ContRACt enFoRCeMent PRoBLeMs In CReDIt MARKets petition, imperfect or asymmetric information, public goods, and externalities (see, for example, Arnold 2005). Imperfect In an ideal credit market, loans are traded competitively competition occurs in monopolies or oligopolies when there and interest rates are determined by supply and demand. are only a few sellers with price-setting ability. Lack of com- Theoretically, the best investment opportunities will be petition in financial markets could be an example. Imperfect financed because individuals with the best investment op- information occurs when consumers, producers, or both do portunities are willing to pay the highest interest rates. In not know the true costs and benefits associated with a good practice, however, credit markets diverge from an ideal mar- or activity. Asymmetric information exists when one party to ket because information is imperfect and loan contracts are a transaction (such as a lender) has less information than the difficult to enforce (Besley 1994). other party (the borrower). A lender’s willingness to lend depends on having sufficient Investment in agricultural research has long been justified information to evaluate the borrower’s reliability, capacity to as a public good because a single farmer or group of farm- repay, and intention to use borrowed funds wisely. A bor- ers cannot finance a socially optimum amount of research. rower may claim to be unable to repay (owing to crop fail- Even if they had the resources, they could not control the ures, livestock losses, or low prices) when in fact he or she is free-rider problem in which other farmers benefit, say, from unwilling to repay (if there are insufficient sanctions against using a new technology without paying for its development. default). The farmer-borrower may promise to work diligently S U B S I D I E S A S A N INST RUME NT IN AGRICULT URE FINA NCE: A REV IEW 9 to repay, but if default occurs, it is difficult for the lender to farmers. The complicated environmental, material, and pro- determine if it was due to bad luck, farmer mismanagement, duction features of agriculture that inhibit demand for and or unwillingness. Faced with insufficient information and un- supply of credit and insurance are well known and are sum- certainties about repayment, lenders may simply choose to marized here (Binswanger and Rosenzweig 1986). deny loans. Lenders may try to overcome contracting prob- Agriculture has important spatial and risk characteristics. lems by requiring substantial loan collateral, finding effective Land is immobile; production is dispersed; and transport, collateral substitutes, charging higher interest rates, lending communication, and travel costs are high. variations in tem- only to borrowers with proven track records of repayment, or perature, rainfall, and sunshine lead to seasonality of produc- lending only to borrowers with diversified and reliable cash tion, which creates a demand for seasonal credit to bridge flows. But such strategies pose limitations and usually ex- the gaps between receipts and expenditures. Parcels of ad- clude the poorest farmers who are the frequent concern of jacent pieces of land have similar weather conditions, leading policy makers. to covariate yields. several kinds of risks and uncertainties Analysts have used these conceptual arguments to under- exist: (1) yield risks due to weather, diseases, and insects; (2) stand bank reluctance to lend in the face of large unmet de- market price risks due to local and global weather and mar- mand for loans. One analysis found the ratio of liquid assets ket variations; (3) timing uncertainties due to farm-specific to total bank deposits averaged 19 percent for 5 developed weather variations; (4) uncertainties in the timing of repairs countries compared with 45 percent for 35 developing coun- and reinvestments; and (5) illness, accidents, and other life- tries, a difference that indicated ample opportunities for ad- cycle risks. ditional lending. The primary reasons for reluctance to lend These characteristics have important implications for credit were identified as higher reserve requirements for banks due markets. seasonality and synchronic timing imply that bor- to greater macroeconomic risk and volatility; a legal and regu- rowers prefer to borrow at the same time at planting and latory environment in which it is difficult to enforce loan con- repay at the same time at harvest, thereby creating liquidity tracts and foreclose on collateral; widespread availability of management problems for financial institutions. Covariance low-risk, high-yielding government bonds; substantial asym- of yields implies covariance of default risks, and lenders must metric information so lenders know little about prospective carry large cash reserves to meet depositor withdrawals at borrowers; and inadequate banking skills for assessing risk times when borrowers may be slow in repaying or unable to and managing loans. Implementing reforms is daunting and repay. High communication costs make it difficult to man- time-consuming. age the large banking network needed to reduce covariate Lack of financial access is most serious in Africa, but finan- risks. When information costs are high, in the absence of cial sector development programs there have produced insurance, only small loans will be made without collateral. disappointing results. Low levels of financial intermediation, The most valuable forms of collateral are assets that (1) are relatively high interest rates, wide intermediation spreads, easiest to appropriate in the case of default, (2) do not easily and substantial bank profitability persist. The causes were lose their value, and (3) have high use value to the borrowers identified as currency and macroeconomic uncertainties; so they will not want to easily part with them. Where there is high government demand for loan funds; lack of competition; an active land market, land facilitates access to formal loans, relatively small bank sizes; and contractual problems includ- as it tends to have the highest value as collateral. Collateral ing weak creditor rights, compromised courts, a deficient substitutes become important for lenders when borrowers insolvency framework, and a general disrespect for contracts possess few high-quality assets to pledge as loan collateral. (Honohan and Beck 2007). In these environments, many The characteristics of agriculture also have important implica- governments consider interest rate and other subsidies the tions for insurance markets. Expected and actual yields dif- instrument of choice. fer enormously by field and farm, and average plot sizes are small, implying high costs for loss assessments. Insurance 4 3 sPeCIAL CHALLenGes FoR CReDIt may induce moral hazard problems associated with poor MARKets to seRVe RURAL AReAs AnD animal and plant husbandry. Yield risk introduces covariance FARMeRs risk, and large insurance payouts require holding large finan- cial reserves against losses. Typhoons, hurricanes, droughts, In addition to the general problems discussed, special chal- and other catastrophic events are important sources of lenges exist for financial systems serving rural areas and 10 C H A PTER 4 — TH E CH A LLENGE OF D EVELOPING CR ED IT M A R K ETS weather risk in many countries. The smaller the country, the more difficult it is to diversify risks across a large area to overcome covariance problems.7 In the absence of insur- ance, individuals employ insurance substitutes such as cash reserves, traditional technologies, conservative input levels, and social ties that provide aid in the event of emergencies. Implications. The complex environmental, material, and pro- duction features of agriculture inhibit the demand for and supply of credit and insurance. Not surprisingly, efforts to increase formal credit supplies have had a spotty record and quick fixes have not worked. Most successes have occurred when careful long-term institutional development has been at the heart of the strategy. Lessons learned from microfinance provide insights into success factors and are discussed below. 7 This assumes that insurers and reinsurers operate only within that country and cannot diversify their risks internationally. S U B S I D I E S A S A N INST RUME NT IN AGRICULT URE FINA NCE: A REV IEW 11 Chapter 5: tHe PARADIGM sHIFt In DeVeLoPInG CReDIt MARKets There is a long tradition of viewing cheap agricultural credit and controlling interest rates on retail loans to priority groups as a development tool. From the 1960s to the 1980s, sub- and sectors. One-size-fits-all credit models were created for sidized, directed agricultural credit programs were common farm lending rather than relying upon local institutions and in top-down government and donor policies and programs. credit officers to design products for individual borrowers. Unfortunately, attempts to resolve conditions viewed as mar- ket failure often ended up as government failure. The key Generally, the results from this paradigm failed to meet ex- features of these failures are summarized here to contrast pectations. Increased lending may have contributed to some with the new financial systems paradigm that emerged in short-term increases in food supplies but did not lead to sus- late 1980s and guided the microfinance revolution.8 tainable credit supplies. Low interest rates created an excess demand for credit, and rationing logically tended to favor richer and more politically powerful farmers.9 High transac- tion costs for loans coupled with long delays reduced the 5 1 tHe oLD-PARADIGM APPRoACH advantage of formal loans relative to informal sources. Low The old-paradigm, directed agricultural credit approach em- interest rate margins and poor loan recovery undermined ployed in many countries had several typical features. At the the financial sustainability of institutions. Many institutions farm level, the approach was often implemented without failed, and others required bailouts and repeated recapitaliza- careful analysis of the nature and causes of the supposed tions. A culture of nonrepayment developed among borrow- credit market failures. Interventions were considered neces- ers, especially when loans were viewed as coming from the sary to overcome the risk aversion of conservative lenders government. Most old-paradigm approaches ended by the who failed to provide credit that farmers needed to purchase 1980s. Government failure replaced market failure because inputs for adopting Green Revolution production packages. the directed credit programs generally failed to resolve the Moreover, loans at artificially low interest rates were justi- screening, incentive, and enforcement problems in rural fied to accelerate farmer adoption of these packages. Cheap lending (Hoff and stiglitz 1990). formal credit was viewed as a way to introduce competition for usurious moneylenders and reduce farmer dependency In reality, the picture can be complex, as the example of on informal sources. the Thai Bank for Agriculture and Agricultural Cooperatives (BAAC) shows. The bank was established in 1966 and is At the national level, it was believed that supply-leading fi- still licensed as a special financial institution (sFI) and not a nance could accelerate the economy and that credit could be full-service commercial bank. Today it is one of the country’s force-fed by imposing lending targets on financial institutions 10 largest financial institutions and serves most farmers in and providing incentives for extending bank branches into the country directly through group and individual loans or rural areas. specialized agricultural development banks and through loans to farmer associations and cooperatives. It cooperatives were created to supplement commercial banks has participated in government credit programs and been that resisted serving rural areas. Interest rates were reduced subject to controlled interest rates but has operated with by subsidizing wholesale loans made to first-tier institutions relatively modest subsidies. It has maintained a fairly suc- cessful firewall against the most distortionary governmental 8 some of the most comprehensive and accessible publications initiatives, but for some time it has been affected by gov- of the vast literature that analyzed the debates and presented ernment debt relief and flood assistance programs. This the evolution in thinking about agricultural credit include Donald (1976); von Pischke, Donald, and Adams (1983); Adams, Gra- ham, and von Pischke (1984); World Bank (1989); Yaron, Benja- 9 Gonzalez-vega (1984) explained this result as a logical outcome min, and Piprek (1997); and Conning and Udry (2007). of the Iron Law of Interest-Rate Restrictions. 12 CH A PTER 5 — TH E PA RA D IGM SH IFT IN D EVELOPING CR ED IT M A R K ETS Village market in Rangpur, northwest Rice paddy in India (Photo: Mark Wenner) Bangladesh (Photo: Michael Hamp) situation is refl ected in a 20 percent provision against bad 5 2 eMeRGenCe oF A neW PARADIGM debts, as evidenced by a 2009 IFAD-Asian Pacifi c Rural and By the late 1980s, the accumulated criticisms of the old para- Agricultural Credit Association supervision mission (see also digm combined with the emerging success of microfinance siamwalla et al. 1990; Yaron, Benjamin, and Piprek 1997; and led to a rejection of the directed credit approach and the Meyer and Nagarajan 2000).10 The Bank Pertanian Malaysia, emergence of a new financial systems paradigm.11 The term Banrural s.A. in Guatemala, the Unit Desas of Bank Rakyat “financial system� covers all financial institutions, financial in Indonesia, and the Agricultural Bank of Mongolia are ad- markets and instruments, the legal and regulatory environ- ditional examples of state-owned rural fi nancial institutions ment, and financial norms and behavior. Building the system that successfully reformed by switching from subsidized to requires efforts at three levels: more market-driven organizations. ƒ micro—understanding the financial needs and behav- Although this paradigm has been largely replaced, important iors of different clientele, building financial institutions, remnants remain. In 2004, nearly 40 countries reported hav- and creating financial products and services; ing some form of interest rate ceilings, and several had re- ƒ meso—creating the infrastructure needed for financial cently introduced interest rate restrictions (Helms and Reille intermediation services; and 2004). Interest rate ceilings can discourage lenders from ƒ macro—creating conducive national policies and making small loans with high operating costs. India liberal- strategies, complementary nonfinancial services, and ized many aspects of its financial system but still has a quota a supportive enabling environment. requiring that 18 percent of bank lending must go to agricul- tural services, and it only recently lifted interest rate ceilings Key elements of the new paradigm include the following: for small loans. Poorly performing agricultural finance institu- tions still exist, and several are located in North Africa and ƒ a broadened view of rural finance that includes financ- the Near East. Although MFIs and most commercial banks in ing of farming and rural nonfarm activities; the region recover almost all of their loans, agricultural banks collect only 40–70 percent of their loans (Mustafaet al. 2010). 11 The ideas summarized here are drawn largely from Yaron, Ben- jamin, and Piprek (1997); FAO/GTZ (1998); World Bank (2003); and IFAD (2010). This new approach was formally incorporated in the policies of the international agencies in the 1990s. A World Bank report briefl y summarized the specifi c documents within 10 see reports in the country’s leading English-language newspa- the CABFIN institutions that demonstrate adoption of the new per, The Bangkok Post, in May and June 2010. paradigm (World Bank 2003). S U B S I D I E S A S A N INST RUME NT IN AGRICULT URE FINA NCE: A REV IEW 13 ƒ recognition of the importance of local savings mobilization; ƒ belief that market interest rates for both savings and credit reinforce market discipline of both financial institutions and clients; ƒ the granting of loans in response to demand rather than supply targets; ƒ the evaluation of financial institutions on the basis of viability rather than simply loans disbursed; ƒ recognition that successful finance is dependent on favorable macroeconomic, agricultural sector, and financial sector policies as well as an appropriate legal framework for private sector development; ƒ a perception of informal finance as complementary rather than usurious and harmful; ƒ an understanding that financial sector reform is es- sential for improved performance and wider outreach of financial institutions; and ƒ a role for donors in helping create a favorable policy environment, improving the legal and regulatory framework for rural financial markets, building insti- tutional capacity, and supporting innovations to lower transaction costs and improve risk management. Implications: The new paradigm reversed the strong focus on dispersing cheap credit and focused attention instead on creating sustainable institutions, treating borrowers and sav- ers as clients rather than beneficiaries, developing products that clients demand, and pricing products and services to cover costs and risks. Long-term relationships with clients were encouraged through stepped loans that are small ini- tially and are gradually increased with subsequent loans. Donor agencies reduced the use of credit lines in favor of grants, loans, and technical assistance to assist with the de- sign of appropriate products, institutions, and policies. The new paradigm contributed to the success of the emerging microfinance industry. 14 CHAP T E R 6 — T HE MICR OFINA NC E IND US TRY: S UCC ES S FUL A PPLICATION OF TH E NEW PA R A D IGM Chapter 6: tHe MICRoFInAnCe InDUstRY: sUCCessFUL APPLICAtIon oF tHe neW PARADIGM Asian microcredit experiments began in the mid-1970s, ex- Many institutional models and financial delivery systems for panded in the 1980s, and exploded thereafter. The history microfinance have emerged (Zeller 2006). Credit unions, co- of microfinance, however, starts long before the 1970s with operatives, village banks, specialized microbanks, NGOs, and the creation of the first pawn shops in Italy in the 15th cen- commercial and development banks all provide microfinance tury and the large expansion of cooperative movements in services. Many are new start-ups, but existing institutions the 19th century in Europe and Canada (see Helms 2006, 3). also began offering microcredit when profitability was dem- Thousands of microfinance programs exist today, millions onstrated. Where widespread banking networks exist, as in of clients are being served, financial services are evolving India, informal self-help groups are created and linked with beyond just loans, and social and commercial sources now financial institutions (Nair 2005). surpass donors and governments in funding the most suc- cessful programs. Many clients are as poor or poorer than those targeted in the old-paradigm agricultural credit pro- 6 1 GUIDeLInes FoR DeVeLoPInG grams. This section presents a highly simplified overview MICRoFInAnCe of the microfinance revolution and its role in agricultural The Consultative Group to Assist the Poor (CGAP) was cre- credit.12 ated in June 1995 to increase the resources for microfinance and deepen the success of the pioneer institutions. Donor Many MF programs started as bottom-up attempts by members endorsed Key Principles of Microfinance to guide nongovernmental organizations to assist the poor in sharp their support for the industry. CGAP created guidelines at the contrast to the earlier top-down efforts to expand agricul- micro, meso, and macro levels to address the appropriate tural credit. An important factor in MF success has been the use of external support and subsidies without undermining creation of innovative lending technologies and institutional the growth of the private sector (CGAP 2006). designs (Gonzalez-vega 2003). Joint liability group lending proved to be a major breakthrough to resolve the lack of The guidelines include recommendations such as the collateral by the poor. Groups have incentives to undertake following: good borrower selection, monitoring, and contract enforce- ment. Peer pressure encourages prompt loan repayment ƒ verify that credit is actually needed (the main con- because the entire group is denied future loans if any mem- straints may lie elsewhere, such as in weak infra- ber defaults. small loans reduce risks for both lenders and structure, poor production technology, limited market borrowers. Frequent loan payments reduce risks and help access). lenders efficiently monitor clients. High interest rates are ƒ Avoid using microcredit merely as a resource transfer charged to cover operating costs and potentially high loan mechanism for high-risk groups when other methods losses. may be more efficient (such as safety net programs for vulnerable groups). ƒ Provide flexible grant funding to cover research, product refinement and development, and technical 12 There is a vast MF literature. sources that describe important assistance for capacity building. aspects of the industry’s evolution and key features of institu- ƒ support financial service providers in progressively tions, products, clients, and policies include Otero and Rhyne (1994), Hulme and Mosley (1996), Ledgerwood (1998), Mor- intermediating commercial funds and deposits. duch (1999a), Robinson (2001 and 2002), Zeller and Meyer ƒ Allow financial service providers to set their own (2002), Armendariz and Morduch (2005), Hartarska and Holtman (2006), Meyer and Nagarajan (2006), and Zeller (2006). pricing policies, encourage them to be transparent, S U B S I D I E S A S A N INST RUME NT IN AGRICULT URE FINA NCE: A REV IEW 15 and avoid compelling them to charge below-market magnitude of the many direct and indirect subsidies granted interest rates on loans or rates lower than necessary to the Grameen Bank in Bangladesh has been carefully calcu- to cover costs in the medium term. lated. For the period 1985–96, it was estimated that the bank ƒ Price loans to financial institutions at commercial or would have needed to raise nominal rates on ordinary loans near-commercial rates to avoid undermining incen- from 20 to 33 percent to become free of subsidies (Morduch tives to mobilize deposits or tap other local sources of 1999b). But like many MFIs, Grameen worked toward self- capital. sufficiency, eventually implemented massive savings mobili- ƒ Phase out grants and subsidized loans gradually as zation, and tapped commercial sources. Likewise, estimates local and international commercial capital markets and have been made of the large amount of subsidies granted to domestic savers become viable sources of capital. BAsIX, one of the early MFIs that stimulated development of ƒ Promote transparency and accountability through the industry in India (sinha 2008). regular financial reporting and third-party performance The microfinance industry has become increasingly commer- assessments and ratings. cialized, and the large profits earned by the most successful ƒ support research and development on the use of MFIs have attracted huge investments by private and social technology for, for example, points of service, transfer investors. CGAP reported that 61 donors and investors had and payment mechanisms, and credit bureaus. committed $14.8 billion to microfinance as of December ƒ support interest rate liberalization through education 2008, of which $3.9 billion were new 2008 commitments. and advocacy. At least $3 billion was disbursed in 2008, two-thirds of it by ƒ Avoid direct provision of credit services by a govern- investors. Eighty-four percent of the total committed fund- ment, government-mandated portfolio quotas, direct- ing went to finance the portfolios of retail institutions. Total ed credit, borrower loan guarantees, or operational funding for capacity building at the retail level amounted to subsidies. Exceptions may be appropriate for well-run $1 billion (CGAP 2009b). programs that serve hard-to-reach populations. ƒ Encourage adaptation of policy and legal frameworks that increase competition and improve the quality of 6 2 MICRoFInAnCe PenetRAtIon Into RURAL services available for poor people. AReAs AnD AGRICULtURe ƒ Build the capacity of key government staff in minis- Until recently, microfinance mostly involved microcredit. The tries of finance and central banks. high cost of making and recovering small loans required high volumes per loan officer, so operations tended to involve Financing of the industry. Huge amounts of funds have been group lending in urban and peri-urban areas or rural areas spent to develop the industry. During the 2000s, donors spent with high population densities. Frequent loan payments, close to $1 billion a year in MF programs (CGAP 2006). The often collected in weekly or monthly group meetings, were considered essential to maintaining financial discipline in the absence of physical collateral. This approach, and the use of manual bookkeeping, led to highly standardized one-size-fits- all loans. Borrowers with frequent sources of cash inflows, often engaged in retail or petty trading, were naturally at- tracted to the product, so penetration into rural areas and to farmers with seasonal cash flows was limited. This situation is particularly clear in Bangladesh, a leading MF country. It has excelled in highly standardized loans but has lagged behind other countries in serving farm households. In 2008, some 25–30 million borrowers had access to micro- credit and 6–7 million of these people were engaged in crop farming, but only 1–1.5 million borrowed loans specifically designed for seasonal or investment lending in agriculture (Alamgir 2009). The MFIs reportedly financed the growth of poultry and livestock and dairy production, but loans were Women and microfinance (Photo: Lotte!) 16 CHAP T E R 6 — T HE MICR OFINA NC E IND US TRY: S UCC ES S FUL A PPLICATION OF TH E NEW PA R A D IGM not tailored for seasonal agriculture. several problems ex- The MFI successes in agricultural lending involved a combi- plain the slow development of products for farmers. Most nation of practices such as treating the borrower as a firm MFIs work largely with female clients whereas crop farming household, basing loan size and repayment schedules on tends to be done by men; most still use manual bookkeep- the cyclical cash flow of the entire household, and delinking ing so standardized annual loans are easier to manage; loans loan payments from loan use. MFIs reduce lending risks by with periodic rather than lump-sum payments make it easier lending to borrowers with diversified cash flows, diversifying to monitor clients efficiently; agriculture is perceived as risky portfolio risks by lending to farmers in different regions with so most programs prefer clients less affected by weather; different crop and livestock enterprises, linking credit with and considerable training and decentralization of decision area-based index insurance, and insulating themselves from making would be required to develop flexible loans geared political interference (CGAP/IFAD 2005). to farmers’ cash flows. Most NGO-MFIs have neither the commitment nor capacity to enter into this market segment MFIs also have to deal with negative political events. For ex- (BWTP Network 2009). The International Fund for Agricultural ample, some of the Nicaraguan MFIs mentioned increased Development (IFAD) and the Asian Development Bank their share of lending for livestock but encountered repay- (ADB), however, have sponsored projects that support the ment problems when meat prices fell during the food and second-tier funding institution (PKsF - Palli Karma-sahayak fuel crises. The No Pago movement in that country is push- Foundation) to encourage more experimentation with loans ing for a debt moratorium that will damage MFIs (Campion, designed for seasonal agriculture. Ekka, and Wenner 2010). Political problems have also affect- ed Bolivian MFIs. For example, the Agrocapital Foundation, Although there are no comprehensive data, there are many founded in 1992, was unique in that its clients had registered examples demonstrating that MFIs have found methods to land titles, and it was the first NGO in Bolivia to grant loans deal with the costs and risks of agricultural lending (CGAP/ to farm and nonfarm businesses using individual lending IFAD 2005). As of the end of 2006, 20 MFIs in Nicaragua re- (Alvarado and Galarza 2003). In 2007, the government imple- ported that 47 percent of their portfolios were in agriculture mented controversial changes in land laws that prohibited and forestry.13 In 2007, 37 MFIs in Uganda reported that 38 owners with less than 50 hectares of land from using it as percent of their total portfolios were in agricultural loans.14 loan collateral. Agrocapital added a village banking model to The Economic Credit Institution in Bosnia and Herzegovina, reach smaller-scale borrowers and substitute for loan collat- the Banco del Estado de Chile, small Farmer Cooperatives, eral. These legal changes, along with public unrest in rural Ltd. in Nepal, the Cresol and sICREDI systems of savings areas over land reform and other issues, contributed to a de- and loan cooperatives in Brazil, Confianza in Peru, and sev- cline in Agrocapital’s agricultural lending. By the end of 2008, eral community-managed village savings and credit organiza- the agricultural share of the loan portfolio had fallen to about tions (CvECAs) in parts of West Africa developed innovations 30 percent (Fitch Ratings 2009). to serve agriculture. One of the leading promoters of individualized lending, IPC 6 3 InnoVAtIons AnD PRosPeCts FoR FUtURe – Internationale Projekt Consult, a German consulting firm, is AGRICULtURAL LenDInG credited with initiating agricultural lending by Caja Los Andes Many innovations are taking place, so prospects are good in Bolivia, Uganda’s Centenary Rural Development Bank, for further reductions in agricultural lending costs and risks. and Calpia in El salvador (CGAP/IFAD 2005). A comparative successful innovations will be especially important for analysis of Calpia’s rural and urban portfolios showed that dealing with the small transactions and high communica- credit officer efficiencies, operating costs, and portfolio qual- tion costs that impede financial development. MFIs are ity were similar. These data demonstrated that with careful updating their management information systems (MIss), client selection and appropriate lending technologies the reducing the constraints imposed by manual bookkeep- rural portfolio could perform as well as the urban (Buchenau ing systems. Technological innovations to speed transac- and Meyer 2007). tions and reduce transaction costs involve using the vast emerging mobile phone networks in emerging economies for deposit and credit transactions. Earlier experiments in- volved debit, credit, and smart cards; electronic passbooks; 13 REDCAMIF, “Microfinanzas en Centroamerica,� Bulletin No. 6, April 2007. remote transaction systems; and point-of-sale devices 14 Bank of Uganda, unpublished survey data, 2007. (Nagarajan and Meyer 2005). The high cost of opening bank S U B S I D I E S A S A N INST RUME NT IN AGRICULT URE FINA NCE: A REV IEW 17 Mobile banking, Mozambique (Photo: Renate Kloeppinger) Sign for mobile banking, Uganda (Photo: Renate Kloeppinger) branches may be reduced by mobile banking, and Brazil is integration with formal financial institutions.15 The first prod- the leader in implementing a massive bank agent system, uct launched was savings accounts. Data suggest that cell including about 150,000 “banking correspondents,� many of phones may already have important community-level effects whom specialize in credit. More than 50,000 are authorized on money transfers, bill paying, and business expansion to open accounts and handle deposits (CGAP 2010a). Cell (Plyer, Haas, and Nagarajan 2010). phone banking is exploding, especially in Africa where land lines are especially sparse. M-PEsA in Kenya is leading the Opportunity Bank in Malawi used several technologies to way by facilitating the transfer of funds through cell phones. expand rural outreach but has faced problems in trying to Demand has far exceeded expectations, and an important develop mobile banking when an appropriate cell phone constraint has been the creation of a network of certified partner does not exist.16 CGAP summarized alternative ap- agents (Eijkman, Kendall, and Mas 2010; Lonie 2010). In May 2010, safaricom (the mobile operator behind M-PEsA) 15 “M-KEsHO in Kenya: A new step for M-PEsA and mobile bank- and Equity Bank launched M-KEsHO, a co-branded suite of ing,� posted by Ignacio Mas, May 27, 2010, on the FAI Blog financial products that will open the door to more cell phone http://financialaccess.org/blog downloaded on July 9, 2010. 16 For information about the several efforts of Opportunity Bank to extend rural outreach, see McKay (2009) and Opportunity Inter- national (2008, 2009). Cambodia (Photo: Renate Kloeppinger) Kenya (Photo: Calvin Miller) 18 CHAP T E R 6 — T HE MICR OFINA NC E IND US TRY: S UCC ES S FUL A PPLICATION OF TH E NEW PA R A D IGM proaches for integrating microfinance services with mobile 6 4 InsIGHts GAIneD FRoM MICRoFInAnCe phone banking in circumstances when cell phone partners ABoUt tHe IMPACt oF FInAnCe do and do not exist. Improved customer service, reduced The current debate about the impact of microfinance raises costs of offering savings accounts, a potential to increase the questions about the role of credit in poor households and number of savings customers, and possible cost reductions businesses. The common narrative has been that microloans large enough to warrant interest rate reductions, but not a help poor households create employment and start or ex- large expansion in new credit customers, seem to be the pand businesses that lift them out of poverty. Early impact logical outcomes of these integrations so far (Kumar, McKay, studies tended to support those results, but the frequently and Rotman 2010). cited gold standard evaluations (such as Khandker 1998) have been challenged recently, and more robust impact Community and member-based institutions, including sav- evaluation techniques are advocated (Duflo, Glennerster, and ings and credit cooperatives and credit unions, may be the Kremer 2008). most logical institutions to provide cost-effective financial services in remote areas. In these organizations, members Microcredit seemed to produce some effects on business have the responsibility for owning, managing, and operating outcomes and the composition of household expenditures the financial institutions in addition to being the main or only among randomly selected households in the slums of customers. Often they are built on the principles of informal Hyderabad, India, where an MFI opened branches (Banerjee rotating savings and credit associations and accumulating et al. 2009). Existing business owners appeared to use mi- savings and credit associations that are found in many coun- crocredit to expand their businesses, and business profits tries. CvECAs are being formed at minimum cost as savings- increased. Households with a high predicted propensity to first institutions in many African countries.17 Member-owned start a business reduced nondurable spending, presumably institutions often suffer from poor governance and have op- to finance an even bigger initial investment than could be erations limited to small geographic areas that are subject to financed with just the loan. covariate risks. External assistance is often provided to start and strengthen these organizations, but credit lines from The issue of impact of microfinance is far from settled, but external funds for on-lending have often damaged their self- the results of new studies suggest a need to reexamine help discipline and induced rent-seeking behavior (Hirschland the traditional view of how expanding credit supplies will et al. 2008). benefit the poor. similar caution is needed when advocat- ing the expansion of credit for rural areas and for agriculture. MFIs are also actively seeking ways to participate in agricul- Enthusiasm for expanding credit supplies has surpassed our tural value chain financing, which may open new possibilities understanding of how target groups actually use finance and for strengthening financial operations by linking with nonfi- what this implies for the design of products and programs. nancial groups and institutions (Miller and Jones 2010). In Comparatively high interest rates of MFIs reduce the positive India, BAsIX is famous for learning that microcredit alone, impact of loans at the borrower level and can also influence which it started in 1996, failed to produce a major impact on the composition of economic activity in a given catchment the lives of the rural poor. In 2002, therefore, it broadened area toward retail and other trading activities. But this is not its approach to include enterprise and institutional develop- a call for reducing or subsidizing interest rates. Rather, the ment services. It also, however, reduced its seasonal crop institutional inefficiencies of MFIs should not be passed on lending from more than 21 percent of its overall portfolio in to borrowers. 2001/02 to less than 2 percent in 2006/07, while the allied agriculture (mostly animal husbandry) category rose from 22 Implications. Microfinance thrived by following a market- to 33.5 percent. The share of nonfarm microenterprise loans oriented approach that was in sharp contrast to the subsi- also increased to 53.4 percent compared to 45.3 percent in dized directed agricultural credit paradigm. As a result mil- 2001/02 (sinha 2008). lions of poor people have obtained access to microloans, and the industry is beginning to supply other financial services. some MFIs have made significant inroads into agriculture and rural areas, but generally the industry has been slow to design products and methodologies to fit the seasonal 17 Data for more than 20,000 savings groups show a range of cost between $15 and $85 per member assisted with an average cash flow patterns of farm households. Managing the costs of $31.80 (personal correspondence with Hugh Allen, July 29, and risks of agricultural lending has been challenging. Major 2010). see www.savingsgroups.com for comparative data. S U B S I D I E S A S A N INST RUME NT IN AGRICULT URE FINA NCE: A REV IEW 19 developments in the application of technology to reduce the cost of small financial transactions, the use of banking corre- spondents, and greater linkages with nonfinancial groups and institutions offer possibilities for greater MFI penetration into agriculture. The integration of financial services and mobile phones offers interesting possibilities where the appropriate conditions exist. The current debate about the impact of mi- crofinance demonstrates the need for better understanding of how target groups use additional credit and the impact to be expected from improved access to loans. 20 CHAP T E R 7 — C R ED IT D EMA ND , RATES OF RETURN, A ND INTERES T RATE S ENS ITIV ITY Chapter 7: CReDIt DeMAnD, RAtes oF RetURn, AnD InteRest RAte sensItIVItY Improved understanding about the demand for and use of transaction costs may have affected their decision to apply agricultural credit is important in developing effective prod- (Magill and Meyer 2005). ucts, institutions, projects, and policies. The rapid growth of microfinance suggests there may be a large unmet demand some MFIs have experienced high dropout rates—as high as for agricultural loans, but three issues need to be anticipated. 13 to 60 percent a year in East Africa. some dropouts occur First, there may be a tendency to overestimate demand, as because of success—that is, the borrowers make progress has occurred with microfinance. second, farmer demand and are able to self-finance their financial requirements or for loans may be limited if the interest rates charged are as graduate from an NGO to a commercial bank or credit union. high as MFIs need for sustainability in making small micro- some dropouts simply want to take a break before borrowing enterprise loans. This issue raises the question of the rates the next loan. Others fail when they have difficulty repaying of return in farming relative to the cost of borrowing. Third, and refuse to borrow a second time or their group members there is the question of how sensitive borrowers are to in- do not want them to borrow again. some learn that stan- terest rates relative to other factors affecting demand. This dardized loans with frequent meetings and payments are too section provides a summary of research, often drawing from time consuming and inflexible to meet their needs (CGAP microfinance, on these three issues.18 2000). Evidence from Bangladesh supported the view that most dropouts occur because MFIs do not meet the needs of the market (Wright 2001). 7 1 DeMAnD FoR CReDIt A recent CGAP publication described how demand for micro- 7 2 RAtes oF RetURn In AGRICULtURe loans is often overestimated (Anand and Rosenberg 2008). Borrowers use credit for many purposes—working capital, Estimates derived from population estimates or number investment, consumption smoothing, education, and health of enterprises, the usual starting point for credit demand expenditures—so it is difficult to determine the impact estimates, need to be reduced for three reasons: (1) many of interest rates on demand. Moreover, credit demand is people simply do not want microloans; (2) some who might influenced by other factors, such as transaction costs and want loans are not considered creditworthy; and (3) people attitudes toward the risk of borrowing. Borrowers obtain who want and qualify for loans do not necessarily borrow cash from a variety of sources to make loan payments, but continuously. For example, the 17,000 microenterprises in the rate of return earned from enterprises is of special con- the Ecuador study mentioned reported that 85 percent of cern because it can affect both supply and demand. Higher both men and women respondents had not applied for a loan profits improve debt repayment capacity, so lenders are during the preceding 12 months. The greater the wealth of usually more willing to lend to farmers with more profitable the respondents, the higher the application rates, with the enterprises. Farmers may also be inclined to take risks and richest 10 percent (decile) applying at a 24 percent rate com- demand more credit when they expect to earn higher rates pared to 9 percent for the poorest decile. The overall success of return in their enterprises. rate for applicants actually receiving loans was a surprisingly high 97 percent so the rejection rate should not have discour- Analysts of the directed-credit paradigm (the old-paradigm aged applicants. Two thirds of the applicants were located approach detailed in section 5.1) often concluded that farm- within two kilometers of the lender so familiarity and low ers were more concerned about obtaining loans on time and in the correct amount than with low interest rates. This 18 several ideas presented in this section are discussed in more finding was a logical reaction to the high transaction costs detail by Karlan and Morduch (2010). S U B S I D I E S A S A N INST RUME NT IN AGRICULT URE FINA NCE: A REV IEW 21 and credit rationing attributed to low interest rates (Adams destroying a crop. A little medicine may save a sick cow, and von Pischke 1992). Borrowing transaction costs were and a small amount of fertilizer may increase crop yields in high as a natural result of credit rationing and inefficiencies in depleted soils. directed credit delivery systems (Meyer and Cuevas 1992).19 MFI interest rates today are low relative to moneylender Empirical studies of the productivity of credit use in agricul- rates but tend to be high relative to the rates for larger loans ture, however, have yielded mixed results. For example, one from commercial banks because of higher operating costs. study suggested that many Chinese farmers were credit In 2006, the median interest income for sustainable MFIs in constrained, but the use of additional credit was estimated the MIX Market, weighted by gross loan portfolio, was 26.4 to yield a low return, implying that much production credit percent of loans outstanding. The majority of MFI interest might actually be diverted to consumption purposes (Feder income goes to pay operating costs (salaries and other ad- et al. 1990). A study of small farmers in northern Peru pro- ministrative costs), which are about 60 percent of total MFI ducing rice, cotton, and corn concluded that relaxing credit costs (Rosenberg, Gonzalez, and Narain 2009). constraints would raise the value of output per hectare by 26 percent, a rate that should encourage borrowing (Guirkinger MFIs argue that they must charge relatively high interest and Boucher 2007). A study in southern Ghana estimated rates because of these high operating costs. They report, returns to capital for small farmers producing maize, cassava, however, that many clients thrive and expand their busi- and food crops for local markets and pineapple for export. nesses while paying interest rates of 20 or 30 percent a year, Real returns to capital were estimated at 250–300 percent or even higher. The high interest rates charged in informal for the new technology of pineapple cultivation and 30–50 credit markets are evidence that borrowers already pay high percent for well-established food crop cultivation (Udry and rates and thus value access to formal loans more than low Anagol 2006). interest rates (Rosenberg 2002). This argument is consistent with the view espoused by Nobel Prize winner Muhammad A recent Kenya study analyzed the profitability of fertilizer Yunus, that returns to access to credit are bound to be large use, an important issue because of high fertilizer prices fol- because the poor are starved for capital.20 But credit ob- lowing market liberalization. Kenyan farmers switch back and tained for business loans may be diverted because returns forth between using and not using fertilizer, and many never could be equally high for, say, investing in education.21 use it for maize production. The study tested the possibility that fertilizer and hybrid seed may increase maize yields on Do rates of return in agriculture justify paying the levels model farms but may not be profitable on small farms where of interest rates that millions of microentrepreneurs pay? conditions are not optimal (Duflo, Kremer, and Robinson Comparatively little analysis has been conducted in recent 2008). Yield increases on small farms due to fertilizer use years on rates of return earned in farming relative to interest were found to be in the range of the estimates found on rates for agricultural loans. Logic suggests that returns to model farms. The return for the most profitable quantity of capital should be high for farmers without access to credit. fertilizer (only 1/2 teaspoon of top dressing fertilizer per plant- For example, the returns from applying a small amount ing hole) was 36 percent over a season, or 69.5 percent on of insecticide should be high if it prevents insects from an annualized basis. Other levels of fertilizer use, however, were not profitable for sampled farmers, so the demand for credit to purchase fertilizer will depend on the farmer know- ing the level of fertilizer application that will earn the higher return. 19 Borrower transaction costs ranged from a low of 4 percent to a high of 180 percent of explicit interest charges in the several These results are promising because they point to the pos- case studies summarized. sibility of earning higher returns in agriculture. However, 20 Yunus (2006) as cited by Karlan and Murdoch (2010). 21 Policy makers are tempted to impose interest rate controls to even if returns are high relative to the interest rate for loans, “protect� borrowers. This practice can lead to unexpected con- households may voluntarily withdraw from the credit market. sequences. For example, in May 2010 the Central Bank of Ec- High borrowing transaction costs may reduce borrowing, and uador reduced the maximum lending rate for retail microlenders from 33.9 to 30.5 percent. It was reported that lenders made farmers who could access may decide not to borrow because larger loans to existing clients rather than expand lending to poorer clients, and some lenders may have needed to sell their portfolios to larger institutions with lower fixed costs (Business News Americas, May 6, 2010, downloaded from www.bnamericas.com/new/banking on May 12, 2010). 22 CHAP T E R 7 — C R ED IT D EMA ND , RATES OF RETURN, A ND INTERES T RATE S ENS ITIV ITY they are risk averse and will therefore settle for lower-return, A study by de Mel, McKenzie, and Woodruff (2007) involved lower-risk activities.22 an experiment in which small firms in sri Lanka were given small grants, half in cash and half in the form of equipment or inventories. If returns were low at low levels of capital stock, 7 3 RAtes oF RetURn In MICRoenteRPRIses this result would show that lack of capital was an important Recent microenterprise studies report returns to capital that barrier to opening a business. If, on the other hand, returns are surprisingly high. McKenzie and Woodruff (2006) used were high, this would mean that an entrepreneur could open Mexican data to estimate returns to capital that ranged from a business and grow by reinvesting profits. The average real 10 to 15 percent a month for the smallest firms with capital return to capital was found to be 5.7 percent a month, sub- stocks of less than $500. For each additional $100 invested in stantially higher than the market interest rate. The analysis the enterprise, earnings rose $10–$15 a month. The authors did not address whether enterprises lacked credit because acknowledge the possibility that capital investment is corre- they were screened out or because of lack of demand. lated with ability that was not measured in the study; there- Further analysis of the sri Lankan data produced an unex- fore, they conducted a follow-up randomized experiment to pected result because of the popular focus on women in mi- generate data for a consistent measure of returns to capital crocredit programs. About half of the participating enterprise (McKenzie and Woodruff 2008). Data were collected from owners were women. Contrary to expectations, the grants a panel of male-owned retailing microenterprises. A 1,500 resulted in large, sustained increases in income for male peso ($140) treatment was estimated to increase monthly owners, but no increase in income for female owners (de profits by at least 292–487 pesos, a marginal return to capital Mel, McKenzie, and Woodruff 2009). Apparently men turned of 20–33 percent. If financially constrained households tend the grants into a sustained source of income by making to be poorer, this result suggests that poorer households profitable investments in their enterprises, resulting in an ac- have more ability to pay for capital than better-off house- cumulation of household durable goods and financial assets. holds, and interest rates of even 10 percent a month seem Women, on average, did not generate a sustained source of reasonable.23 income from the grant, either because they did not invest the grant in their enterprise or because they did not earn ad- ditional profits when the grant was invested. The reasons for these gender differences could not be determined. A common problem in studies of returns to capital concerns the measurement of all resources used in the enterprises (Karlan and Morduch 2010). Enterprise profits are often measured without explicitly valuing some inputs used in pro- duction, such as unpaid family labor or the value of manage- rial inputs. Therefore, the estimated profit levels reflect the returns to three inputs—capital, unpaid labor, and manage- Women’s enterprise, Rwanda (Photo: Renate Kloeppinger) ment—thereby overestimating the returns to capital alone. On the other hand, if entrepreneurs have little opportunity to Part of the reason for high returns could be the lack of access employ their resources elsewhere, investing in their enter- to credit. The authors acknowledge a puzzle: given the high prises may be their best option even if they earn a low return returns to capital and the lack of any meaningful minimum for their labor. The key result is that high profits imply that investment threshold, why have the enterprises not grown borrowers could pay high interest rates and still have funds through reinvesting profits? left over for other purposes. 7 4 sensItIVItY oF LoAn DeMAnD to 22 These concepts are discussed more fully in Boucher, Guirkinger, InteRest RAtes and Trivelli (2006). 23 This point was emphasized by Karlan and Murdoch (2010). Of A few studies have attempted to directly measure the sensi- course, the ability of borrowers to pay high interest rates does tivity of loan demand to interest rates. Dehejia, Montgomery, not justify inefficiencies in financial institutions that unnecessar- and Morduch (2009) analyzed the impact of interest rate ily drive up interest rates. S U B S I D I E S A S A N INST RUME NT IN AGRICULT URE FINA NCE: A REV IEW 23 changes when a microfinance NGO operating in Bangladeshi by male- and female-headed firms. As Banerjee and Duflo slums raised interest rates on loans from 2 to 3 percent a (2004) suggest, very high and very low rates of return may month. Borrowers were highly sensitive to the rate increase. coexist in the same economy. Even if entrepreneurs report They tended to take out smaller, more frequent loans and high returns on average, there may be wide variations, a find- repay them more quickly, leading to a reduction in overall ing that complicates policy making. loan balances. The less wealthy households, however, were second, some microenterprises earn high rates of return to particularly sensitive to the rate increase, so the NGO’s port- capital, implying an ability to pay high MFI rates of interest. folio shifted slightly away from its poorest borrowers when There is also evidence to contradict the notion that rates it raised interest rates. The rate increase raised revenues of return in agriculture are so low that farmers cannot pay enough to cover the NGO’s costs even though some cus- interest rates as high as those paid by microentrepreneurs. tomers changed their borrowing behavior. If there is great heterogeneity in rates of return, as implied Karlan and Morduch (2010) report on an experiment by the earlier, farmers may not demand credit until they learn the Compartamos bank in Mexico to randomize interest rates correct combination of inputs to use or until enterprises and at the community level in 80 geographic clusters. Half the value chains are developed that generate higher returns than clusters were randomly assigned to receive a 0.5 percent- are possible with traditional agriculture. age point reduction in the monthly interest rate (roughly a 10 Third, the low returns reported for women is surprising con- percentage point reduction in the annual percentage rate). sidering the importance MFIs place on reaching women, but The microcredit borrowers were found to be quite sensitive this result may reflect occupational choice and traditional to interest rates. The interest reduction led to an increased gender roles in a particular setting (sri Lanka) rather than rep- number of clients and larger loan sizes, and this demand resenting a generalizable phenomenon. Given that studies response was large enough to generate higher revenue. frequently find differences in how women and men spend Costs increased with the higher volume of lending, but not money, it would not be surprising to discover that the profit- as much as revenues (and consequently profits) increased. ability of women’s businesses may be different from that of A recent study of interest rate sensitivity involved an effort male businesses. to estimate individual (rather than average) demand elas- Fourth, Turvey’s finding of great heterogeneity in credit de- ticities (Turvey 2010). Experimental techniques were used to mand among farms requires further investigation. If further estimate household credit demand functions for farm house- research supports that result, it would signal an important holds in China in 2009. The findings suggest considerable complication for policy making—namely, that responses to heterogeneity in credit demand, that the need to use a vari- a general interest rate policy in agriculture will be hard to ety of credit policies should be used to address differences predict and could produce unexpected results. Interest rate across farms. For example, nearly 20 percent of the farm policies may have a greater impact on consumption, but households had nearly perfectly inelastic demand for credit, policy makers normally consider production rather than con- but nearly 20 percent had elasticities above -0.75, including sumption when debating credit policies. some 15 percent with elasticities greater than -1.0. Higher- risk farms had lower demand elasticities, whereas farms Fifth, demand for credit is driven by many factors besides in- with mean revenues had more elastic demand elasticities. terest rates. Transaction costs of borrowing, the complexity Households with higher savings rates had more inelastic of administrative procedures, access to other financial prod- demand than low savers, which suggests that high savings ucts, loan sizes, and repayment schedules may all have an groups substitute savings for credit and low savings groups impact. Microfinance provides useful insights, but more in- view savings and credit as complements. The elasticity of formation is needed about farmer financial behavior, produc- credit demand was higher for consumption goods than for tion and marketing risks, farmer risk aversion and mitigation agricultural production, suggesting that interest rate policies techniques, and other factors to derive robust conclusions have a greater impact on consumption than production. about demand for loans. Implications. These studies of profitability and credit demand have five implications. First, each study represents a unique situation, so it is difficult to generalize about the profitabil- ity of farm and nonfarm enterprises and the returns earned 24 C H A PTER 8 — S TR ENGTH ENING A GR ICULTUR A L CR ED IT: TH E WAY FORWA R D Chapter 8: stRenGtHenInG AGRICULtURAL CReDIt: tHe WAY FoRWARD There is a consensus that agricultural finance in many parts be used to justify government interventions into financial of the world could work better than it presently does for markets. This chapter outlines issues to be considered when small farmers. It seems that financial sector reforms in many evaluating the role of subsidies in these interventions. countries did not percolate down to the small farmers, and access to finance is still a challenge for them. There is sur- prisingly little quantitative evidence, however, to document 8 2 FoRMs oF sUBsIDIes credit flows before and after reforms.24 Nonetheless, there several forms of subsidies are used to improve the supply is a widespread belief that farmers with good investment and demand for financial products and strengthen financial projects are stymied because they lack access to formal systems. Direct subsidies at the level of the financial institu- credit. tion may fund interest rate subsidies and result in “cheap loans� for end borrowers or support institutional develop- Agricultural credit was overlooked in the 1980s and 1990s ment of the financial institution itself. Others focus on the because of the focus on microfinance, but interest is now economic environment, rules and regulations, and support reemerging. Advances are being made in institutions, institutions that indirectly affect the performance of the fi- products, services, processes, and enabling environments nancial system. Other subsidies aid specific financial institu- that contribute to increased outreach and sustainability tions, such as innovation grants to help MFIs design and test (Nagarajan and Meyer 2005). Many financial innovations are products, develop MIss, and provide management and staff being tested as already noted, but the pace of financial sec- training. tor development seems slow for critics. Therefore the ques- tion arises: How can agricultural finance be improved in tune Loans supplied to financial institutions at below-market in- with the requirements of rapidly developing agriculture and terest rates for on-lending to farmers were a popular form the needs of smaller farmers? of direct subsidy under the old agricultural credit paradigm. Indirect subsidies for financial institutions take the form of tax exemptions and the provision of goods and services at 8 1 tHe Use oF sUBsIDIes AnD GRAnts In below-market prices. some subsidies are designed to re- FInAnCIAL seCtoR DeVeLoPMent duce the cost of resources used in the production of goods This chapter delves into contentious issues at the heart of or to reduce the cost of services, such as loans, to improve the old-paradigm debate—namely, the appropriate use of competitiveness and increase demand. subsidies and grants to strengthen rural finance. Donor pro- grams operate with nonrepayable grants or below-market subsidized investments in financial infrastructure, such as rate loans to governments and partner organizations. In quite credit bureaus and collateral registries, are justified as public a few cases, these entry-level subsidies are carried forward goods. Likewise, legal reforms that improve secured transac- as grants and loans as part of foreign assistance programs. tions and enhance property rights encourage financial devel- subsidies are frequently justified as temporary measures to opment by improving borrower creditworthiness, reducing improve economic efficiency or for long-term redistributive lender risks, and increasing demand for investment credit. objectives. As already outlined, economic arguments can subsidies for guarantee funds and insurance products also reduce lender risks. subsidies to aid clients, such as financial literacy programs and training in production technology to 24 A Latin American study concluded that the ratio of agricultural help small farmers effectively participate in value chains, also credit to agricultural gross domestic product (GDP) had fallen in benefit financial institutions. some countries but risen in others (Wenner and Proenza 2000). S U B S I D I E S A S A N INST RUME NT IN AGRICULT URE FINA NCE: A REV IEW 25 subsidies are also used to influence choices concerning pro- alleviating world hunger (Pingali and Kelley 2007; Evenson duction and consumption decisions. For example, training and Gollin 2007). Grants to NGOs and service providers programs and capital subsidies are designed to encourage have been important in expanding the supply of fertilizer and financial institutions to modernize financial systems and up- other agricultural inputs, improving small farmer access to grade MIss. Another subsidy of great interest, especially in agricultural value chains, and developing microfinance insti- Africa, is the use of cash subsidies or vouchers to encourage tutions. Although economic theory recognizes that grants farmers to increase their uptake of inorganic fertilizers. are potentially useful instruments to solve or compensate for market failure, it is important to anticipate the risks of grants. If grants are based on wrong assumptions and analyses, they 8 3 CRItICIsMs oF sUBsIDIes can have harmful effects on the development of private mar- Critics of subsidies focus on potential distortions, such as the kets. since a grant is a subsidy, the question of justification overuse of fertilizer and the undermining of the emerging pri- is critical. vate sector when governments supply free food to the poor The World Bank identified several problems with using or cheap loans and grants to farmers. There are concerns, grants to overcome market failure in agriculture, and similar especially in crises, that when costs are too great to provide problems may exist in other institutions.28 The grants ana- subsidies to all, political manipulation will occur and only an lyzed were designed as one-time subsidies for expenditures elite few will be served. Moreover, subsidies for financial on economic services with mixed public and private goods institutions may create subsidy dependence because it is characteristics. The analysis found that grant users justified difficult to avoid creating the expectation of future subsidies the grants by describing obstacles encountered in develop- in increasing amounts.25 It is often difficult to withdraw well- ment, but it was often unclear if these obstacles were really established agricultural subsidies once they are in place. market failures. In many cases, the argument was simply The recent highly successful initial public offerings of that the target population is poor and lacks assets (Donovan Compartamos in Mexico and sKs in India, and the great 2006). Disagreements exist over whether it is legitimate, profits that private and institutional investors earned from effective, and efficient to provide grants to address market them, prompted questions about the ethics involved in the failures and if grants are the best instrument for this purpose. use of and gains from public subsidies and from high interest Uncertainties about the best remedies to mitigate market rates charged for microcredit (von Pischke 2008). On the one failures create opportunities for inserting political and ideo- hand, high returns have caused excitement among investors logical preferences in public expenditures regarding grants and attract more private capital into the sector. On the other (van der Meer and Noordam 2004). hand, there are concerns, especially in the Compartamos Twelve World Bank rural development projects that provided case, that the high profits were earned by charging poor bor- grants for agricultural research and development, private rowers excessively high interest rates.26 enterprise, and community development were reviewed in detail. The review identified three major weaknesses. The first was lack of discussion about the market failures to be 8 4 AnALYsIs oF tHe Use oF GRAnts to solved and the justification for using grants. second was stIMULAte InVestMents the lack of economic evaluation of the project, and, third, Grants are used as a special form of subsidies by govern- insufficient information about implementation modalities. To ments, international agencies, and foundations to stimulate improve the design and implementation of grants designed investments.27 Grants for national and international agri- to reduce market failure, four categories of issues need to cultural research and extension are widely recognized for be considered: contributing to the Green Revolution technologies critical in ƒ analysis of obstacles to private investment and their possible remedies; 25 Jacob Yaron wrote several World Bank publications using his ƒ analysis of the costs and benefits of resolving the subsidy Dependence Index to analyze the dependency of finan- market failure; cial institutions on subsidies (1994). 26 The Compartmos case was analyzed by Rosenberg (2007). 27 The Terra viva Grants web site manages information about 28 This section draws heavily on the World Bank analyses dis- grants for agriculture, energy, environment, and natural resourc- cussed by van der Meer and Noordam (2004) and Rajalahti and es in developing countries. Farley (2010). 26 C H A PTER 8 — S TR ENGTH ENING A GR ICULTUR A L CR ED IT: TH E WAY FORWA R D ƒ determination of the optimal grant percentage and BoX 1: Examples of Subsidies That Contribute to the amount of grants to be given to particular invest- Sustainable Rural Finance ments, to individual applicants, and to identified target groups; and Pillar 1. Subsidies for Financial Intermediaries ƒ methods for implementation of the grants scheme. subsidies to financial intermediaries must be In addition, a set of questions was recently drafted to ask in ƒ transparent, targeted, and capped; determining whether to use a competitive grant scheme to address market failure.29 ƒ funded explicitly through the government budget or other sources subject to effective control and regular review; 8 5 Use oF GRAnts AnD sUBsIDes FoR ƒ fiscally sustainable; RURAL FInAnCe ƒ fair, not giving an unfair advantage to some The international agencies reviewed their experiences in us- intermediaries vis-à-vis other qualified and directly ing grants and subsidies under the old-paradigm agricultural competing institutions; and credit projects. As a result, the World Bank identified a new ƒ economically justified. strategy involving three pillars of support for rural finance Appropriate subsidies could sector development: (1) government policies and the legal, ƒ provide technical assistance to financial interme- regulatory, and supervisory framework; (2) financial sector diaries to improve systems that enhance efficien- and real sector infrastructure; and (3) financial institutions cy, such as management information systems; (World Bank 2006). Likewise, IFAD recently developed a ƒ develop and introduce demand-responsive prod- new rural finance policy statement based on six guiding prin- ucts on a pilot basis; ciples for intervention at the micro, meso, and macro levels ƒ help develop or improve service delivery mecha- (IFAD 2009a, 14). Both institutions support the financial nisms that enable greater outreach into rural systems approach discussed in the FAO/GTZ Agricultural areas; and Finance Revisited monograph series. The financial systems ƒ cover a portion of the cost of establishing new paradigm shifts the focus from target groups to financial branches in areas that do not have financial inter- systems and institutions (and their supporting infrastructure) mediaries that serve the poor. based on the argument that only well-managed sustainable Pillar 2. Subsidies for Financial Infrastructure financial intermediaries can guarantee the long-term supply of financial services to rural customers. A range of financial Time-bound subsidies may be appropriate to institutions, models, and delivery channels and demand-driv- ƒ create capacity within regulatory and supervisory en and innovative approaches to creating financial products bodies; and services are advocated. The previous poor experience ƒ support the creation of industry associations; and with credit lines suggested caution about their use, and ƒ develop training institutes and credit information emphasis was placed instead on investments in institutional agencies. capacity building (World Bank 2005a). World Bank projects Pillar 3. Subsidies for Economic and Social also need to comply with policies on financial intermediary Infrastructure lending (OP8.30), which stipulate that interest rates to end- subsidies in this category involve investments in eco- customers not be subsidized. nomic and social infrastructure that facilitate the carry- ing out of income-generating activities by members in The World Bank identified several subsidies that could con- the community. such subsidies should tribute to rural development and poverty reduction without distorting the development of sustainable rural finance. The ƒ decline over time, as the local organizations build general rule is that subsidies should be time-bound, limited, up capacity to cover costs through user fees; and and decreasing over time. Examples are given in Box 1. ƒ include a match from the beneficiaries, prefer- ably in cash but also in kind, depending upon the beneficiaries’ economic circumstances. Source: Excerpted from World Bank (2006, 13–16). 29 see Box 4, Page 12 in Rajalahti and Farley (2010). S U B S I D I E S A S A N INST RUME NT IN AGRICULT URE FINA NCE: A REV IEW 27 The revised rural finance strategy by the international agen- cies recognized that potential borrowers of financial sector ƒ Developing a cost-recovery mechanism can help loans must attain a minimum level of economic capacity ensure that only people with serious intentions before they can effectively use and repay loans. Those who receive grants. One possibility would be to are extremely poor, living in post-conflict or emergency situ- establish local savings and credit associations to ations, or seriously ill may not be able to profitably manage capture recoveries and hold beneficiary savings. an economic activity. Therefore, grants may be useful to help The recoveries would help capitalize the entities kick-start an economic activity by providing the very poor with for future lending within the groups. an income-generating asset, if these grants are followed by a ƒ Grants are sometimes made to groups to finance package of assistance to help beneficiaries graduate to sus- expensive assets that cannot be provided by tainable sources of financing. However, since grants are not grants to individuals. However, conflicts can a source of sustainable financing, their use should be limited arise from group ownership of an asset. If group in time. The World Bank also drafted guidelines for grants to ownership does not have clear advantages that the poor to help them accumulate assets and thereby build significantly outweigh these potential conflicts, it their capacity for future access to loans. These guidelines are might be preferable to provide grants to carefully listed in Box 2. targeted individuals. ƒ For poor people with some assets and income- earning capacity, financing a portion of the invest- BoX 2: Subsidies to the Poor for Asset Acquisition ment with a grant and the remainder with savings General guidelines for grants for economic activities in- and a loan from a financial institution should be clude the following: considered. There should be a strict separation between the financial intermediary issuing the ƒ Grants for economic activities should be limited loan and the body issuing the grants, even if the to (1) very poor people who are too vulnerable to funding comes from the same financial institution. take on the risk of a loan, (2) poor people living in This way, it can be made clear to the beneficiary communities that are beyond the reach of finan- that the loan is indeed a loan and needs to be cial institutions willing and able to extend services paid back. If both sources of funding appear to to the poor, and (3) poor people with some assets come from the same organization, confusion and earning capacity but unable to earn enough among beneficiaries is likely to result in poor to pay the investment costs within a reasonable repayment and damage to the local credit culture. time frame. ƒ Grants for income-generating activities should in ƒ Grants must be carefully targeted with strong many cases be combined with training in select- eligibility criteria to avoid capture of benefits by ing, planning, and managing economic activities. elites. The World Bank Institute has an established ƒ Grants should be made on a matching basis, and grassroots management training program, which beneficiary equity contributions should be made includes household management, business in cash if possible. In-kind contributions would skills, and financial skills. such training programs only be appropriate in situations such as emer- improve the ability of targeted groups (especially gencies or post-conflict situations, where the rural women) to manage their income-earning majority of participants cannot be expected to activities and finances, often obviating the need save for a cash contribution. to seek credit and making them more successful ƒ To ensure that beneficiaries value and care for the when they do. such programs are sometimes assets financed by the grant, they should contrib- linked with literacy and health programs. ute as high a percentage as is reasonable, given their overall economic circumstances. This should Source: Excerpted from World Bank (2006, 13–16). be at least 10 percent of total cost, and in many cases, a much greater percentage. 28 C H A PTER 8 — S TR ENGTH ENING A GR ICULTUR A L CR ED IT: TH E WAY FORWA R D some MFIs have experimented with temporary grants for Two important caveats about fertilizer subsidies apply to all the very poor. BRAC, the huge NGO/MFI in Bangladesh, is subsidies. First, there are significant opportunity costs in de- a well-known leader in providing this type of infant industry voting substantial public resources to the supply of fertilizer, support for some of its poorest members. CGAP and the Ford a private good (as is credit), at the expense of public goods, Foundation are testing nine graduation models that target such as infrastructure, education, or public health services, the ultra-poor (El-Zoghbi, Montesquiou, and Hashemi 2009). that may have a greater impact in reducing poverty (Minot However, such subsidies raise the possibility that pressures and Benson 2009). second, could develop to convert the temporary client subsidy into …although there is an increasing perception among po- a permanent subsidy. Grants provided to the poor by MFIs litical leaders that there is a huge and unacceptable hu- imply that richer borrowers pay higher interest rates to cover man cost in waiting for markets to develop well enough the cost of the grants or that the MFI obtains a continuous to support agricultural intensification in Africa, it may be supply of external subsidies to finance this component of its equally important to ask what is the human cost of not operations (Armendariz and Morduch 2005). taking active steps now to make markets work in the subsidies for savings rather than credit could be even more future. There is a very real possibility that quick fix ap- important for the poor. IFAD’s strategy notes that savings are proaches to promote fertilizer use may leave inadequate important because they enable poor households to withstand resources and little political will for effectively improving income shocks and mitigate the effect of emergencies and the situation for the long run (Crawford, Jayne, and Kelly crises. Access to secure savings services is also expected 2006, 46). to promote financial discipline and help borrowers service The use of subsidies to meet short-term objectives, there- their loans on a timely basis. However, customer education fore, potentially implies high opportunity costs in the form of and protection are critical, savings should be adequately pro- insufficient resources for and lack of attention to long-term tected, and any risks should be clearly explained to savers development needs. (IFAD 2009a). 8 7 sMARt sUBsIDIes FoR MICRoCReDIt AnD 8 6 tHe ConCePt oF sMARt sUBsIDIes AGRICULtURAL CReDIt The debate over the proper use of subsidies has focused The pros and cons of subsidies for microfinance have been attention on the idea of designing so-called “smart� subsi- analyzed based on their implications for agricultural credit dies. smart food subsidies aim to increase food availability in the short term while stimulating growth and rural develop- ment and increasing (or at least not suppressing) effective demand for and commercial distribution of inputs in the long run (Dorward et al. 2008). The concept of smart subsidies seems to be most advanced by supporters of fertilizer sub- sidies in Africa when they propose that governments avoid past mistakes and implement instead “smart subsidies� designed to target the poor and support rather than undercut private input distribution markets (Minot and Benson 2009). Although these arguments paint an enticing picture for smart subsidies, they provide little guidance on what form they should take in practice, how the traditional problems of elite capture and resale can be avoided, how subsidies can best Savings Credit Cooperative associated with XacBank in the Hentii be administered (for example, through private or state-con- Aimag Province of Mongolia (Photo: Michael Hamp) trolled systems), and how leakages and distortions can be minimized. Little evidence is provided on the relative costs (Armendariz and Morduch 2005). smart subsidies were of subsidies versus other forms of income or food transfer defined as carefully designed interventions to minimize dis- (Crawford, Jayne, and Kelly 2006). tortions, mistargeting, and inefficiencies while maximizing social benefits. The authors accept the financial system’s S U B S I D I E S A S A N INST RUME NT IN AGRICULT URE FINA NCE: A REV IEW 29 argument of “subsidize the institution but not the borrower� likelihood that the benefits will be spread among all but recognize that this is impossible in practice because any members. subsidy to an institution means it has fewer costs to pass ƒ Indirect subsidies that benefit many borrowers may on to borrowers. A variation of this idea is to subsidize start- generate more total benefits than direct interest-rate ups but not operating costs, but this is hard to accomplish subsidies to borrowers. in practice unless the provider of the subsidy (government ƒ Quantitative performance measures should be in- or donor) has a firm and credible exit strategy.30 It is easier cluded in project agreements so subsidies to financial to justify indirect subsidies for the creation of public goods institutions do not dull incentives for achieving high useful to several financial institutions or direct subsidies used performance levels. For this reason, subsidies need for specific time-bound institution-building tasks. to be time-bound with explicit exit strategies specified for the supplier of the subsidies. One example of a smart subsidy is the World Bank’s new Agriculture Finance support Facility, launched in June 2009 ƒ Comparative cost-benefit studies are needed to to subsidize the expansion of agricultural finance. The facility, identify which subsidies generate the greatest payoff with the support of the Bill and Melinda Gates Foundation, in practice. will provide capacity-building grants to retail financial institu- ƒ Recipients of grants should provide matching cash tions in Africa and Asia for up to 50 percent of development or in-kind contributions to demonstrate their commit- costs for initiatives that will enable them to enter into or sig- ment to the projects funded. nificantly scale up agricultural lending. Eligible expenses in- ƒ The provision of grants to financial institutions should clude technical advice, rural outreach infrastructure, staff and be designed so recipients clearly understand the client training, consultancy services, and staff salaries on a difference between grants and loans that need to be declining basis. Grantees will exchange experiences through repaid. peer-learning and networking opportunities.31 Conclusions. The following general guidelines for “smart or market-friendly subsidies� emerge from this summary of subsidy issues: ƒ subsidizing the institution but not the borrower is the best way to reduce distortions even if this implies a degree of direct subsidy to borrowers. ƒ Projects to subsidize selected institutions should explicitly consider the interest rates to be charged rela- tive to competing institutions so the subsidies do not undermine competition. ƒ subsidies that successfully create public goods for the benefit of the entire financial sector may generate higher returns than subsidies for specific institutions because no single institution can justify making the investment alone when the benefits accrue to many. ƒ subsidies for institution-building of individual financial institutions are easier to justify if there is a natural pos- itive spillover to nonsubsidized institutions. subsidies to finance innovations created through networks of financial institutions may be preferred because of the 30 The authors use the example of the difficulty of removing favor- able tariff protections for infant industries. 31 Information on this facility can be found at www.agrifinfacility. org. 30 C H A P T E R 9 — F IVE M Aj OR A P P ROACHE S TO SUPPORT A GRICULTUR A L FINA NC E A ND DES IGN OF SUBS IDY INFLOW S Chapter 9: FIVe MAjoR APPRoACHes to sUPPoRt AGRICULtURAL FInAnCe AnD DesIGn oF sUBsIDY InFLoWs This chapter discusses five major interventions frequently 9 1 1 Microinsurance supported by governments and international agencies to Microinsurance aims to protect low-income people against support financial services. These interventions are (1) micro- specific risks in exchange for regular premium payments insurance and weather index-based insurance, (2) guarantee proportionate to the likelihood and cost of the risk involved. funds, (3) warehouse receipts, (4) agricultural development Much of the emphasis of microinsurance is on reaching poor banks, and (5) agricultural investment funds. These inter- people who work in the informal economy and therefore ventions may be useful in their own right, but they are also have less access to commercial insurance provided through important because they may induce financial institutions to employment (Churchill 2006a). Yet insurance for poor rural serve more rural clients and farmers. All involve some kind people is not limited to microinsurance alone; where popu- of subsidy intended to kick-start a private-sector activity. lation densities and service costs permit, health insurance, This section summarizes the rationale for each and offers conventional crop insurance, and other types of insurance an analysis of the extent to which they seem to achieve the are increasingly provided by national insurance agencies, objective of contributing to expanded agricultural credit.32 especially on the Indian subcontinent. Four major challenges face commercial insurers attempting 9 1 MICRoInsURAnCe AnD WeAtHeR InDeX- to serve the poor. First, their products are not ideally designed BAseD InsURAnCe to meet the small irregular cash flows typical of the informal economy. second, the commercial insurance industry lacks Access to insurance for agriculture and poor people has a distribution network necessary to reduce transaction costs. received comparatively little attention relative to access to Third, insurance companies lack experience and data to cal- credit in most developing countries, but that has changed culate the risks faced by the poor and adequate methods to dramatically in recent years because of two major lines of control adverse selection and moral hazard. Fourth, the poor experimentation. The first is microinsurance for poor people are often skeptical about insurance and refuse to pay premi- offered as a standalone product or in conjunction with other ums for insurance products offering future benefits that they MFI products. The second is indexed crop and livestock may never receive. insurance targeted to agricultural producers. The rationale for providing insurance to the poor is that they live in risky Microinsurance can be considered either a private market environments, and without insurance they are less likely to product or an activity that plays a redistributive social func- take advantage of income-generating opportunities to reduce tion worthy of public subsidies. Therefore, advocates argue poverty. In agriculture, poor uninsured farmers may choose that expanding sustainable insurance should involve dual ob- to produce low-value, low-risk crops rather than adopt higher- jectives: work with the private sector to supply appropriate income alternatives. Lenders may refuse to grant credit be- products and systems and work with the public sector to cause they perceive that uninsured farmers are too risky. As increase demand for insurance and subsidize the costs for with credit markets, informal and village-based risk-sharing the poor (Jacquier et al. 2006). Therefore, the insurance fron- and insurance mechanisms are considered inferior to formal tier lies in finding ways to successfully wed social protection insurance. against extreme risks with private market protection against smaller risks (Hill and Torero 2009). 32 see also GTZ’s Rural 21, Issue No. 4, 2010. S U B S I D I E S A S A N INST RUME NT IN AGRICULT URE FINA NCE: A REV IEW 31 To protect their loan portfolios, MFIs prefer clients with insur- ƒ improved institutional capacity, investments in new ance, and because they already have financial transactions delivery channels, and creation of new insurers; with the poor, it may be most cost-effective for them to link ƒ improved operating efficiency to reduce premiums; insurance with other products. some MFIs offer insurance, ƒ enhanced business models including the increased frequently developed by or in conjunction with insurance use of reinsurance; companies, along with other financial products and make ƒ regulations providing a balance between prudence and insurance purchase obligatory for borrowers to reduce costs entry barriers; and and ensure viability.33 ƒ strong macroeconomies with efficient financial mar- Long-term sustainability of microinsurance has been difficult kets and infrastructure to facilitate long-term invest- to achieve. subsidies have been a mixed blessing and are ment strategies (Botero et al. 2006). not recommended. For example, government subsidies 9 1 2 Indexed crop and livestock insurance have kept the cost of health care insurance artificially low in several countries. Although subsidies speed uptake in the Creating sustainable insurance for crops and livestock is es- early stages of programs, they make insurance vulnerable to pecially challenging. Besides the normal adverse selection political influence and policy changes. When subsidies are and moral hazard problems, returns to farming are typically discontinued, premiums must be raised, leading to a contrac- covariate, and frequent health risks impose logistical chal- tion in renewal rates (Churchill and Garand 2006). lenges. The problems of providing sustainable crop insur- ance are well known (Hazell, Pomerada, and valdes 1986; The future for expanded microinsurance will depend on Roberts 2005). It is costly to write insurance contracts for ƒ development of an insurance culture to increase de- large numbers of small farmers and implement farm-level mand for insurance; a logical place to start is through inspections. Because farmers are unwilling to pay the full credit life insurance offered by MFIs; cost of all-risk crop insurance, traditional programs are public ƒ product designs responsive to evolving customer schemes subject to political pressures and are often used as demands in which health insurance may be the logical income transfer mechanisms for farmers. A distinction ex- next step after credit insurance; the limits of market- ists between, on the one hand, insurance that “protects� the based solutions may emerge in this process so public livelihoods and assets of the poor from catastrophic losses social protection services may become a logical that must be subsidized as part of the social safety net and, demand; on the other hand, insurance that is linked to agricultural development through private intermediaries. The latter may be sold on an unsubsidized basis if the insurance enables 33 several public and private sector microinsurance models are dis- cussed in Churchill (2006) and form the basis for conclusions presented in that volume and summarized here. Ethiopia (Photo: Renate Kloeppinger) Poultry farming by a microfinance client/loan beneficiary of BRAC in Rangpur, northwest Bangladesh (Photo: Michael Hamp) 32 C H A P T E R 9 — F IVE M Aj OR A P P ROACHE S TO SUPPORT A GRICULTUR A L FINA NC E A ND DES IGN OF SUBS IDY INFLOW S farmers to participate in new technologies or high-value mar- (Hazell et al. 2010) suggest the following key drivers of sus- kets that significantly raise expected incomes. tainability and scalability of weather index insurance: The creation of index-based insurance has raised hopes of ƒ Create a proposition of real value to the insured, and breaking the major barriers to protecting individual farmers offer insurance as part of a wider package of service. against production risks. With an index, insurance is paid out ƒ Build the capacity and ownership of implementation when an independently observed trigger (such as level of stakeholders. rainfall or temperature) shows that an insurable event has ƒ Increase client awareness of index insurance occurred. Index insurance has the potential of reducing sev- products. eral problems. First, it reduces moral hazard because the in- ƒ Graft onto existing, efficient delivery channels, engag- sured cannot significantly influence the index value and the ing the private sector from the outset. indemnity paid by the insurance contract. second, adverse ƒ Get access to international risk-transfer markets. selection problems are reduced because the contract’s in- ƒ Improve the infrastructure and quality of weather data. demnity schedule and premium rate are based on publicly available rather than privately held information. Third, it does ƒ Promote enabling legal and regulatory frameworks. not require individually tailored terms of indemnification or ƒ Monitor and evaluate products to promote continuous verification of individual loss claims so administrative costs improvement. are reduced, making it more affordable, particularly for poor A sequential strategy for developing insurance markets farmers or cooperatives, farmer associations, or lenders that beginning with linkages to lending was proposed by skees purchase insurance for them. The disadvantage is that basis et al. (2007, 10). This strategy proposes using index-based risk (the risk that payouts may not match the losses a farmer weather insurance to first address the biggest risks of major experiences) may be substantial, making it difficult for farm- catastrophes. This step will then facilitate the development ers to understand and accept (skees 2008). of other products for different categories of farmers and rural households. The strategy advocates long-term sustainability several pilot index schemes have been launched, but it is and limits the role of government to that of a facilitator and too early to judge their success. A small number failed to not a direct deliverer of insurance. Governments may choose generate demand and were discontinued. In India, however, to fund insurance for catastrophic losses as part of the social 1.25 million farmers were reached in 2009. Experience to safety net but generally not provide insurance36 nor provide date suggests the following lessons: (1) insurance needs direct premium subsidies that will undermine incentives to be accompanied with improved access to technology for private sector insurance. In practice, subsidies tend to and credit so that farmer incomes are likely to rise with favor wealthier farm households, thereby eroding poverty adoption, (2) basis risk needs to be reduced, (3) improving objectives, while targeted premium subsidies rarely work understanding and trust are key to increasing demand for as planned. Governments should establish an appropriate the insurance, (4) scaling up will require public goods invest- enabling environment, provide certain public goods, sup- ments in weather data infrastructure and in creating new port improvements in the legal and regulatory environment, products, and (5) uptake will be affected by whether lenders improve data systems and collection, conduct educational offer loans to clients to pay for insurance premiums and the efforts about weather insurance, assist with product devel- extent to which insurance premiums are subsidized.34 The opment, and facilitate access to global insurance and reinsur- use of smart subsidies to kick-start insurance markets must ance markets.37 be carefully justified, and robust impact studies are needed to learn from pilot schemes and demonstrate their economic and social benefits.35 IFAD and the World Food Programme insurance contracts; and (3) the cost of marketing the products, especially to the smallholder sector. They use the public good argument to advocate for a public role in underwriting innovation 34 Miranda and Gonzalez-vega (2010) demonstrate that subsidies costs, creating reliable long-term information, and sharing some for index insurance premiums may affect borrower incentives to of the excess risks until more long-term information is available. repay loans and, therefore, the returns earned by lenders. Lend- 36 Clarke and Dercon (2009) are more open to the use of subsidies ers may be in a better position to buy insurance than borrowers. for insurance for low-income people. 35 Carter, Galarza, and Boucher (2007) conducted an analysis of the 37 In a recent paper, Teh and Martina (2008) focus on the catastro- potential for weather-based index insurance in Peru and identi- phe issue and demonstrate their belief that conglomerates of fied three major reasons for private market’s failure to provide it: intermediate financial institutions may need to be formed in de- (1) the novelty of the product and the costs associated with its veloping countries to acquire risk-transfer financial instruments. innovation; (2) the scarcity of long-term data on which to base They argue that the preferred instrument in responding to natural S U B S I D I E S A S A N INST RUME NT IN AGRICULT URE FINA NCE: A REV IEW 33 The current insurance experiments need to be evaluated of first addressing the biggest risks of major catastrophes with attention to improved understanding of basic questions requires careful analysis. such as: In addition, there may be short-term opportunities for donors 1. developing a better understanding of traditional risk and governments to accelerate progress and speed innova- mitigation strategies of households, such as savings, tions for both microinsurance and index-based crop and live- accumulation of saleable assets, and on-farm man- stock insurance by supporting financial institutions commit- agement practices; ted to developing sustainable insurance. Bundling insurance 2. learning more about the preferences of farm house- with loans and savings is emerging as a logical first step to holds, their willingness to pay for insurance, and the reduce costs and speed adoption. Careful feasibility studies appropriate role of subsidies to kick-start insurance are needed to evaluate where conditions are most appropri- markets;38 ate for new pilot index-based insurance projects.40 3. alternative ways to deliver insurance products since The logical role of governments and donors is to focus on not all potential buyers will be borrowers of financial long-term investments in public goods, such as weather- institutions; and reporting stations and basic data collection and analysis, 4. the role of the government in creating disincentives needed to create the necessary conditions for thriving insur- for insurance in crisis situations through programs of ance markets. Innovations, technologies, and ongoing ex- emergency relief, loan forgiveness, and subsidized perimentation will make it more likely that better insurance emergency loans. protection will emerge for poor households. But complemen- tary investments are also needed in the basic methods of Conclusions. A considerable amount of analysis is required reducing risks through low-cost irrigation, drought-resistant concerning the design of insurance products and appropriate seed varieties, improved sanitation, and better preventive delivery systems for the poor and for agriculture. Donors can health care (Hill and Torero 2009). play a useful role in conducting or financing careful evalu- ations of insurance experiments to improve understanding about basic questions. Additional experiments in more di- 9 2 CReDIt GUARAntee FUnDs41 verse environments and with different product designs and The objective of credit guarantee funds is to reduce default delivery systems will be important to eventually develop risks for lenders as an inducement to lend to specific target best practices. Robust evaluations are needed to assess groups or types of institutions. Although the nature of the whether insurance investments produce the desired effects specific market imperfection is often not well analyzed, guar- and to determine the appropriate role for public subsidies in antee subsidies are often justified to accelerate learning so developing private insurance markets and for catastrophe lenders become more effective in credit analysis. The enthu- insurance.39 The proposed insurance development strategy siasm for guarantees is somewhat surprising. Even though they are widely used in commercial credit transactions, several studies in the 1990s were cautious about advocating guarantees to stimulate lending or expecting significant im- disaster floods and droughts is subsidized catastrophe bonds, not reinsurance or lump-sum foreign disaster assistance. Foreign pacts from credit guarantee projects. There was no interna- assistance should focus on the idea of subsidizing the issuing of tional consensus that such schemes widen access to formal catastrophe bonds. stutley (2010) also discusses the role of gov- ernment in the development of index insurance for major natural bank credits for small and medium enterprises (sMEs).42 disasters in Africa. 38 A discussion of the pros and cons of subsidies for weather-index insurance and on the design of appropriate insurance products can be found in Hazell et al. (2010). 40 see Hartell and skees (2009) for an example of a prefeasibility 39 some weather risks may be impractical to insure conventionally, study that revealed the complications of introducing weather- and derivatives and other alternative risk transfer mechanisms index insurance in Mali. can provide a solution. Insurance-linked securities or cat bonds expanded until 2007, when cat bond sales reached Us$7 billion 41 I am indebted to Calvin Miller for useful comments on this sec- of insured covered globally. The market then contracted because tion. of the global financial crisis but has recovered again somewhat. 42 several aspects of this discussion are included in Meyer and A pilot scheme has been launched in the Caribbean, and cat Nagarajan (1996), Llisterri and Levitsky (1996), Gudger (1997), bonds are expected to replace a significant percentage of con- Levitsky (1997), Doran and Levitsky (1997), vogel and Adams ventional disaster relief in developing countries. (1997), and Gudger (1998). 34 C H A P T E R 9 — F IVE M Aj OR A P P ROACHE S TO SUPPORT A GRICULTUR A L FINA NC E A ND DES IGN OF SUBS IDY INFLOW S A recent example of the enthusiasm for and impacts ex- expected that they will see lending to the target group as pected from loan guarantees can be found in a report by less risky than envisioned and will make nonguaranteed the Alliance for a Green Revolution in Africa (AGRA).43 It loans when the guarantees end. Guarantees are perceived reported using $17 million in loan guarantee funds to le- as cost-effective for donors; by offering partial guarantees, verage $160 million through four major lending programs. donors can benefit more borrowers than they could by using This included a $10 million line of credit that the National the same funds directly for credit lines. Microfinance Bank in Tanzania agreed to lend to agro-dealers at an interest rate of 18 percent, compared to the typical There are at least two problems with this standard rationale. rate of 46 percent charged by MFIs. In Kenya in 2008, AGRA First, if the reluctance to lend is attributed to problems such and IFAD provided $2.5 million each as a loan guarantee that as legal and administrative frameworks, regulatory require- leveraged $50 million from Equity Banks. As of May 2009, ments, high rates of return paid on government securities, the program had loaned more than 679 million Kenyan shil- or other problems, then the most appropriate strategy is a lings (about $9.8 million) to almost 20,000 small-scale farm- market-oriented one that addresses these problems directly ers. The bank reportedly hired 100 new staff to expand and so that all borrowers, not just a specific target group, will improve the program’s outreach and effectiveness. In March benefit. 2009, standard Bank in Africa agreed to offer $100 million second, the underlying concept of an international guarantee in loans to smallholder farmers and agricultural businesses: appears naïve. Guarantors, who know little about the local $25 million each went to Ghana, Mozambique, Tanzania, and environment or conditions faced by borrowers, imply that Uganda. With several contributing partners, AGRA devel- they are better able to evaluate credit risks than are local oped a loan guarantee fund of $10 million for these loans lenders. As a result, they are willing to offer guarantees to (AGRA 2009). No details were provided on the design of absorb part of the credit risk. the guarantees or the circumstances of the lenders involved to show whether the guarantees actually induced these What the guarantors probably mean is that they have learned amounts of additional lending. to correctly evaluate the risk profile of the target group else- where and because of this experience will be effective in this new environment, as long as (1) their proven lending technol- 9 2 1 Rationale for guarantee schemes ogy is used, and (2) the local environmental and other con- Arguments in favor of guarantees follow a similar pattern straints are no worse. Therefore, some minimum conditions (for example, Doran, McFayden, and vogel 2009). Banks must be met (such as appropriate regulations or comple- have liquid funds but are risk averse so they invest in se- mentary inputs such as training and technology transfer) for cure government securities rather than make retail loans to the guarantee to succeed. These complementary activities small farmers and sMEs or wholesale loans to cooperatives, may actually be more important than the guarantee itself. In NGOs, and other MFIs that serve these neglected markets. extreme cases, the guarantee may not even be necessary or Therefore, it is believed that potential borrowers have profit- will make little additional impact if the complementary activi- able projects but are starved for credit. These problems are ties are undertaken. exacerbated when legal and administrative frameworks do not support collateral contracts effectively, resulting in larger collateral requirements than necessary for prudent lending.44 9 2 2 Review of guarantee experiences in the 1990s Collateral substitutes, in the form of third-party sureties or The reviews of guarantees in the 1990s focused on two main partial guarantees from external funds, are sometimes avail- issues. First, there was little clear evidence of additionality— able, but guarantee funds are considered necessary to break that is, evidence that loans made with a guarantee would not the credit bottlenecks. Once lenders gain experience, it is have been made without it. It is difficult to evaluate addition- ality without a clear counterfactual (what lenders and borrow- ers would have done without the guarantee). The fact that 43 AGRA was initiated in 2006 with initial funding of $150 million a lender guarantees a loan does not “prove� that it would by the Rockefeller Foundation and the Bill and Melinda Gates not have been made otherwise. Guarantees may have sev- Foundation to help millions of small-scale farmers and catalyze an African Green Revolution. Other organizations subsequently eral other possible impacts on lenders. They might induce provided support. larger loans, longer-term loans, reduced collateral require- 44 An analysis of loan collateral problems and the potential impact ments, reduced interest rates, speedier loan processing, and of resolving them is presented in Fleisig, safavian, and De La increased marketability of loans in secondary markets. Just Pena (2006). S U B S I D I E S A S A N INST RUME NT IN AGRICULT URE FINA NCE: A REV IEW 35 as it is difficult to evaluate the impacts of microcredit, it is difficult to evaluate the impacts of guarantees on borrowers. Potential economy-wide benefits of loan guarantees, such as higher employment and economic growth, also need to be tested. second, many guarantee schemes were not sustainable with- out subsidies, and data were sparse to assess whether or not the magnitude of subsidization decreased over time. Many guarantees suspended or delayed payments on claims to preserve the sustainability of their funds, which undermined confidence among lenders. Poorly designed subsidized guar- antees may have crowded out the development of private nonsubsidized guarantees and hindered rather than aided the Selling farm tools in Madagascar (Photo: Maria Pagura) development of a market-oriented financial system. begin to develop a long-term relationship with a local bank, The design of guarantees affected additionality, sustainability to diversify funding, or to gain prestige by associating with an and other possible impacts. The critical design features were international institution. eligibility, leverage, risk sharing, fees, and claims procedures, The results of the review were mixed regarding MFI gradu- but no clear best practices were defined. Many countries ation to unguaranteed borrowing, for three reasons. First, believed that guarantees were the most economic way to lenders may continue to use guarantees indefinitely if they provide financial support to sMEs, but there was no consen- are more efficient than other methods of risk management. sus that they widened access to formal bank credit (Levitsky second, when guarantees are used to overcome regulatory 1997). A more recent report also concluded that government barriers, they will continue to be necessary for future trans- or donor-financed loan guarantee schemes generally have not actions unless regulations are changed. Third, MFIs often led to significant additionality but that mutual guarantee as- find other funding sources with better terms (such as lower sociations seem to be more useful (World Bank 2003). costs, longer terms, lower collateral requirements) than the retail terms offered by local banks. For regulatory, political, 9 2 3 CGAP study of guarantees for MFIs and other reasons, banks in some markets chose to lend to A CGAP study reviewed guarantees for supporting microfi- MFIs on more favorable wholesale terms. Larger MFIs could nance (Flaming 2007). The review analyzed loan guarantees raise funds through savings, certificates of deposits, and issued to back up loans mostly from local banks to rapidly bonds. growing, small, profitable MFIs. The MFIs had a strong de- The review found that guarantee agencies subsidize their mand for funding because they did not have access to sav- guarantees, but few supply financial reports, so subsidy ings. The guarantees were expected to enable them to obtain levels cannot be estimated. Operating expenses are often bank loans that were otherwise unavailable, and the expe- supported by general agency budgets and are not allocated rience would increase the participating bank’s appetite for to the guarantee. Typically, agencies determine fees by esti- nonguaranteed lending in the future. Loan guarantees could mating what banks and MFIs would be willing to pay rather also increase the MFI’s collateral so that local banks would than by evaluating the agencies’ costs or the true market comply with banking regulations concerning unsecured lend- value of the risk involved. ing. In addition, the guarantees had potential spillover effects in terms of inducing non-MFI lenders to experiment with MFI The overall conclusion was that the cost of guaranteeing technologies to lend to farmers and rural residents. MFIs loans would be unsustainable without considerable subsidization. The benefits of guaranteed loans were typi- Both the guarantors and MFI managers reported that the cally modest, and the costs of the loans were high despite guarantees helped the MFIs obtain loans from banks, but substantial subsidies by the guarantors. As competitive they had little impact on the terms of loans offered. The guar- MFIs grew, they found better sources of funding than retail antor’s annual fees, when added to the bank’s interest rate, loans from local banks. For most MFIs, bank loans are not made these funds more costly than the MFIs’ other sources. a long-term sustainable source, especially when they face Therefore, the primary MFI motivation for participating was to 36 C H A P T E R 9 — F IVE M Aj OR A P P ROACHE S TO SUPPORT A GRICULTUR A L FINA NC E A ND DES IGN OF SUBS IDY INFLOW S competition and can no longer pass on high funding costs included in these estimates. Important details are sketchy to borrowers. Guarantors realize their greatest potential by concerning the selection of institutions and borrowers for focusing on lenders that use guarantees to structure loans to guarantees, guarantee designs and fees, and how the guar- MFIs in conditions competitive with other funding options. antees support local missions. No effort was made to compare guarantee benefits with Only one evaluation report specifically covered agriculture: costs. the José Maria Covelo Foundation (FJMC) in Honduras, which began to offer direct credit as an MFI in 1995. In 9 2 4 USAID guarantees January 2008 it established the Banco Popular Covelo The U.s. Agency for International Development (UsAID) (Bancovelo), a licensed commercial bank. The guarantee implements guarantees through its Development Credit helped FJMC jumpstart its agricultural lending. The number Authority (DCA) to support the programs of its field mis- of loans supplied increased along with average loan size and sions, and it has posted several evaluations.45 The UsAID length, while the bank kept interest rates low. However, it microfinance program also conducted evaluations of DCA is unlikely that these positive outcomes will be sustained guarantees that help MFIs obtain loans. The primary objec- because Bancovelo put ceilings on its agricultural lending tive of the DCA guarantees is to influence lender behavior and may sell its agricultural portfolio. The FJMC experience in favor of market segments (such as agriculture and sMEs) did not stimulate other MFIs to get involved in agricultural that are underserved by lenders. The UsAID missions pay lending. Past governmental debt forgiveness schemes and the U.s. Treasury an amount estimated as potential losses, subsidized interest programs deter lending to agriculture. but sustainability of a guarantee fund is not an objective, and no costs are imputed for management of the guarantee and The evaluation of DCA guarantees in Uganda was more rigor- the oversight provided by the UsAID missions. ous than the other five evaluations. The UsAID Rural savings Promotion and Enhancement of Enterprise Development Four types of partial guarantees are offered: single project (Rural sPEED) project managed three DCA loan guarantees loan guarantees in which specific lenders and borrowers are with six banks and one microfinance depository institution identified up front; loan portfolio guarantees from lenders (UsAID/Rural sPEED 2007). One guarantee supported lend- to a pool of borrowers; bond guarantees; and portable guar- ing to sMEs and MFIs by seven commercial banks, and the antees in which targeted borrowers shop for the best loan second targeted small loans to micro, small, and medium package. It is expected that the effects of guarantees will be enterprises (MsMEs), especially in rural areas. The third was sustainable, but sustainability depends on factors beyond the a collateral management program to encourage lending to guarantee. Positive impacts can occur at the level of borrow- the grain industry, but it was canceled for reasons unrelated ers and participating financial institutions through potentially to the guarantee. lower interest rates and collateral requirements and through greater credit availability. The DCA covers up to 50 percent of About 40 percent of the guaranteed loans went to agricul- defaults on loans made by private financial institutions. since ture in the first guarantee, and 78 percent in the second. The it was established in late 1999, the program reported more guarantee coverage of 50 percent on net principal losses than 225 partial credit loan and bond guarantees in more than was thought to have increased lending. Repeat loans to MFIs 60 countries. Claims have been approximately 1 percent, and were larger than the original loans, suggesting that the guar- the total cost to UsAID was reported at approximately $61 antees helped cover the additional risk. some banks reduced million, but no information was provided to explain the costs collateral requirements, others began to accept different types of collateral, and a few offered unsecured lending to proven clients. The guarantees helped two large MFIs obtain 45 Unless otherwise noted, information about the DCA and the credit, and two could now get access to commercial credit five evaluations was obtained from the material posted at http:// without a guarantee. One bank self-selected to “graduate� www.usaid.gov/our_work/economic_growth_and_trade/development_ credit/. The evaluations include (1) EcoBank, a prominent Nigeri- and lend to rural MsMEs without guarantees. Claims paid an-owned retail bank in Ghana; (2) Bank Danamon, a large and by UsAID under the first guarantee were only 1.6 percent profitable private commercial bank in Indonesia; (3) José Maria of portfolio lent, and no claims had yet been made under Covelo Foundation in Honduras; (4) Local Government Unit Guar- antee Corporation in the Philippines; and (5) portfolio guarantee the second guarantee. These low claim rates may show that with Bank Center-Invest, a regional bank in the southern Federal lenders were being too selective and not effectively testing District, Russia. For unknown reasons, UsAID requested that the evaluations discuss only findings and conclusions, not les- the market in spite of the guarantees. sons learned or recommendations. S U B S I D I E S A S A N INST RUME NT IN AGRICULT URE FINA NCE: A REV IEW 37 Technical support was credited with facilitating the use of The most interesting possibility is that the training and tech- guarantees and building understanding of sME lending. nical assistance components of guarantee schemes are This support ranged from intensive two-week credit officer more important than the guarantees themselves in stimulat- training to follow-on loan mentoring to specialized training in ing lenders to work with a new client group. This suggests agricultural lending. Banks were positive about the comple- that “guarantee plus� programs may be critical in affecting mentary support activities in the loan mentoring program, performance. The guarantees may be the frosting on the and all banks liked the ability to interact with the DCA portfo- cake, not the cake itself. lio manager at any time for clarifications essential to speedy operations. 9 3 WAReHoUse ReCeIPts Conclusions. The case for expecting major impacts from Warehouse receipts are an old form of collateralized com- guarantee schemes continues to be unclear. The methodol- modity transaction now being considered in several countries ogy used to evaluate the impacts of guarantees has been as a catalyst to stimulate agricultural lending where other weak, so questions of additionality and sustainability are as attempts have failed.46 The basic rationale is that instead valid today as in the 1990s. It is possible that guarantees may of taking a credit risk by lending against the borrower’s ex- provide an additional bit of comfort for financial institutions pected future cash flow and repayment capacity, the lender that are interested in testing the feasibility of lending to a takes a minimal performance risk because the collateralized new client group. It is unlikely, however, that a guarantee commodity can be easily liquidated in the event of nonper- alone will induce much additional lending by lenders who do formance (World Bank 2005b). The commodity becomes the not have such an interest. first source of repayment rather than the second source, as International agencies could perform a valuable service by in typical loans. conducting a few robust evaluations to determine if and un- Warehouse receipts play a limited role in agricultural credit by der what conditions guarantees really produce the expected facilitating postharvest financing. Except in the case of dou- results. The evaluations should assess how the details of ble or triple cropping, credit obtained after harvest does not guarantee design affect performance. It is also critical to directly solve the problem of supplying the working capital evaluate whether they distort markets and prevent rather required to plant a new crop. After harvest, the commodities than encourage private credit market development. The sequencing of market development may be important in af- fecting guarantee performance. For example, if a new credit 46 several publications discuss the concept of warehouse receipts registry system is created, a guarantee may help nudge lend- and provide examples from several countries (such as Fries and ers to begin to use it. Akin 2004; Coulter 2009; and World Bank 2005b). Warehouse storage and receipt, India (Photo: Mark Wenner) Uganda (Photo: Renate Kloeppinger) 38 C H A P T E R 9 — F IVE M Aj OR A P P ROACHE S TO SUPPORT A GRICULTUR A L FINA NC E A ND DES IGN OF SUBS IDY INFLOW S standards for warehouse operations and effective licensing and monitoring of warehouses; (6) an effective indemnity fund or bond as insurance against potential fraud or negli- gence by warehouse operators; and (7) local financial institu- tions willing to experiment with a new product. Oftentimes these prerequisites can be most easily met for traditional ex- port commodities such as coffee and cotton, but their use for cereals and other nontraditional export crops often requires preliminary work to create essential support systems. The liquidity that warehouse receipts financing provides Savings Credit Cooperative, Ethiopia (Photo: Renate Kloeppinger) may be useful for traders, millers, and processors and may indirectly benefit farmers through increased competition for are stored in a licensed and bonded warehouse that issues a their harvest. But it makes only a limited direct contribution receipt certifying the amount and quality stored. The owner to the objective of increasing production credit for small of the commodity (farmer, miller, trader, company, coopera- farmers. Private warehouse operators normally have little tive, or farmer association) gives the receipt to the lender in interest in dealing with the small transactions of individual exchange for a loan. The lender offers a percentage of the farmers. Farmer associations and cooperatives may be more value to cover the cost of selling the commodities in the important users and managers of warehouses, but they are event of loan default, as well as any price declines that may often poor credit risks for lenders because of poor man- occur before liquidation. More sophisticated lenders can use agement and weak governance. Moreover, it is difficult for hedging to reduce price volatility. The warehouse operator member-owned institutions to enforce grades and standards will not release the commodity without authorization from for commodities delivered by their members. the lender. Usually the borrower who owns the commodity plans to use or dispose of it and repay the loan before the 9 3 2 Warehouse receipts financing in Africa due date, but the lender has the legal right to sell it in case of default. Repurchase agreements, factoring, and other forms Warehouse receipts financing seems to be less well de- of collateralized transactions are used in more advanced veloped in Africa. It is often reported to be unavailable for markets. smallholders and is used mainly by a few large borrowers, usually importers, under expensive collateral management Warehouse receipts lending can reduce the risk and trans- agreements involving international inspection companies. action cost impediments for agricultural lenders caused by Models funded by donor NGOs have often failed because small size transactions and high information and supervision of limited scale economies and government policies that costs, so interest rates may fall for borrowers. securely damage incentives for storage by importing commodities in stored commodities may be second only to land as preferred the event of expected grain shortages (Coulter and Onumah collateral, providing that the legal procedures for selling 2002; Coulter 2009). stored commodities are efficient. Through the conversion of commodities into collateral for short-term loans, borrowers A more inclusive model of warehouse receipts financing was can preserve their land, buildings, and other assets as col- proposed for Zambia. This proposal suggested making com- lateral for long-term loans. mercial finance more readily available through a network of privately managed warehouses authorized to issue transfer- able warehouse receipts. To engender confidence, an arms- 9 3 1 Prerequisites for warehouse receipts lending length, self-financing regulatory agency, insulated from direct There are at least seven prerequisites for successful ware- government control would be created to certify and inspect house receipts lending: (1) an appropriate legal system that warehouse operators. Certification would be based on meet- essentially treats warehouse receipts as cash; (2) active ing criteria such as suitability of warehouses, experienced commodity markets to value and liquidate commodities; (3) management, minimum net worth, insurance, a bond, and a system of grades and standards for classifying commodity acceptance of frequent unannounced inspections. The cer- quality; (4) regular patterns of postharvest seasonal price in- tification agency was designed to operate on user fees but creases sufficient to compensate for storage and borrowing with subsidies in its initial years. The objective was to break costs; (5) appropriate financial, technical, and administrative even quickly by increasing the number of warehouses and S U B S I D I E S A S A N INST RUME NT IN AGRICULT URE FINA NCE: A REV IEW 39 the range of crops stored. The service would be available to The second case is Uganda, where UsAID implemented a producers, processors, and traders with a minimum grain pilot project with a warehouse receipts component for the deposit of 10 to 30 tons. Commodities to be stored initially 2,100-member Kapchorwa Commercial Farmers Association would be maize, wheat, and soybeans meeting prescribed (KACOFA). It was designed to increase maize farmers’ in- weight and grading standards, with expansion later to comes by overcoming the cyclical nature of farm income other crops. The certified warehouse operators would own and lack of access to credit (UsAID/ Rural sPEED 2006). or lease sheds or silos on commercial terms and would The stanbic Bank branch agreed to lend up to 80 percent of be free to charge economic storage rates (Onumah 2003). the value of farmers’ maize stored in the association ware- Although the features of this design seem sound, Coulter house while it participated in UsAID’s DCA loan guarantee (2009) discussed the problems encountered in implement- program, also managed by the Rural sPEED project (see ing a warehouse receipts program in the country. A ma- section 8.2.2.4. on guarantees). Two other UsAID projects jor remaining constraint is the need to make appropriate worked with KACOFA to increase farmer production and im- changes in the agricultural credit act. prove postharvest handling to improve maize quality. UsAID helped develop improved warehouse receipts legislation, Two African case studies summarize the challenges and while other donors supported farmer development activities accomplishments of projects to expand warehouse re- in the region. To increase procurement from small farmers, ceipts lending. A village-level rice inventory credit product the World Food Programme (WFP) agreed to buy high-quality called “Grenier commun villageois� (GCv) is offered by maize from the warehouse at 350 Uganda shillings per kilo- the Caisses d’Épargne et de Crédit Agricole Mutuels de gram, compared to the 120–180 shillings paid by local trad- Madagascar (CECAM) in Madagascar (Bouquet, Wampfler, ers. UsAID subsidized the warehouse collateral manager for and Ralison 2009).47 six-month loans allow producers to the first year, but KACOFA was expected to produce enough store harvests until the lean season, when market prices grain to make subsidy unnecessary the second year. are normally higher. After repaying the loan, the producer can either (1) consume the stock or (2) sell it, thereby real- The amount of grain delivered to the warehouse dramatically izing the difference between the harvest and lean period exceeded initial expectations. As a consequence, KACOFA prices. The minimum quantity required for a loan is only considered expanding the warehouse receipt system to bar- 75 kilograms, so it is easily accessible to small-scale pro- ley and beans, and similar programs are being developed in ducers who cannot provide the collateral required for other other parts of Uganda. However, because many donor proj- loan products. storage is provided in local warehouses se- ects were involved, it is difficult to disentangle the impact cured by two locks, one kept by the CECAM credit officer of the warehouse receipts on lender behavior or to clearly and one by the representative of the warehouse. Interest identify all prerequisites for success. The complex reality of rates are charged at 3 percent per month for a minimum of operating a warehouse receipts system was becoming clear five months. and was viewed as an expensive undertaking for the nascent farmers and farmer groups. The substantial requirements of No third party is involved. CECAM monitors the storage, operating a suitable collateral-managed warehousing facility and the members who store their inventories are respon- (especially insurance, secure premises, and cleaning and sible for maintaining the stock in good condition. Both rich drying facilities), in addition to the high cost of managing and poor producers use the storage, but poorer households collateral, were considered too large and complicated to be reportedly use it for consumption smoothing more than the handled by a small farmers’ organization without substantial rich households. They value the GCv as a means to forcibly external financial support (Besigye 2009). save rice for family consumption until the lean season. But the poor are also highly dependent on agricultural income Another recent effort in Uganda attempts to deal with these and unable to engage in remunerative off-season activities, problems. It involves an agreement between the WFP and so may have a hard time repaying the loan. This situation the Uganda Commodities Exchange to construct and reha- may force them to take expensive informal loans or sell bilitate warehouses for the WFP maize procurement and their stored rice in advance to a local trader at a discounted storage program. Discussions are under way with banks price. concerning the issuance of warehouse receipts for use as loan collateral (Bashaasha and Odeke 2010). 47 village-level warehouse receipts systems operated by banks are reported to exist in India, but no information was obtained about them. 40 C H A P T E R 9 — F IVE M Aj OR A P P ROACHE S TO SUPPORT A GRICULTUR A L FINA NC E A ND DES IGN OF SUBS IDY INFLOW S several public good investments have been identified to en- 9 4 sPeCIALIzeD AGRICULtURAL hance warehouse receipts systems in Africa (Coulter 2009). DeVeLoPMent BAnKs48 For example, to achieve widespread provision of such ser- several forms of institutions deliver financial services in rural vices, it will be necessary to establish accreditation or licens- areas, and they offer different combinations of advantages ing to build up confidence in the industry. such systems will and limitations. For example, cooperatives, credit unions, ensure standardized documentation, particularly electronic and other member-owned institutions may be able to reach documentation, and establish uniform performance guar- more distant locations with cheaper services because they antees to protect depositors against warehouse failure or are located closer to the clients, have access to local infor- bankruptcy. such systems and safeguards increase financial mation, and rely on volunteers for administrative functions. institutions’ confidence and reduce their transaction costs in However, membership organizations frequently experience dealing with the systems. These systems will also require governance problems. small unit or community banks are strict regulatory processes to prevent fraud. If countries also located close to clients and share the problem of be- opt for direct public sector regulation, authority needs to be ing vulnerable to localized shocks, such as crop failures, that vested in the body least susceptible to political interference. damage many clients simultaneously. Commercial banks Coulter also argues for a strategy of developing warehouse have the capacity to offer multiple services, provide more receipts systems within a framework of overall market devel- safety for savings because they are regulated, and have ex- opment and integrating all market participants, including the tensive branch networks, but they are expensive institutions WFP, to achieve sufficient scale. for managing small loans (Zeller 2006). Conclusions. There are too few careful analyses of ware- specialized state-owned agricultural development banks house receipts systems to evaluate when and where they (AgDBs) are of particular interest. Many were created as part make important contributions to improved access to agricul- of the subsidized directed-credit paradigm. Generally they tural credit, especially for small farmers. They may improve performed poorly, although there have been notable excep- commodity storage and marketing functions in value chains, tions. Many have been privatized or closed, especially in with trickle-down benefits in terms of prices paid to farmers. Africa and Latin America; some are technically bankrupt and The high costs associated with creating, operating, and mon- continue to limp along but are unable to attract substantial itoring these systems imply that scale is a serious challenge, new funding. The loss of rural banking outlets that occurs so simple, small-scale village-level systems may be the most with closure, plus some well-known successful examples appropriate way to benefit small farmers directly. Moreover, of reform, has renewed the debate regarding the appropri- the critical need for small farmers may be production loans ate strategy for dealing with these institutions (World Bank that meet seasonal cash outflows at the beginning of the 2006). planting season rather than marketing loans after harvest. There is a also a critical need for more analysis of farm-level 9 4 1 Principles for reforming agricultural development commodity price data to determine which crops normally banks experience seasonal price variations large enough to com- Advocates make strong arguments in favor of AgDB reform. pensate for storage and borrowing costs. If price variations They emphasize the potential for serving the rural poor if are too small or irregular, there is little economic rationale for reforms are implemented properly. The successful reforms the public sector to subsidize a warehousing strategy. The of BRI in Indonesia and the evolution of BAAC in Thailand fact that warehousing is common for export crops suggests provide evidence of the possibilities (seibel 2000). seibel, that analysis is needed to understand the economic barriers Giehler, and Karduck (2005) propose that AgDBs should be that constrain expansion into grains and other commodities transformed into self-reliant, sustainable financial intermedi- produced primarily for local markets. sequencing may also aries based on the following four principles: be important. Once farmer associations and cooperatives have established a successful track record of performance, they may consider undertaking warehouse management and linkages with financial institutions to benefit their members. 48 This section focuses on the long-term development impacts of agricultural development banks. Except where specifically several long-term public good investments have been identi- noted, many of the ideas presented here are drawn from the lit- fied to make warehouse receipts financing work effectively in erature reported in Nagarajan and Meyer (2005). Rudolph (2010) Africa. Many of them may be appropriate for donor funding. looked at the question of the role of state banks in responding to the financial crisis. S U B S I D I E S A S A N INST RUME NT IN AGRICULT URE FINA NCE: A REV IEW 41 ƒ mobilization of domestic resources and provision of ƒ obtaining a commitment from political stakeholders positive real returns to depositors; that they will refrain from interfering in the banking ƒ repayment of loans and coverage of costs from opera- business (Kanathigoda and steinwand 2003). tional income; ƒ production of sufficient retained earnings to offset Additionally, banking decisions can be shielded from politi- the erosion of resources from inflation and to finance cal influence by requiring reformed banks to obtain funds for expansion; and refinancing from capital markets rather than from govern- ƒ continually increased outreach to savers and borrow- ments. Good economic performance requires improving ers and improved quality of services provided to all loan recovery though suitable lending policies and incentive segments of the rural population, including the poor. systems, improved efficiency and staff productivity, and free- dom to set interest rates that cover costs and losses. several Latin America has had many negative experiences with reformed institutions have found that demand from rural sav- AgDBs, and several have been closed. There are cases (such ers has been greater than loan demand (World Bank 2006).49 as in Ecuador) where they continue to operate but provide poor-quality financial services and depend on periodic recapi- 9 4 3 Adopting microfinance technologies talizations. Attempts at reform in some countries have failed Another strategy for reforming AgDBs is to broaden their because of lack of commitment by major stakeholders. One functions, adopt microfinance lending technologies, and, successful reform (as measured during the first four years in several cases, reduce their exposure to agriculture.50 A following reforms) was the transformation of BANDEsA, a successful Asian reform case was the Agricultural Bank of government-owned development bank, into BANRURAL Mongolia (AgBank, now called XAAH), which designed new s.A. in Guatemala in 1998. It was turned from a government- loan, deposit, and money transfer products using lessons owned bank into a mixed capital company (70 percent private learned in microfinance, upgraded the staff, required greater sector and 30 percent public sector ownership), and the tar- accountability for managers, and developed new reporting get market was specified as farmers, merchants, artisans, systems. By 2002, the reforms were so successful that it and micro, small, and medium entrepreneurs. Guatemalan was completely privatized (Dyer, Morrow, and Young 2004). autonomous organizations, including Mayan and Xinca indig- enous groups and small farmers’ legally recognized organiza- An African success story is the reform that created the tions, were included as owners. These changes in owner- National Microfinance Bank (NMB) in Tanzania. In 1997 the ship and governance were important factors in improving state-owned National Bank of Commerce (NBC) was sepa- outreach and financial performance (Alfaro-Gramajo 2003). rated into two entities: the old NBC retaining mainly urban outlets while the new NMB received the rural network with 100 outlets. The NMB also processes government payments 9 4 2 Ownership and governance throughout the country. The government helped the bank Ownership and governance have been identified as the most make loans to creditworthy clients and resisted political in- important success factors in reforming Asian public banks. terference. The transfer products were revamped, and loan Most loan losses are attributed largely to politically motivated products were developed for microenterprises, small-scale interference rather than nonrepayment from “normal� bor- farmers, and employees. Financial performance improved rowers. several approaches have been identified to limit enough so that by 2002 it was in the process of privatization political influence: (Dressen, Dyer, and Northrip 2002). ƒ separation of politically initiated programs from normal banking business; 9 4 4 Risk management techniques ƒ replacement of political stakeholders with managers If financial institutions, whether private or public, are going from the private sector; to specialize in agriculture, they must find ways to manage ƒ diluting the degree of political ownership through issu- ance of shares; 49 Trivelli and venero (2007) report that development banks in Latin ƒ imposing stricter regulatory requirements similar to America that lend to agriculture are successful for similar rea- sons. commercial banks; and 50 A risk management technique widely used by successful agricul- tural lenders in Latin America is to limit portfolio exposure to less than 40 percent of total lending (Wenner et al. 2007). 42 C H A P T E R 9 — F IVE M Aj OR A P P ROACHE S TO SUPPORT A GRICULTUR A L FINA NC E A ND DES IGN OF SUBS IDY INFLOW S risks. The CRDB bank in Tanzania, a state-owned rural de- microfinance technologies. As institutional capacity grows, velopment bank privatized in 1996, offers an example. Its a more ambitious strategy with larger loans can be slowly borrowers are heavily involved in cotton and coffee and are implemented. exposed to international price risks that caused other banks to withdraw from lending. The government export guarantee fund is costly, and there are difficulties in settling default 9 5 AGRICULtURAL InVestMent FUnDs claims. To reduce risks, CRDB requires all coffee and cot- Investment funds are a relatively new method of financing ton clients to use collateral management arrangements and agriculture. They are expanding rapidly and are of interest is working to get access to international markets for price here both because of the positive direct effects they may hedging. By purchasing options contracts for clients, it ex- have on agricultural investments and because of the addi- pects to lock in floor prices that protect against downward tional agricultural lending that may be induced because of price movements (World Bank 2005b). the investments made. By observing the performance of the investments made by the funds, local investors and financial Conclusions. Three key lessons emerge from these reforms institutions may also identify other creditworthy investments of agricultural development banks. First, successful reforms to be financed, as well as pitfalls to avoid. of AgDBs are possible only if governments are prepared to make fundamental changes in ownership, governance, prod- An agricultural investment fund is similar to a financial mutual ucts, and services, and perhaps even the clientele served. fund that pools capital from different investors and allocates second, selected successful microfinance procedures and the funds for agricultural investments meant to generate lending technologies need to be adopted and applied to profits for the investors. Funds offer investors an opportunity agriculture. Third, more sophisticated risk management for risk pooling through diversified investments while em- techniques will be needed for those financial institutions that ploying specialized professional fund managers to manage expect to devote a large proportion of their loan portfolios to investments. The managers conduct risk assessments of large loans to farmers and nonfarm businesses closely linked alternative investment opportunities, administer the invest- to agriculture. A significant challenge is that most financial ment portfolio, and have fiduciary responsibilities to the in- institutions do not have experience using more sophisticated vestors. The funds may have social or altruistic objectives, risk management techniques (such as insurance, hedging, such as combating hunger and poverty, but private sector futures markets, derivatives, and swaps), and a minimum investors increasingly recognize that attractive financial re- scale is often necessary for their adoption (Wenner et al. turns can be realized if invested properly in agriculture. 2007; Nair 2008). A greater knowledge and use of value chain finance may also help resolve problems of lending 9 5 1 Microfinance investment funds costs and risks. Although investment funds are new to agriculture, they have A problem for donors is to develop a response in countries a longer track record of public-private sector partnerships in where local leaders create a new AgDB out of frustration funding microfinance, and this experience may have insights with the slow pace of agricultural lending by other financial for agricultural investment funds. The success of these funds institutions. Although the World Bank reports success with has attracted considerable investor interest. The first com- the creation of two banks in the former soviet Union,51 the mercially focused investment structure targeted for MFIs negative experiences with political interference in many was Profund launched in 1995. By mid-2005, 23 investment AgDBs do not bode well for such projects. A necessary con- funds provided equity to MFIs with total assets amounting dition for successful start-ups is an institutional design that to about 536 million euros (equivalent to about $725 million solves the governance and management problems already at the time). About 262 million euros ($355 million) of the noted and maintains a successful firewall between credit total was invested in microfinance (Goodman 2009). About operations and political interference. A critical component is two-thirds of the assets were in microfinance development government commitment to charging full cost recovery inter- funds that place more emphasis on furthering development est rates. It may also be possible to avoid political capture by than on earning financial returns. Commercially oriented adopting an initial strategy of targeting small loans and using microfinance investments funds with expected financial re- turns higher than those of microfinance development funds actually made a larger share of equity investments in MFIs. 51 The Agricultural Development Bank of Latvia and the Kyrgyz Ag- ricultural Finance Corporation (World Bank 2006). S U B S I D I E S A S A N INST RUME NT IN AGRICULT URE FINA NCE: A REV IEW 43 Microfinance was just beginning to attract private investors at 9 5 2 Investment funds for agriculture that time (Goodman 2009). The FAO studied 31 investment funds for agriculture involv- ing a total of $7 billion (Miller et al. 2010). The majority have The Profund history is a revealing case. It invested in what less than $100 million in capital and are global or focus on became some of the well-known Latin American success sub-saharan Africa. Most are public-private partnerships that stories (Compartamos, Bancosol, and Mibanco), but it also draw in private capital to leverage public resources. About incurred losses on other investments. At the final 2005 one-third of the funds were created during the past three shareholders’ meeting, it had provided an average return to years with only private capital. Their recent creation reflects shareholders of 6 percent a year. An important conclusion donor interest in becoming more heavily engaged in foreign was that it initially expected to balance its investments be- investment as the demand for food, fiber, and other agricul- tween intermediaries catering to small businesses and those tural products continues to rise. The funds invest in equity serving microenterprises. However, market forces and prag- only, debt only, mixed equity/debt, microfinance, guarantees, matism led it to concentrate on the latter for three reasons: and miscellaneous categories. (1) microentrepreneurs are more flexible and resilient and better at coping with downturns; (2) they have fewer financ- Private capital comes from individuals, institutional inves- ing options, so they see credit from a formal institution as a tors, and foundations whereas public funds are provided by privilege and feel a more serious obligation to repay loans; international donors and development finance institutions. and (3) they normally operate with low levels of capitalization, Most funds target the “missing middle,� including larger so money lent to them generates large marginal productivity, processing companies or agricultural sMEs that are too large allowing them to thrive while still paying high interest.52 to receive MFI loans or that require equity investments. Two- thirds of the funds have a social and development mission, A huge expansion in investment funding for MFIs was report- such as agribusinesses with sound environmental and social ed by MicroRate in its 2010 survey (MicroRate Incorporated practices or investments supporting women entrepreneurs. 2010). It concluded that 88 microfinance investment vehicles Public funding is often used for the considerable costs and (MIvs) had more than $6.0 billion in total assets at the end time involved in setting up funds. At least 50 percent of of 2009. At the end of 2009, the MIvs held a total of 3,033 funds provide technical assistance to strengthen capacity so microfinance investments with an average investment size the investments are more productive. Technical assistance of $1.4 million. Debt comprised approximately 81.6 percent, may also contribute to mitigating investment risks and help followed by equity at 17.6 percent, guarantees at 0.5 percent, cover the costs of helping small farmers participate in value and other microfinance assets at 0.3 percent. Most MIv as- chains where investments will be made. sets were in Latin America and the Caribbean (37 percent) followed by Eastern Europe and Central Asia (35 percent), but Expectations for returns vary between 3 and 25 percent the fastest growth, admittedly from a low base, occurred in depending on the fund’s orientation. Most investors have East Asia and the Pacific. mixed social and profit objectives. some public investors treat their capital as a grant while some private investors Although the MIvs attracted more than $1.0 billion in new expect a close-to-market rate of return. Impact data showing funding during the year, they found it difficult to place the large returns to investors, employees, suppliers, consum- funds, with the result that less than half of the funding mo- ers, competitors, and the community were provided for the bilized ended up in microfinance. Part of the explanation investment funds managed by small Enterprise Assistance was the decline in demand for funds from MFIs owing to a Funds (sEAF 2007). Although considerable data were col- slowdown in loan disbursements. MicroRate concluded that lected from the firms in which the fund invested, the meth- liquidity has reached unsustainable levels and that pressures odology used was too weak to argue that the fund “caused� to disburse funds, coupled with the decrease in investment the impacts reported. opportunities, could lead to a deterioration of portfolio qual- ity as fund managers seek to reduce liquidity. This situation New investment funds continue to be developed. A re- could be an opportunity for the MIvs to strengthen their op- cent example was the 2009 announcement of the 10-year erations and focus on delivering the products and services African Agricultural Fund, a $300 million fund supported that microfinance institutions truly require. by the African Development Bank, Agence Française de Développement, Alliance for a Green Revolution in Africa, 52 Reported in an unofficial history entitled “ProFund Internacional, Banque Ouest Africaine de Développement, Ecowas Bank sA� (n.d.) supplied courtesy of Tomas Miller, IADB. 44 C H A P T E R 9 — F IVE M Aj OR A P P ROACHE S TO SUPPORT A GRICULTUR A L FINA NC E A ND DES IGN OF SUBS IDY INFLOW S for Investment and Development, and IFAD. It will invest costs mean that special measures will be needed to inte- in agricultural businesses (established as companies or co- grate poor farmers into value chains that benefit from these operatives) operating in food production industries, or pro- investments. Moreover, unless these funds invest in devel- vide financial services to small agribusiness operators and oping financial institutions that serve agriculture, they will not sMEs, cooperatives, or farmers’ organizations. A Technical contribute to broadening the supply of rural savings, insur- Assistance Facility, funded through grants from bilateral and ance, and other financial services important to farmers and multilateral institutions and private foundations, will finance rural people. feasibility studies, training programs, and other external ser- vices useful to the implementation and long-term monitor- International agencies can play an appropriate and productive ing of the fund. The fund’s targeted rate of return is 6 or 8 role by subsidizing intensive monitoring and analyses of fund percent, depending on the category of shareholder.53 activities. Providing subsidies for the technical assistance components of the funds may also be helpful to strengthen An Oxfam report presented a cautious interpretation of the local capacities, make the investments more productive, potential impact of socially responsible investment funds mitigate risks, and cover some of the costs of helping small (Doran, McFayden, and vogel 2009). The authors argue farmers participate in the value chains where investments that high transaction costs and fund economics mean that are being made. If the technical assistance helps facilitate such funds must necessarily exclude small investments. direct investments in financial institutions, the funds may With small management teams of highly paid professionals, make important contributions to broadening the supply of it is not feasible for funds to make and closely supervise rural financial services. investments of substantially less than $1 million, even if there may be subsequent larger financing rounds for some investments. Most funds are based in Europe or the United states, so transaction costs will be high unless they can ef- fectively resolve the problem of linking with a local agent or company to facilitate deal-sourcing, due diligence, and post- investment support. Countries where external investment is rare and information and contract problems are serious may attract little investment because legal protections for investors may be weak and exit options limited. Conclusions. Evaluations are needed of the activities of in- vestment funds to determine their impact and the extent to which they induce more local lending by financial institu- tions. Estimates of agricultural investment requirements in developing countries are very large, so a logical argument can be made for more external investments. Important learn- ing about the potential returns and risks of investing in ag- riculture can occur if the funds will share their experiences. It is likely that the market segment that will benefit most from fund investments will be the more affluent and en- trepreneurial farmers and agribusinesses, so wealth and income distribution implications may be important, as they are with many improvements in access to finance. Benefits in the form of better access to inputs, markets, and jobs flow through the value chain linkages to smallholders, and some other benefits may trickle down to the poor in general. But high information, transaction, and contract enforcement 53 African Agriculture Fund, “summary of Principal Terms and Con- ditions,� draft dated March 15, 2010. S U B S I D I E S A S A N INST RUME NT IN AGRICULT URE FINA NCE: A REV IEW 45 Chapter 10: ConCLUsIons AnD eMeRGInG ReCoMMenDAtIons This paper reviewed the challenges faced in developing sus- credit systems. With a few noteworthy exceptions, tainable agricultural credit supplies for small farmers in devel- the old-paradigm approach managed to push out loans oping countries. Donors and governments have spent billions but generally did not lead to sustainable agricultural subsidizing programs and policies to develop and strengthen credit institutions. success requires careful develop- financial institutions. Nonetheless, national decision makers, ment of products, policies, institutions, and supportive international donors, and farmers in nearly all countries are infrastructure. dissatisfied with the supply and cost of agricultural credit. Market-oriented financial reforms were implemented follow- 2. Back to basics. Economists are asking some funda- ing the collapse of the directed-credit paradigm, but critics mental questions, such as, Why are there such large claim they have failed because in most countries agriculture wedges or gaps between the rates for depositing and continues to receive only a small share of a country’s total lending in developing countries? Why would some formal credit. Most farmers report that they rely on their own people pay so much more than others for loans? Why savings or loans from informal credit suppliers, family, and don’t banks simply raise interest rates high enough to friends to finance working capital and investments. make lending to the poor remunerative? Are the high costs of small loans explained largely by the high fixed The first part of this paper reviewed how the financial sys- administrative costs of lending, or are they also af- tems approach was successfully used to develop the micro- fected by adverse selection and moral hazard? Is the finance industry. This experience provides lessons useful for impressive success of microcredit explained, at least developing agricultural credit markets. Likewise, the debates in part, by its ability to generate reductions in moni- about the use of grants and subsidies in food, fertilizer, and toring costs? Does group liability work, not because credit markets were reviewed as guides to future interven- of the formal structure of liability, but because, after tions by governments and international agencies. being together for a while, the group members began to value relationships with other members? Is one of The second part of the paper highlighted the literature cover- the basic values of microcredit the fact that it commits ing the rationale for and experiences of international agencies the borrower to a savings plan and helps avoid tempta- in five major program areas in support of agricultural finance: tion spending (Banerjee and Duflo 2010)? The rigorous microfinance, microinsurance, and weather-index-based methods used by this new group of researchers pro- insurance, credit guarantee funds, warehouse receipts, vide good prospects for deepening our understanding specialized agricultural development banks, and agricultural of human behavior and how it influences credit mar- investment funds. ket operations. This research, plus the lessons learned What are the major conclusions derived from the literature from the many innovations being tested around the consulted for this review? What lessons can be learned, and world, needs to be widely disseminated for the ben- what are the frontiers for policy and projects that CABFIN efit of the entire financial industry. supporting and members might support in their projects and programs? learning from research and innovations will provide international agencies with many opportunities to help push out the frontier of agricultural credit in develop- 10 1 ConCLUsIons ing countries and learn how to use selective subsidies and investments to make the greatest impact. 1. No magic bullets. There are no simple magic solu- tions in the toolkits available to governments and inter- national agencies for creating sustainable agriculture 46 C H A PTER 10 — C ONC LUSIONS A ND EMERGING REC OMME ND ATIONS 3. Microfinance offers a partial solution. The financial variety of techniques, MFIs have found ways to re- systems approach has contributed to a thriving micro- duce information and contract enforcement problems finance industry that is now slowly penetrating rural typical in credit markets. Group lending was an im- areas, but the highly standardized MF products and portant early innovation, but much of the industry has lending methodologies are not well suited to farming successfully evolved to provide individual lending. clients. MFIs in many countries are still clustered in Innovation grants made to MFIs have contributed to urban and peri-urban areas or in densely populated the search for new innovations, and several reformed rural areas. There may be considerable payoff to sub- AgDBs owe their success to their emulation of MF sidizing the learning and innovation costs of MFIs and approaches. Agricultural credit needs similar careful other local financial institutions that are committed to innovation and design to be successful. value chain entry into agriculture lending. The use of agents and finance helps identify opportunities for innovations. mobile phones are currently among the promising 7. Diversified loan portfolios. specialized agricultural methods for driving down the costs of rural banking. development banks were considered key to suc- cessful agricultural finance under the old paradigm. Although specialization may lead to improved knowl- 10 2 eMeRGInG ReCoMMenDAtIons edge on how to serve agriculture, evidence from re- 4. Avoid interest rate controls. The directed-credit par- formed AgDBs and successful MFIs and agricultural adigm often involved subsidies to farmers in the form lenders suggests that a diversified loan portfolio is of subsidized interest rates, which saved them money needed to help manage the risk associated with co- on interest charges but undermined the financial insti- variate agricultural incomes. No data have been found tutions that lent to them. These subsidies also tended to confirm the argument that agricultural loans are to distort the allocation of loans in favor of richer farm- more risky than others, so an important empirical ers. Microfinance has thrived in spite of high operat- task is to measure the correlation in incomes among ing costs and risks because it has generally been able farm and nonfarm enterprises to evaluate how well to avoid interest rate controls, although the industry portfolio diversification may reduce credit risk. In the is threatened in several countries that have imposed meantime, prudent policies should be followed, such interest rate ceilings. International agencies need to as setting a ceiling on the share of agricultural loans continue their strong efforts to inform and advocate in a loan portfolio, especially in the early stages of ex- on behalf of market-oriented interest rates. There is perimentation. International agencies can make a use- evidence that rates of return in agriculture are often ful contribution by helping financial institutions reduce higher than assumed, so cheap interest rates may be their exposure to systemic risks by improving their ca- less critical to borrowers than policy makers expect. pacity to lend to both farm and nonfarm activities in ru- 5. Subsidizing institutions and infrastructure rather ral areas and to lend in multiple geographic areas. The than borrowers. subsidies for building institutions present enthusiasm for value chain finance needs to and financial infrastructure contributed to the success be tempered with the recognition that excessive lend- of microfinance, and such subsidies are generally ac- ing to participants in one chain leads to portfolio risk. cepted as part of the market development strategy for 8. Subsidizing public goods. subsidies to create pub- rural finance. This type of indirect subsidy to borrow- lic goods that benefit the entire financial sector can ers is generally considered to be less distorting than generate higher returns than subsidies to specific the direct interest rate subsidies granted to borrow- institutions and may be critical once the immediate ers under the old paradigm. The key to lower interest priorities of creating products and institutions are met. rates for agricultural credit is increased efficiency and Improving property rights for agricultural assets, col- competition, as experiences with microfinance have lateral registries, credit bureaus, special courts for shown in several countries. credit defaulters, and other support institutions help 6. Investments in lending technologies and institu- the entire financial sector resolve information and con- tional design. MFIs have also demonstrated that de- tract enforcement problems. International agencies sign matters. They have successfully created lending can play their logical role in championing a longer-term technologies and institutional designs to match the view of credit market development by identifying needs and capacities of a poor clientele. Through a S U B S I D I E S A S A N INST RUME NT IN AGRICULT URE FINA NCE: A REV IEW 47 gaps in support institutions and proposing methods to analyzing alternative designs and for long-term public address them. goods investments in networks of weather reporting 9. Subsidizing networks. The many national and inter- stations and data collection and analysis. subsidies national networks related to microfinance provide pu- for individual countries may be justified: the lack of bic goods by facilitating the exchange of information private sector initiative implies first-mover problems and innovation within the industry. subsidies for inno- in which private investors resist making initial invest- vations may produce a larger payoff when channeled ments in research and development, fearing the ease through networks because the benefits will be spread with which competitors can copy their products. Two among all members. subsidies received by CGAP, the key types of analyses are needed. The first is analysis MIX, and the Microcredit summit generated substan- of the demand for such crop and livestock insurance, tial benefits because of the leadership and sources recognizing that uptake by farmers is poor in most de- of innovation they provided to microfinance. By com- veloped countries without huge subsidies. The second parison, the networks for agricultural credit and rural consists of robust ex post evaluations to determine if finance are not as well developed and could benefit insurance has the expected effects on farmers’ liveli- from this same type of investment and leadership. hood strategies and incomes and if it helps protect 10. Experiment and evaluate. What gets measured gets lives and assets, enabling people to avoid or escape attention. Compared with the large amount of funds poverty. These results will be especially important in invested in programs and projects to support agricul- determining the appropriate role of public subsidies tural credit, surprisingly little is spent on monitoring for catastrophe insurance. and evaluation to determine if the expected results 12. Credit guarantee schemes. Credit guarantee are being achieved. The evaluations that are conduct- schemes play a surprisingly large role in development ed lack the same robustness that is now being advo- programs considering the lack of international consen- cated and designed for microfinance impact analysis. sus about their impact on access to finance. Recent Effective project design requires that a few robust evaluations of UsAID guarantees used weak meth- analyses be conducted in the five major areas of do- odologies and thus constitute an unreliable source of nor support discussed in the second part of this paper. evidence on effectiveness. The training and technical 11. Microinsurance and weather-index-based insur- assistance provided to financial institutions along with ance. Microinsurance is expanding quickly, and the guarantees may be far more important in encourag- proper role for private and public sector support will ing lenders to improve credit access to a target group eventually become clearer. Weather-index-based in- than the guarantee itself. Once again, design matters, surance needs international support for testing and but critical design features of eligibility, leverage, risk sharing, fees, and claims procedures are not well discussed or evaluated in projects. The international agencies should conduct a few robust evaluations to determine if credit guarantees produce the expected results, to identify the best designs for assuring ad- ditionality and sustainability, and to assess the pre- conditions (such as commitment and interest of the participating financial institutions) and complementary activities (such as training and technical assistance) required for success. 13. Warehouse receipts. Warehouse receipts proj- ects were pioneered by the European Bank for Reconstruction and Development in Eastern Europe and the Commonwealth of Independent states coun- tries in the 1990s. They are not yet popular, especially in Africa, because this traditional form of commodity Monitoring and evaluation, Capetown, South Africa (Photo: IMA International) collateralization is mostly limited to export crops. This mechanism is expected to improve access to finance 48 C H A PTER 10 — C ONC LUSIONS A ND EMERGING REC OMME ND ATIONS and possibly reduce interest rates by reducing the the United states, so transaction costs are high and risk and transaction costs for agricultural lenders fac- they have small management teams of highly paid ing small-size transactions and high information and professionals. It is not feasible, therefore, for them to supervision costs. However, it will make only a small make small investments important for small farmers. contribution to the objective of increasing production Countries with minimal external investment and se- credit for small farmers, and warehouse operators rious information and contract problems may attract may have little interest in dealing with small transac- little investment because legal protections for inves- tions of individual farmers. Farmer associations, co- tors may be weak and exit options limited. The mar- operatives, and private traders may be more frequent ket segment most likely to benefit from fund invest- users. systems of grades and standards must be cre- ments will be the more affluent and entrepreneurial ated and enforced in parallel for these systems to be farmers and agribusinesses, so it will be important to effective. The large expense of operating these sys- monitor wealth and income distribution implications. tems is a serious challenge for small farmers, so sim- The critical role for donors will be financing subsidies ple, small-scale village-level systems may be more ap- for technical assistance components of the funds. propriate. There is a critical need for farm-level price Technical assistance will strengthen local capacity to data to identify which crops normally have seasonal make investments more productive and will mitigate price variations large enough to compensate for stor- the risks and costs of helping small farmers partici- age and borrowing costs. The fact that warehousing pate in the value chains where investments will be is common for export crops suggests a need to look made. Moreover, if the technical assistance facilitates at economic barriers that constrain expansion into direct investments in financial institutions, the funds commodities produced for local markets. Resolving will contribute to broadening the supply of rural sav- these underlying barriers is a logical use for subsidies. ings, insurance, and other financial services important 14. Specialized agricultural development banks. A to farmers and rural people. more optimistic view has emerged about the possibil- ity of successfully reforming state-owned agricultural development banks in recent years. Ownership and governance have been identified as important suc- cess factors. A number of reformed banks have been successful by broadening their functions, adopting microfinance lending technologies, and reducing their exposure to agriculture. When these banks choose to specialize in agriculture, they must find ways to reduce lending risks, but most do not use the more sophisticated forms of risk management, such as in- surance, hedging, futures markets, derivatives, and credit swaps. International agencies should work with financial institutions and farmer groups to develop awareness and capacity so eventually they can adopt these methods. 15. Agricultural investment funds. Investment funds for agriculture are a relatively new method of invest- ing, so their impact cannot yet be assessed. The public-private sector partnerships in these funds are expected to help meet the huge projected demand for investments in developing countries. A large pro- portion of these funds have a social and development mission, so there is hope that sound environmen- tal and social practices or investments will be sup- ported. Most funds, however, are based in Europe or S U B S I D I E S A S A N INST RUME NT IN AGRICULT URE FINA NCE: A REV IEW 49 ReFeRenCes Adams, Dale W., Douglas H. Graham, and J. D. von Pischke, Arnold, Roger A. 2005. Microeconomics. 7th ed. Mason, OH: eds. 1984. Undermining Rural Development with Cheap south-Western. Credit. Westview special studies in social, Political and Banerjee, Abhijit v., and Esther Duflo. 2004. “Growth Theory Economic Development series. Boulder, CO, and London: through the Lens of Development Economics.� Working Westview Press. Paper 05-01, Department of Economics, Massachusetts Adams, Dale W., and J. D. von Pischke. 1992. “Microenterprise Institute of Technology, Cambridge, MA. Credit Programs: Déjà vu,� World Development 20 (10): ———. 2010. “Giving Credit Where It Is Due.� Journal of 1463–70. Economic Perspectives 24 (3): 61–80. AGRA (Alliance for a Green Revolution in Africa). 2009. “Early Banerjee, Abhijit v., Esther Duflo, Rachel Glennerster, and Accomplishments, Foundations for Growth.� June. Cynthia Kinnan. 2009. “The Miracle of Microfinance? Nairobi, Kenya. www.agra-alliance.org. Evidence from a Randomized Evaluation.� Unpublished Alamgir, Dewan A. H. 2009. “Microfinance in sAARC paper, Department of Economics, Massachusetts Region: Review of Microfinance sector of Bangladesh.� Institute of Technology, Cambridge, MA. Bangladesh. Unpublished paper. Bashaasha, Bernard, and Elvis Odeke. 2010. “Developments Alfaro-Gramajo, Luis Noel. 2003. “Reverting the Tendency in WFP’s Purchase for Progress (P4P) Programme.� In in Developing Finance: The Case of Banrural s.A. in Agricultural Finance Yearbook 2009, ed. Richard Roberts Guatemala.� Paper presented at “Paving the Way and Robert Ocaya, 140–43. Kampala: Bank of Uganda and Forward for Rural Finance: An International Conference Plan for the Modernization of Agriculture. on Best Practices,� organized by the U.s. Agency for Besigye, Asaph. 2009. “Financial Institutions Client International Development, Washington, DC, June 2–4. Perspectives from Kapchorwa.� In Agriculture Finance www.basis.wisc.edu/rfc. Year Book 2008, ed. Richard Roberts and Robert Ocaya, Alvarado, Javier, and Francisco Galarza. 2003. “The Agrocapital 59-62. Kampala: Bank of Uganda/GTZ-sida Financial Foundation of Bolivia: Pioneering Individual Loans in systems Development Program. Rural Areas.� In Promising Practices in Rural Finance: Besley, Timothy. 1994. “How Do Market Failures Justify Experiences from Latin America and the Caribbean, ed. Interventions in Rural Credit Markets?� World Bank Mark D. Wenner, Javier Alvaredo, and Franciso Galarza, Research Observer 9 (1): 27–47. 213–42. Washington, DC: Inter-American Development Bank. Binswanger, Hans P., shahidur R. Khandker, and Mark R. Rosenzweig. 1993. “How Infrastructure and Financial Anand, Malika, and Richard Rosenberg. 2008. “Are We Institutions Affect Agricultural Output and Investment in Overestimating Demand for Microloans?� Brief, India.� Journal of Development Economics 41: 337–66. Consultative Group to Assist the Poor (CGAP), Washington, DC. Binswanger, Hans P., and Mark R. Rosenzweig. 1986. “Behavioural and Material Determinants of Production Armendariz de Aghion, Beatriz, and Jonathan Morduch. Relations in Agriculture.� Journal of Development Studies 2005. The Economics of Microfinance. Cambridge, MA: 2 (3): 503–39. MIT Press. A j O I N T P U B L I C AT ION OF T HE WORL D B ANK, GLT N, FIG, A ND FA O 50 R EFER ENC ES Botero, Felipe, Craig Churchill, Michael J. McCord, and ———. 2009a. “Creating Pathways for the Poorest: Early Zahid Qureshi. 2006. “The Future of Microinsurance.� In Lessons on Implementing the Graduation Model.� Protecting the Poor: A Microinsurance Compendium, ed. Washington, DC: CGAP. Craig Churchill, 583–601. Geneva: International Labour ———. 2009b. “Microfinance Funding Continues to Grow: Office. Results of the 2009 Funder survey.� Washington, DC: Boucher, steve, Catherine Guirkinger, and Carolina Trivelli. CGAP. 2006. “Direct Elicitation of Credit Constraints: Conceptual ———. 2010a. “Branchless Banking Agents in Brazil: Building and Practical Issues with an Empirical Application to viable Networks.� Washington, DC: CGAP. Peruvian Agriculture.� Working Paper No. 07-004, Department of Agricultural and Resource Economics, ———. 2010b. “How Is the Financial Crisis Impacting Our University of California, Davis, CA. World and Our Work?� CGAP spreadsheet, Washington, DC. Bouquet, Emmanuelle, Betty Wampfler, and Eliane Ralison. 2009. “Rice Inventory Credit in Madagascar: Conditions CGAP/IFAD (Consultative Group to Assist the Poor/ of Access and Diversity of Rationales around a Hybrid International Fund for Agricultural Development. 2005. Financial and Marketing service.� Paper presented at the “Managing Risks and Designing Products for Agricultural First European Research Conference on Microfinance, Microfinance: Features of an Emerging Model,� Brussels, June 2–4. Occasional Paper No. 11, CGAP, Washington, DC. Buchenau, Juan, and Richard L. Meyer. 2007. “Introducing Chen, s., and M. Ravallion. 2009. “The Impact of the Global Rural Finance into an Urban Microfinance Institution: The Financial Crisis on the World’s Poorest.� http://tinyurl. Example of Banco Procredit, El salvador.� Paper pre- com/ccagl8. sented at the International Conference on Rural Finance Churchill, Craig, ed. 2006a. Protecting the Poor: A Research: Moving Research Results into Policies and Microinsurance Compendium. Geneva: International Practice, Food and Agriculture Organization of the United Labour Office. Nations, Rome, March 19–21. ———. 2006b. “What Is Insurance for the Poor?� In BWTP Network (Banking with the Poor Network). 2009. Protecting the Poor: A Microinsurance Compendium, “Microfinance Industry Report: Bangladesh.� Dacca. ed. Craig Churchill, 12–24. Geneva: International Labour Campion, Anita, Rashmi Kiran Ekka, and Mark Wenner. Office. 2010. “Interest Rates and Implications for Microfinance Churchill, Craig, and Denis Garand. 2006. “strategies for in Latin America and the Caribbean.� Inter-American sustainability.� In Protecting the Poor: A Microinsurance Development Bank, Washington, DC. Compendium, ed. Craig Churchill, 564–82. Geneva: Carter, Michael R., Francisco Galarza, and stephen Boucher. International Labour Office 2007. “Underwriting Area-Based Yield Insurance to Clarke, Daniel, and stefan Dercon. 2009. “Insurance, Credit, Crowd-In Credit supply and Demand.� Savings and and safety Nets for the Poor in a World of Risk.� DEsA Development 31 (3): 335–62. Working Paper No. 81, Department of Economic and CGAP (Consultative Group to Assist the Poor). 2000. “Those social Affairs, United Nations, New York, NY. Who Leave and Those Who Don’t Join: Insights from Conning, Jonathan, and Christopher Udry. 2007. “Rural East African Microfinance Institutions.� Focus Note 16, Financial Markets in Developing Countries.� In CGAP, Washington, DC. Handbook of Agricultural Economics, vol. 3, Agricultural ———. 2003. “Financial services for the Rural Poor.� Development: Farmers, Farm Production and Farm Donor Brief No. 15, CGAP, Washington, DC. Markets, ed. Robert Evenson and Prabhu Pingali, 2857– 2908. Amsterdam: Elsevier. ———. 2006. Good Practice Guidelines for Funders of Microfinance: Microfinance Consensus Guidelines. Washington, DC: CGAP. S U B S I D I E S A S A N INST RUME NT IN AGRICULT URE FINA NCE: A REV IEW 51 Coulter, Jonathan. 2009. “Review of Warehouse Receipt Dorward, Andrew, Ephraim Chirwa, Duncan Boughton, system and Inventory Credit Initiatives in Eastern and Eric Crawford, Thom Jayne, Rachel slater, valerie Kelly, southern Africa.� A working document for comment and Maxton Tsoka. 2008. “Towards ‘smart’ subsidies commissioned by the United Nations Conference on in Agriculture? Lessons from Recent Experience in Trade and Development (UNCTAD), under the All ACP Malawi,� Natural Resource Perspectives 116, Overseas Agricultural Commodities Programme (AAACP). Development Institute, London. Overseas Development Institute Coulter, J., and G. Onumah. 2002. “The Role of Warehouse Receipt systems in Enhanced Commodity Marketing and Dressen, Robert, Jay Dyer, and Zan Northrip. 2002. “Turning Rural Livelihoods in Africa.� Food Policy 27 (4): 319–37. Around state-Owned Banks in Underserved Markets.� Small Enterprise Development 13 (4): 58–67. Crawford, Eric W., T. s. Jayne, and valerie A. Kelly. 2006. “Alternative Approaches for Promoting Fertilizer Use in Duflo, Esther, Rachel Glennerster, and Michael Kremer. Africa.� Agriculture and Rural Development Discussion 2008. “Using Randomization in Development Economics Paper 22, World Bank, Washington, DC. Research: A Toolkit.� In Handbook of Development Economics, vol. 4, ed. T. Paul schultz and John strauss, Dehejia, Rajeev, Heather Montgomery, and Jonathan 895–962. Amsterdam: North-Holland. Morduch. 2009. “Do Interest Rates Matter? Credit Demand in the Dhaka slums.� Unpublished paper, Wagner Duflo, Esther, Michael Kremer, and Jonathan Robinson. 2008. Graduate school, New York University, New York, NY. “How High Are Rates of Return to Fertilizer? Evidence from Field Experiments in Kenya.� Paper presented at de Mel, suresh, David McKenzie, and Christopher Woodruff. the American Economic Association Meetings, January, 2007. “Returns to Capital in Microenterprises: Evidence New Orleans. from a Field Experiment.� Policy Research Working Paper 4230, World Bank, Washington, DC. Dyer, Jay, J. Peter Morrow, and Robin Young. 2004. “The Agricultural Bank of Mongolia.� In Scaling Up Poverty ———. 2009. “Are Women More Credit Constrained? Reduction: Case Studies in Microfinance. Washington, Experimental Evidence on Gender and Microenterprise DC: Consultative Group to Assist the Poor. Returns.� Unpublished paper, World Bank, Washington, DC. Eijkman, Frederik, Jake Kendall, and Ignacio Mas. 2010. “Bridges to Cash: The Retail End of M-PEsA.� Demirguc-Kunt, Asli, Thorsten Beck, and Patrick Honohan. Consultative Group to Assist the Poor, Washington, DC. 2008. Finance for All? Washington, DC: World Bank. El-Zoghbi, Mayada, Aude de Montesquiou, with syed Donald, Gordon. 1976. Credit for Small Farmers in Developing Hashemi. 2009. “Creating Pathways for the Poorest: Early Countries. Boulder, CO: Westview Press. Lessons on Implementing the Graduation Model.� Brief, Donovan, Graeme. 2006. “When Markets Do Not Work, Consultative Group to Assist the Poor, Washington, DC. should Grants Be Used?� Agriculture and Rural Evenson, Robert E., and Douglas Gollin. 2007. “Contributions Development Notes, Issue 9, World Bank, Washington of National Agricultural Research system to Crop DC. Productivity.� In Handbook of Agricultural Economics, Vol. Doran, Alan, and Jacob Levitsky. 1997. Credit Guarantee 3, Agricultural Development: Farmers, Farm Production, Schemes for Small Business Lending. London: Graham and Farm Markets, ed. Robert Evenson and Prabhu Bannock and Partners Ltd. Pingali, 2419–59. Amsterdam: Elsevier. Doran, Alan, Ntongi McFadyen, and Robert C. vogel. 2009. FAO (Food and Agriculture Organization of the United The Missing Middle in Agricultural Finance: Relieving Nations). 2008. “Initiative on soaring Food Prices.� the Capital Constraint on Smallholder Groups and Other Programme Document, FAO, Rome. Agricultural SMEs. Oxfam GB Research Report. http:// ———. 2009a. “Responding to the Food Crisis: synthesis www.oxfam.org.uk/resources/policy/trade/missing-middle- of Medium-Term Measures Proposed in Inter-Agency agricultural-finance.html. Assessments.� FAO, Rome. ———. 2009b. Declaration of the World summit on Food security. Rome, November 16–18. A j O I N T P U B L I C AT ION OF T HE WORL D B ANK, GLT N, FIG, A ND FA O 52 R EFER ENC ES ———. 2009c. “The state of Agricultural Commodity ———. 2003. “Lessons for Rural Finance from the Markets 2009: High Food Prices and the Food Crisis— Microfinance Revolution.� In Promising Practices in Experiences and Lessons Learned.� FAO, Rome. Rural Finance: Experiences from Latin America and the Caribbean, ed. Mark D. Wenner, Javier Alvaredo, and ———. 2010. “summary Note: Agricultural Investment Franciso Galarza, 53–65. Washington, DC: Inter-American Funds for Developing Countries.� FAO, Rome. Development Bank. FAO/GTZ (Food and Agriculture Organization of the Goodman, Patrick. 2009. “Raising MFI Equity through United Nations/Deutsche Gesellschaft für Technische Microfinance Investment Funds.� In New Partnerships Zusammenarbeit). 1998. Agricultural Finance Revisited: for Innovation in Microfinance, ed. Ingrid Matthaus-Maier Why? Agricultural Finance Revisited Monograph 1. and J. D. von Pischke, 19–47. Berlin: springer. Rome. Gudger, Michael. 1997. “The sustainability of Credit Feder, Gershon, Lawrence J. Lau, Justin Y. Lin, and Xiaopeng Guarantee systems.� Small Enterprise Development 8 Luo. 1990. “The Relationship between Credit and (2): 18–23. Productivity in Chinese Agriculture: A Microeconomic Model of Disequilibrium.� American Journal of Agricultural ———. 1998. Credit Guarantees: An Assessment of the Economics 72 (5 [proceedings issue]): 1151–57. State of Knowledge and New Avenues of Research. FAO Agricultural services Bulletin 129. Rome: Food and Fitch Ratings, 2009. “Agrocapital Rating Fitch 2008,� www. Agriculture Organization of the United Nations. mixmarket.org/mfi/agrocapital/files/medialibrary/ Guirkinger, Catherine, and steve Boucher. 2007. “Credit Fitchett, Delbert A. 1999. “Bank for Agriculture and Constraints and Productivity in Peruvian Agriculture.� Agricultural Cooperatives (BAAC), Thailand (Case study),� Working Paper No. 07-005, Department of Agricultural Working Group on savings Mobilization, Consultative and Resource Economics, University of California, Davis, Group to Assist the Poor, Washington, DC. pdf.usaid. CA. gov/pdf_docs/Pnacj084.pdf. Hartarska, valentina M., and Martin Holtman. 2006. “An Flaming, Mark. 2007. “Guaranteed Loans to Microfinance Overview of Recent Developments in the Microfinance Institutions: How Do They Add value?� CGAP Focus Literature.� Agricultural Finance Review 66 (2 [special Note No. 40, Consultative Group to Assist the Poor, Issue]): 147–65. Washington, DC. Hartell, Jason, and Jerry skees. 2009. “Pre-Feasibility Fleisig, Heywood, Mehnaz safavian, and Nuria De La Pena. Analysis: Index-Based Weather Risk Transfer in 2006. Reforming Collateral Laws to Expand Access to Mali.� Washington, DC: U.s. Agency for International Finance. Washington, DC: World Bank. Development. www.microlinks.org/.../Index-based+Wea Freedman, Paul L., and Reid W. Click. 2006. “Banks ther+Risk+Transfer+in+Mali+-+ 2009+Report.pdf. That Don’t Lend? Unlocking Credit to spur Growth in Hazell, P., J. Anderson, N. Balzer, A. Hastrup Clemmensen, Developing Countries.� Development Policy Review 24 U. Hess, and F. Rispoli. 2010. Potential for Scale and (3): 279–302. Sustainability in Weather Index Insurance for Agriculture Fries, Robert, and Manu Akin. 2004. “value Chains and and Rural Livelihoods. Rome: International Fund for Their significance for Addressing the Rural Finance Agricultural Development and World Food Programme. Challenge.� MicroReport #20, U.s. Agency for Hazell, Peter, Carlos Pomareda, and Alberto valdes, eds. 1986. International Development, Accelerated Microenterprise Crop Insurance for Agricultural Development: Issues and Advancement Project, Washington, DC. pdf.usaid.gov/ Experience. Baltimore: Johns Hopkins University Press. pdf_docs/PNADI913.pdf, Helms, Brigit. 2006. Access for All: Building Inclusive Gonzalez-vega, Claudio. 1984. “Credit-Rationing Behavior Financial Systems. Washington, DC: World Bank. of Agricultural Lenders: The Iron Law of Interest-Rate Restrictions.� In Undermining Rural Development with Helms, Brigit, and Xavier Reille. 2004. “Interest Rate Ceilings Cheap Credit, ed. Dale W. Adams, Douglas H. Graham, and Microfinance: The story so Far.� Occasional Paper and J. D. von Pischke, 78–95. Boulder, CO, and London: No. 9, Consultative Group to Assist the Poor, Washington, Westview Press. DC. S U B S I D I E S A S A N INST RUME NT IN AGRICULT URE FINA NCE: A REV IEW 53 Hess, Ulrich, and Peter Hazell. 2009. “sustainability and Kanathigoda, saliya, and Dirk steinwand. 2003. “The scalability of Index-Based Insurance for Agriculture and Challenge of sustainable Outreach: How Can Public Banks Rural Livelihoods.� In Innovations in Insuring the Poor, Contribute to Outreach in Rural Areas? An Overview.� ed. Ruth vargas Hill and Maximo Torero. Washington, Deutsche Gesellschaft für Technische Zusammenarbeit, DC: International Food Policy Research Institute. Eschborn, Germany. Hill, Ruth vargas, and Maximo Torero. 2009. “Innovations in Karlan, Dean, and Jonathan Morduch. 2010. “Access to Insuring the Poor: Overview.� In Innovations in Insuring Finance.� In Handbook of Development Economics, vol. the Poor, ed. Ruth vargas Hill and Maximo Torero. 5, 4703–84. Amsterdam: Elsevier. Washington, DC: International Food Policy Research Karlan, Dean s., and Jonathan Zinman. 2008. “Credit Institute. Elasticities in Less-Developed Economies: Implications Hirschland, Madeline, with Renée Chao-Béroff, Malcolm for Microfinance.� American Economic Review 98 (3): Harper, and Nanci Lee. 2008. “Financial services in 1040–68. Remote Rural Areas: Findings from seven Member- Khandker, s. R. 1998. Fighting Poverty with Microcredit: Owned Institutions.� Coady International Institute, Experience in Bangladesh. New York: Oxford University Antigonish, Nova scotia, Canada. Press. Hoff, Karla, and Joseph E. stiglitz. 1990. “Introduction: Kloeppinger-Todd, Renate, and Manohar sharma, eds. 2010. Imperfect Information and Rural Credit Markets-Puzzles Innovations in Rural and Agriculture Finance. Focus 18. and Policy Perspectives.� World Bank Economic Review Washington, DC: International Food Policy Research 4 (3): 235–50. Institute and World Bank. Hollinger, Frank. 2004. Financing Agricultural Term Kumar, Kabir, Claudia McKay, and sarah Rotman. 2010. Investments. Agricultural Finance Revisited No. 7. Rome: “Microfinance and Mobile Banking: The story so Far.� Food and Agriculture Organization of the United Nations/ Focus Note 62, Consultative Group to Assist the Poor, Deutsche Gesellschaft für Technische Zusammenarbeit. Washington, DC. Honohan, Patrick, and Thorsten Beck. 2007. Making Finance Ledgerwood, J. 1998. Microfinance Handbook: An Work for Africa. Washington, DC: World Bank. Institutional and Financial Perspective. Washington, DC: Hulme, David, and Paul Mosley. 1996. Finance against World Bank. Poverty, vols. 1 and 2. London: Routledge. Levine, Ross. 1997. “Financial Development and Economic IFAD (International Fund for Agricultural Development). Growth: views and Agenda.� Journal of Economic 2009a. Rural Finance Policy. Rome: IFAD, p. 14. Literature 35 (2): 688–726. ———. 2009b. “The Future of World Food security: Levitsky, Jacob. 1997. “Credit Guarantee schemes for sMEs: Investing in smallholder Agriculture—An International An International Review.� Small Enterprise Development Priority.� Rome: IFAD. 8 (2): 4–17. ———. 2010. IFAD Decision Tools for Rural Finance. Rome: Llanto, Gilberto M.,and Jocelyn Alma R. Badiola. N.d. IFAD. “The Impact of the Global Financial Crisis on Rural and Microfinance in Asia.� Unpublished paper, Philippine IFPRI/Terrafrica/GTZ (International Food Policy Research Institute for Development studies, Manila. Institute/Terrafrica/ Deutsche Gesellschaft für Technische Zusammenarbeit). 2009. The World Food Crisis, Land Llisterri, Juan Jose, and Jacob Levitsky. 1996. Sistemas Degradation, and Sustainable Land Management: de Garantias de Credito: Experiencias Internacionales y Linkages, Opportunities, and Constraints. Washington, Leccions Para America Latina y El Caribe. Washington, DC: IFPRI/Terrafrica/GTZ. DC: Inter-American Development Bank. Jacquier, Christian, Gabriele Ramm, Philippe Marcadent and valérie schmitt-Diabate. 2006. “The social Protection Perspective on Microinsurance.� In Protecting the Poor: A Microinsurance Compendium, ed. Craig Churchill, 45–64. Geneva: International Labour Office. A j O I N T P U B L I C AT ION OF T HE WORL D B ANK, GLT N, FIG, A ND FA O 54 R EFER ENC ES Lonie, susie. 2010. “M-PEsA: Finding New Ways to serve Microfinance Information eXchange (MIX). 2009. the Unbanked in Kenya.� Innovations in Rural and Microbanking Bulletin. December. Washington, DC. Agriculture Finance, ed. Renate Kloepping-Todd and MicroRate Incorporated. 2010. “state of Microfinance Manohar sharma. Brief 8, Focus 18. Washington, DC: Investment: The MicroRate 2010 MIv survey.� MicroRate International Food Policy Research Institute and World Incorporated, Arlington, virginia. Bank. Miller, Calvin, and Linda Jones. 2010. Agricultural Value Chain Magill, John H., and Richard L. Meyer. 2005. “Microenterprises Finance: Tools and Lessons. Rome and Warwickshire, and Microfinance in Ecuador: Results of the 2004 Baseline U.K.: Food and Agriculture Organization of the United study of Microenterprises.� U.s. Agency for International Nations and Practical Action Publishing. Development/Ecuador sALTO (strengthening Access to Microfinance and Economic Liberalization) Project and Miller, Calvin, sylvia Richter, Patrick McNellis, and Development Alternatives, Inc., Bethesda, MD. www. Nomathemba Mhlanga. 2010. Agricultural Investment ruralfinance.org/.../1129121833621_Microenterprises_ Funds for Developing Countries. Rome: Food and and_ microfinance_in_Ecuador.pdf Agriculture Organization of the United Nations. Mahul, O., and C. s. stutley. 2010. Government Support Minot, Nicholas, and Todd Benson. 2009. “Fertilizer subsidies to Agricultural Insurance: Challenges and Options for in Africa: Are vouchers the Answer?� IFPRI Issue Brief 60, Developing Countries. Washington, DC: World Bank. International Food Policy Research Institute, Washington, DC. McKay, Claudia. 2009. “In Malawi, Biometric ATMs Confront Traditional Ways of Moving Money.� December 3, Miranda, Mario J., and Claudio Gonzalez-vega. 2010. Consultative Group to Assist the Poor, Washington, DC. “systemic Risk, Index Insurance, and Optimal http://technology.cgap.org/2009/.../in-malawi-biometric- Management of Agricultural Loan Portfolios in Developing atms-confront-traditional-ways-of- moving-money/. Countries.� Paper presented at the annual meeting of the Allied social sciences Association, Atlanta, GA, January McKenzie, David J., and Christopher Woodruff. 2006. “Do 3–5. Forthcoming in the American Journal of Agricultural Entry Costs Provide an Empirical Basis for Poverty Economics. Traps?� Economic Development and Cultural Change 55 (1): 3–42. Morduch, Jonathan. 1995. “Income smoothing and Consumption smoothing.� Journal of Economic ———. 2008. “Experimental Evidence on Returns to Capital Perspectives 9 (3): 103–14. and Access to Finance in Mexico.� World Bank Economic Review 22 (3): 457– 82. ———. 1999a. “The Microfinance Promise.� Journal of Economic Literature 47 (4): 1569–1614. Meyer, Richard L., and Carlos E. Cuevas. 1992. “Reduction of Transaction Costs of Financial Intermediation: Theory ———. 1999b. “The Role of subsidies in Microfinance: and Innovations.� In Saving and Credit for Development, Evidence from the Grameen Bank.� Journal of report of the International Conference on savings and Development Economics 60 (1): 229–48. Credit for Development, Klarskovgard, Denmark, May Mustafa, M. R., M. Al-sayed Ali, and M. Awaideh. 2010. 28–31, 1990, 285–317. “study on Risk Management in Rural Finance in NENA Meyer, Richard L., and Geetha Nagarajan. 1996. “Credit Region.� Prepared for the Rural Finance Unit of the Guarantee schemes for Developing Countries: Theory, Food and Agriculture Organization of the United Nations, Design, and Evaluation.� Report prepared for the Center Amman, Jordan. for Economic Growth, U.s. Agency for International Mustafa, M. R., M. Al-sayed Ali, M. Awaideh and C. Miller. Development, Washington, DC. 2010. “study on Risk Management in Rural Finance in ———. 2000. Rural Financial Markets in Asia: Policies, NENA Region.� AGs Working Document No. 29, Near Paradigms, and Performance. New York: Oxford East and North Africa Rural and Agricultural Credit University Press. Association (NENARACA) and Food and Agriculture Organization of the United Nations, Rome. ———. 2006. “Microfinance in Developing Countries: Accomplishments, Debates, and Future Directions.� Agricultural Finance Review 66 (2 [special Issue]): 167–93. S U B S I D I E S A S A N INST RUME NT IN AGRICULT URE FINA NCE: A REV IEW 55 Nagarajan, Geetha, and Richard L. Meyer. 2005. “Rural Plyer, Megan G., sherri Haas, and Geetha Nagarajan. Finance: Recent Advances and Emerging Lessons, 2010. “Community-Level Economic Effects of M-PEsA Debates, and Opportunities.� Reformatted version of in Kenya: Initial Findings.� IRIs Center, University of Working Paper No. (AEDE-WP-0041-05), Department Maryland, College Park, MD. of Agricultural, Environmental, and Development PricewaterhouseCoopers. 2010. “AGRA, Impact Economics, Ohio state University, Columbus, OH. Investing Fund for African Agriculture, A Laboratory for www.osu.edu/programs/RuralFinance/publications2005.htm. Development, Enhancement Workshop.� PowerPoint Nair, Ajai. 2005. “sustainability of Microfinance self Help presentation given at IFAD, Rome. Groups in India.� Policy Research Working Paper 3516. Roberts, R. A. J. 2005. “Insurance of Crops in Developing World Bank, Washington, DC. Countries.� FAO Agricultural services Bulletin 159, Food ———. 2008. “Assessing and Managing Credit-risk in and Agriculture Organization of the United Nations, Rome. Agricultural Lending: Findings from a survey of Major Roberts, Richard, and Robert Ocaya, eds. 2010. Agricultural Lenders in Five Countries.� Unpublished draft paper, Finance Yearbook 2009. Kampala: Bank of Uganda and World Bank, Washington, DC. Plan for the Modernization of Agriculture. Onumah, Gideon E. 2003. “Improving Access to Rural Robinson, Marguerite. 2001, 2002. The Microfinance Finance through Regulated Warehouse Receipt systems Revolution, vols. 1 and 2. Washington, DC: World Bank in Africa.� Paper presented at “Paving the Way Forward and the Open society Institute. for Rural Finance: An International Conference on Best Practices,� U.s. Agency for International Development, Rosenberg, Richard. 2002. “Microcredit Interest Rates.� Washington, DC, June 2–4. CGAP Occasional Paper No. 1 (revised November 2002), Consultative Group to Assist the Poor, Washington, DC. Opportunity International. 2008 “Banking Rollout Approaches to Rural Markets, The Experience of Opportunity ———. 2007. “CGAP Reflections on the Compartamos Initial International Bank of Malawi,� White Paper No. 8, Opportunity Public Offering: A Case study on Microfinance Interest International, Oak Brook, IL. www.opportunity.net/Publications/ Rates and Profits.� Focus Note No. 42, Consultative Casestudies. Group to Assist the Poor, Washington, DC. ———. 2009. “Agrifinance: Reaching the Rest of Africa.� Rosenberg, Richard, Adrian Gonzalez, and sushma Narain. White Paper, Opportunity International, Oak Brook, IL. 2009. “The New Moneylenders: Are the Poor Being www.opportunity.net/Publications/Casestudies. Exploited by High Microcredit Interest Rates?� Occasional Paper No. 15, Consultative Group to Assist the Poor, Otero, Maria, and Elisabeth Rhyne, eds. 1994. The New Washington DC. World of Microenterprise Finance: Building Healthy Financial Institutions for the Poor. West Hartford, CT: Rudolph, Heinz P. 2010. “state Financial Institutions: Can Kumarian Press. They Be Relied on to Kick-start Lending?� Crisis Response Note Number 12, World Bank, Washington, DC. Overseas Development Institute. 2009. “The Global Financial Crisis: Poverty and social Protection: Evidence from 10 schreiner, Mark, and Jacob Yaron. 2001. Development Country Case studies.� Briefing Paper 51, Overseas Finance Institutions: Measuring Their Subsidy. Development Institute, London. Washington, DC: World Bank. Pingali, Prabhu, and Tim Kelley. 2007. “The Role of sEAF (small Enterprise Assistance Funds). 2007. “From International Agricultural Research in Contributing to Poverty to Prosperity: Understanding the Impact of Global Food security and Poverty Alleviation: The Case of Investing in small and Medium Enterprises.� sEAF, the CGIAR.� In Handbook of Agricultural Economics, vol. Washington, DC. 3, Agricultural Development: Farmers, Farm Production, seibel, Hans Dieter. 2000. “Agricultural Development Banks: and Farm Markets, ed. Robert Evenson and Prabhu Close Them or Reform Them?� Finance and Development Pingali, 2381–2418. Amsterdam: Elsevier. 37 (2): 45–48. A j O I N T P U B L I C AT ION OF T HE WORL D B ANK, GLT N, FIG, A ND FA O 56 R EFER ENC ES seibel, Hans Dieter, Thorsten Giehler, and stefan Karduck. Triodos Facet. 2008. “EFsE (European Fund for southeast 2005. Reforming Agricultural Development Banks. Europe): Development Impact study, 2007.� Zeist, Eschborn, Germany: Deutsche Gesellschaft für Netherlands. Technische Zusammenarbeit. Turvey, Calum G. 2010. “Risk, savings, and Farm Household siamwalla, Ammar, Chirmsak Pinthong, Nipon Poapongsakorn, Credit Demand Elasticities in Rural China.� Draft work- Ploenpit satsanguan, Prayong Nettayarak, Wanrak ing paper, Department of Applied Economics and Mingmaneenakin, and Yuavares Tubpun. 1990. “The Management, Cornell University, Ithaca, NY. Thai Rural Credit system: Public subsidies, Private Udry, Christopher, and santosh Anagol. 2006. “The Return Information, and segmented Markets.� World Bank to Capital in Ghana.� American Economic Review 96 (2): Economic Review 4 (3; A symposium Issue on Imperfect 388–93. Information and Rural Credit Markets): 271–95. UsAID (U.s. Agency for International Development)/Rural siamwalla, Ammar, and Alberto valdes. 1986. “should sPEED. 2006. “Note from the Field: Warehouse Receipts Crop Insurance Be subsidized?� In Crop Insurance for Turn Corn into Collateral.� Kampala, Uganda: UsAID. Agricultural Development: Issues and Experience, ed. Peter Hazell, Carlos Pomareda, and Alberto valdes, 117– ———. 2007. “Evaluation of DCA Programs and Impact: 125. Baltimore: Johns Hopkins University Press. 2002–2007.� Kampala, Uganda: UsAID. sinha, sanjay. 2008. “Mixing Business with social Change: ———. 2010. Fs Series #8: Rural and Agricultural Finance for Creating BAsIX.� Unpublished paper, Micro-Credit Food Security. Washington, DC. Ratings International Limited, Gurgaon, India. www.m- van der Meer, Kees, and Marijn Noordam. 2004. “The Use cril.com. of Grants to Address Market Failures: A Review of World skees, Jerry R. 2008. “Innovations in Index Insurance for Bank Rural Development Projects.� Agriculture and the Poor in Lower-Income Countries.� Agricultural and Rural Development Discussion Paper 27, World Bank, Resource Economics Review 37 (1): 1–15. Washington, DC. skees, Jerry, Anne Murphy, Benjamin Collier, Michael J. vogel, Robert C., and Dale W. Adams. 1997. “The Benefits McCord, and Jim Roth. 2007. “scaling Up Index Insurance: and Costs of Loan Guarantee Programs.� The Financier 4 What Is Needed for the Next Big step Forward?� Report (1 and 2): 22–29. prepared for Kreditanstalt für Wiederaufbau (KfW), von Pischke, J. D. 2008. “Gandhi vs Gauss: Ethical Issues German Financial Cooperation. in Micro and small Business Finance.� Savings and stiglitz, Joseph E. 1994. “The Role of the state in Financial Development 32 (3): 127–52. Markets.� In Proceedings of the World Bank Annual von Pischke, J. D., Gordon Donald, and Dale W. Adams, eds. Conference on Development Economics, 1993: 1983. Rural Financial Markets in Developing Countries: supplement to The World Bank Economic Review and Their Use and Abuse. Baltimore: Johns Hopkins University The World Bank Research Observer, 19–52. Washington, Press. DC: World Bank. Wenner, Mark, and Francisco J. Proenza. 2000. “Rural stutley, Charles. 2010. “Guidelines for the Design and Finance in Latin America and the Caribbean: Challenges Implementation of Agricultural Weather Index Insurance and Opportunities.� Working Paper, Inter-American schemes in Africa.� Eschborn, Germany: Deutsche Development Bank, Washington, DC. Gesellschaft für Technische Zusammenarbeit. Wenner, Mark D., Javier Alvaredo, and Franciso Galarza, eds. Teh, Tse-Ling, and Alan Martina. 2008. “Developing Countries 2003. Promising Practices in Rural Finance: Experiences spreading Covariant Risk into International Risk Markets: from Latin America and the Caribbean. Washington, DC: subsidised Catastrophe Assistance?� Working Paper No. Inter-American Development Bank. 492, Australian National University, Canberra. Wenner, Mark, sergio Navajas, Carolina Trivelli, and Alvaro Trivelli, Carolina, and Hildegardi venero. 2007. Banca de Tarazona. 2007. “Managing Credit Risk in Rural Financial Desarrollo pare el Agro: Experiencias en Curso en America Institutions in Latin America.� Washington, DC: Inter- Latina. Lima: Instituto de Estudios Peruanos. American Development Bank. S U B S I D I E S A S A N INST RUME NT IN AGRICULT URE FINA NCE: A REV IEW 57 World Bank. 1989. World Development Report 1989: Zeller, Manfred. 2006. “A Comparative Review of Major Financial Systems and Development. New York: Oxford Types of Rural Microfinance Institutions in Developing University Press. Countries.� Agricultural Finance Review 66 (2; special Issue): 195–213. ———. 2001. Finance for Growth: Policy Choices in a Volatile World. Washington, DC: World Bank. Zeller, Manfred, and Richard L. Meyer. 2002. The Triangle of Microfinance: Financial Sustainability, Outreach, and ———. 2003. “Rural Financial services: Implementing the Impact. Baltimore: Johns Hopkins University Press. Bank’s strategy to Reach the Rural Poor.� Report No. 26030, World Bank, Washington DC. Zeller, Manfred, Gertrud schrieder, Joachim von Braun, and Franz Heidhues. 1997. Rural Finance for Food Security ———. 2005a. “Guidance for Design of Microfinance Credit for the Poor: Implications for Research and Policy. Food Lines in social Funds and CDD Projects.� version 1.4, Policy Review 4, International Food Policy Research World Bank, Washington, DC. Institute, Washington, DC. ———. 2005b. Rural Finance Innovations: Topics and Case Studies. Report No. 32726-GLB, World Bank, Washington, DC. ———. 2006. “Meeting Development Challenges: Renewed Approaches to Rural Finance.� Agriculture and Rural Development Department, World Bank, Washington, DC. ———. 2007. World Development Report 2008: Agriculture for Development. World Bank, Washington, DC. ——. 2008. “Framework Document for Proposed Loans, Credits, and Grants in the Amount of Us$1.2 Billion Equivalent for a Global Food Crisis Response Program.� World Bank, Washington, DC. ———. 2009a. Awakening Africa’s Sleeping Giant: Prospects for Commercial Agriculture in the Guinea Savannah Zone and Beyond. Washington, DC: World Bank. ———. 2009b. “Protecting Progress: The Challenge Facing Low-Income Countries in the Global Recession.� Background paper prepared by World Bank Group staff for the G-20 Leaders’ Meeting, Pittsburgh, PA, september 24–25. Wright, Graham A. N. 2001. “Dropouts and Graduates: Lessons from Bangladesh.� Microbanking Bulletin (6): 14–16. Yaron, Jacob. 1994. “What Makes Rural Finance Institutions successful.� World Bank Research Observer 9 (1): 49–70. Yaron, Jacob, McDonald P. Benjamin, Jr., and Gerda L. Piprek. 1997. Rural Finance: Issues, Design, and Best Practices. Washington, DC: World Bank. Yunus, Muhammad. 2006. “Nobel Lecture.� Oslo: Nobel Foundation. A j O I N T P U B L I C AT ION OF T HE WORL D B ANK, GLT N, FIG, A ND FA O ARD AGRICULTURE AND RURAL DEVELOPMENT The World Bank Agriculture and Rural Development (ARD) 1818 H street, NW Washington, D.C. 20433 UsA Telephone: 202-477-1000 Internet: www.worldbank.org/ard This is a joint publication with: