POLICY RESEARCH WORKING PAPER 2129 Agricultural Extension The agriculture sector musL nearly double biological yields on existing farmiand to meet Generic Challenges and Some food needs, which will Ingredients for Solutions double in the next quarter century. A sustainable Gershon Feder approach to providing Anthony Willett agricultural extension services in developing countries- Willem Zijp minimal external Inputs, a systems orientation, pluralism, and arrangements that take advantage of the best incentives for farmers and extension service providers -- will release the local knowledge, resources, common sense, and organizing ability of rural people. The World Bank Development Research Group Rural Development and Rural Development Department May 1999 PoLic-Y REsEARCH WtORKI-NG PAPER 2129 Summary findings Is agricultural extension in developing countries up to Liability for public service functions beyond the the task of providing the information, ideas, and transfer of agricultural knowledge and information. organization needed to meet food needs? What role * Fiscal sustainability. should governments play in implementing or facilitating * Inadequate interaction with knowledge generators. extension services? Feder, Willett, and Zijp show how various extension Roughly 80 percent of the wvorld's extension is approaches were developed in attempts to overcome the publicly funded and delivered by civil servants, providing challenges of extension: a range of services to the farming population, - Improving extension management. commercial producers, and disadvantaged target groups. - Decentralizing. Budgetary constraints and concerns about performance ' Focusing on single commodities. create pressure to show the payoff on investment in * Providing fee-for-service public extension services. extension and to explore alternatives to publicly * Establishing institutional pluralism. providing it. * Empowering people by using participatory Feder, Willett, and Zijp analyze the challenges facing approaches. policymakers who must decide what role governments Using appropriate media. should play in implementing or facilitating extension Each of the approaches has weaknesses and strengths, services. Focusing on developing country experience, and in their analysis the authors identify the ingredients they identify generic challenges that make it difficult to that show promise. organize extension: Rural people know when something is relevant and * The magnitude of the task. effective. The aspects of agricultural extension services * Dependence on wider policy and other agency that tend to be inherently low cost and build reciprocal, functions. mutually trusting relationships are those most likely to Problems in identifying the cause and effect needed produce commitment, accountability, political support, to enable accountability and to get political support and fiscal sustainability, and the kinds of effective interaction funding. that generate knowledge. This paper - a joint product of Rural Development, Development Research Group, and the Rural Development Department-is part of a larger effort in the Bank to identify institutional and policy reforms needed to promote sustainable and equitable rural development. Copies of the paper are available free from the World Bank, 1818 H Street NW, Washington, DC 20433. Please contact Pauline Kokila, room MC3-544, telephone 202-473-3716, fax 202-522-1153, Internet address pkokilafa'worldbank.org. Policy Research Working Papers are also posted on the Web at http:// www.worldbank.org,'html/dec/Publications/Workpapers/home.html. The authors may be contacted at gfederf worildbank.org, awillettca'worldbank.org, or wzijpCa-worldbank.org. May 1999. (32 pages) The Policy Research 'V "orkiog Paper Series disseminates the findinigs of uwork in progress to encourage the exchange of ideas about developpment issues. An objecti,e of the series is to get the findings out qu.sickly, even if the presentations are less than fully polished. The papers carey the namves of the anthors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely tl7ose of the authors. They do not necessarily represenst the view of the World Bank, its Executive Directors, or the countries they represent. Produced by the Policy Research Dissemination Center Agricultural Extension Generic Challenges and the Ingredients for Solutions Gershon Feder, Anthony Willett, and Willem Zijpi 1. Gershon Feder is Research Manager, Rural Development, Development Research Group, the World Bank, 1818 H Street, N.W., Washington D.C. 20433. Tel. 202-473-0378; fax 202-522-1151; Intemet: gfeder@worldbank.org. Willem Zijp is Senior Agricultural Extension Specialist, Rural Development Department, The World Bank. Tel. 202- 473-2456; fax 202-522-3308; Internet: wzijp@worldbank.org. Anthony Willett is a consultant in agricultural education and extension in the Rural Development Department at the World Bank. Home tel. 703-761-9549; Internet: awillett@bellatlantic.net Abstract Agricultural extension faces great challenges. inherent in the nature of extension that make its The agriculture sector needs to nearly double organization difficult - the magnitude of the biological yields on existing farm land to meet task, dependence on wider policy and other gross food needs that will also double in the agency functions, problems in tracing cause and next quarter century. To help meet this effect, and consequently, difficulties in challenge, the role of extension is clear - there obtaining political support for funding, is a great need for information, ideas, and accountability, liability to public service organization. functions beyond agricultural knowledge and About 80 percent of the world's extension is information transfer, fiscal sustainability, and publicly funded and delivered by civil servants, interaction with knowledge generation. providing a diverse range of services to the Subsequently, we identify a range of general population, commercial producers, and innovations that have emerged in order to disadvantaged target groups through a variety of overcome the generic difficulties. These include approaches. Budgetary constraints and concerns improving extension management, about performance are pressuring these services decentralization, single commodity focus, fee- to show the payoff to investment in extension for-service public provision, institutional and explore alternatives to public provision. pluralism, empowerment and participatory This paper analyzes the key challenges facing approaches, privatization, and interconnecting policymakers who must decide what role rural people and the use of appropriate media. governments should play in implementing or Analyzing their varying success alone or in facilitating extension services, and to what combination, we note the ingredients of those extent. approaches that appear most promising to Our focus is on international, primarily increase the effectiveness of agricultural developing country, experience. The core of the extension. paper identifies a number of generic challenges I Extension - A Dilemma for Policymakers Challenge for agriculture and rural Challenge for extension development As the world grapples with these issues, Poverty, hunger, economic growth, food agricultural extension faces at least two production, and natural resource degradation are challenges: all great challenges in today's world. As the Information and organization in the global population climbs to an expected 8,000 agriculture sector must assume greater million by 2025, today at least 800 million importance. People involved in agriculture need people suffer from chronic hunger. Pervasive improved skills, information, and ideas in order poverty will remain largely rural, even as urban to develop an agriculture that will meet complex populations triple in the same time period. demand patterns, reduce poverty, and preserve Ensuring a thriving agricultural economy is or enhance ecological resources. Extension has critical for reducing poverty, enabling food an important role to play. security, and managing natural resources in a Extension funding and delivery face sustainable fashion. Agriculture provides a difficulties inherent in the extension mandate: livelihood for more than 60 percent of * magnitude of the task; developing country populations, and in many * dependence on wider policy and other countries, farm families make up 80 percent or agency functions; more of the population (World Bank, 1990). * problems establishing the cause and effect Agriculture has already reached the limits of necessary to obtain political and financial land and water, thus future increases in food support; production must exploit biological yields on * liability for public service functions beyond existing land (World Bank, 1997). In the face of agricultural knowledge and infonnation this technological challenge, agriculture faces a transfer; crisis in many parts of the developing world. In * fiscal sustainability; and Asia, the growth rate fostered by the green * interaction with knowledge generation. revolution has slowed. In Africa, per capita food Many observers are concerned that public production has declined in most years since extension is not doing enough, not doing it well, 1970 and is reflected in recurrent famine. In and is not always relevant. In developing many parts of Latin America and the Caribbean, countries, bureaucratic inefficiency and poor population pressure and extensive agriculture program design and implementation have led to seriously threaten the environment (World poor performance and incoherent links with Bank, 1990). And in the industrialized world, client farmers and the research sector. Support opposition to high input agriculture is mounting for extension declined in the 1980s and donors in response to such issues as animal rights, fears were unwilling to fund large-scale public-sector of genetically engineered products, and soil and recurrent expenditures, which led to further water pollution. underfinancing, staffing shortages, and contraction of extension services (Amanor and Farrington, 1991). As they seek solutions, policymakers must confront clashing views of what extension 1 should do, and choose among a number of regard as inherent in the nature of extension and extension priorities, products, mandates, and that make extension services difficult to finance models. Given fiscal restraint, there is extreme and deliver. In section IV, we review some of pressure to demonstrate the payoff to investment the most important institutional innovations that in extension and explore alternatives to public were induced by the challenges faced by financing by involving the private sector, local extension, and analyze them within the authorities, and producer groups. framework developed in section ElI. We note the Section II of this paper defines what we ingredients of those approaches that appear most mean by extension and describes the varying promising for the future of agricultural roles that governments have historically played. extension. Section IN identifies eight challenges that we 2 II Government Investments in Extension Defining extension management, and in negotiating financial, The term agricultural extension means different input, and market services. A system includes all public and private Anderson, 1997). Van den Ban and Hawkins institutions that transfer, mobilize, and educate Ande6) arson, 1997). Vaconcden Bn aend kions trural people, as distinct from a service or single (1996) arrive at a concept of extension that inttto tha trdtoal proide advic seems to synthesize diverse perspectives into only (Zijp, 1998). five goals - transferring knowledge from researchers to farmers; advising farmers in their decisionmaking; educating farmers to be able to Public investment in extension make similar decisions in future; enabling Worldwide, agricultural extension employs at farmers to clarify their own goals and least 800,000 extension workers and hundreds possibilities and to realize them; and stimulating of thousands more farmer technicians or leader desirable agricultural developments (rural farmers, reaching about 1,200 million people.3 guidance). They note that stimulating desirable Currently, about 80 percent of the world's agricultural development is the most common extension services are publicly funded and goal of extension directors. delivered by civil servants (World Bank, 1997). We have reviewed a number of efforts to Universities, parastatals, and nongovernment define or characterize extension2. The context organizations deliver about 12 percent of will dictate different rural development, services, and the private sector another 5 agricultural, and human resource development percent. priorities, and hence extension goals and Government involvement tends to be at functions. Our view is that it is helpful to see many levels and in many forms. It may fund, extension as both a system and the set of staff, or facilitate extension by establishing functions performed by that system to induce conducive regulations and policies for other voluntary change among rural people. providers - and it may pursue a range of A set of functions includes: purposes. Some of the generic problems of * transferring technology in multiple extension are caused in part by this complexity. directions for sustainable agricultural The traditional view of the 'public good' of production, transformation, and marketing; many aspects of agricultural knowledge * transferring management to mobilize and diffusion induced most governments to take organize farming, rural groups, and exclusive responsibility for extension delivery communities; and (Birkhaeuser et al., 1991; Umali-Deininger, * transferring capacity to educate, build 1996). Public funding has also been justified for human resources, and enhance local many extension programs by social goals such capacity, for example, in integrated pest as poverty alleviation and targeting specific management, market intelligence, farm groups, including rural women (Wilson, 1991). 2. Albrecht, 1986; Antholt, 1991; Birkhaeuser et al., 1991; Bunting, 1986; Evenson, 1986; Gustafson, 1991; Purcell and Anderson, 1997; Roling, 1986; Russell, 1986; Swanson, 1984; Umali and Schwartz, 1994; World Bank, 3. Based on Swanson et al., 1990, and taking into account 1990. recent Chinese data reported by Songlin, 1998. 3 As recently as 1990, the World Bank (1990) History perceived "no substitute (for a large nationwide Changes and challenges affecting extension are public extension system) for a corps of well- symptomatic of wider forces at work in society. trained and well-supported extension workers Public policy that affects extension tends to and subject matter specialists to serve as parallel development policy, and extension's competent, motivated, and trusted agents of institutional evolution reflects what is occurring change." in other institutional arenas (Rivera and Developing-country governments invested Gustafson, 1991). Major trends and shifts in heavily in agricultural extension, Beting extension praxis have been induced by changes increased agricultural production. Between 1959 in the economic, political, technological, and 1980, spending in real terms for extension sociocultural, and fiscal environments for grew more than six-fold in Latin America, socionv tripled in Asia, and more than doubled in Africa extension. (World Bank, 1990). In developing countries, the early colonial FAO4 surveyed the current status of emphasis on commodity programs, many of agricultural extension in 113 countries, which still exist today, gave way to broader contacting 207 agricultural extension multipurpose rural development efforts, often organizations that were considered to be set up by colonial powers that sent expatriate generally representative of agricultural 'rural agents' to organize communities and serve extension systems throughout the world as contact points for government authorities, (Swanson et al., 1990). Eighty-six percent of input and credit suppliers, and buying agents. thesewansonciesal.,re990undedghty-sixmpeis of Their broadly defined roles often extended into these agencies were funded by a ministry of human health, census taking, and tax collection. agriculture or similar government agency. Based Formation of nation states and state-led, planned on this survey, FAO estimated that in 1988, Fopmation of instationalization of $6,000 million were spent on public agricultural development, and the institutionalization of exte0nsiionagencie werldwide,ersntonpublicnagr ral many national extension services occurred in the extension agencies worldwide, representing an 1 950s. Because agricultural universities were average expenditure per extension worker of 1950s. Bexise agricultural extesion $8,522. Annual per farmer spending on bweak or nonexistent, agricultural extension extension services ranged from $2 to 3 in low- became attached to ministries of agriculture, a income ountrie to $65in highincomestructure that made weak links to research income countries to $65 in high-income (Axinn, 1988). Confidence in Western The rate of public investment in agricultural technology led to the 'diffusion model' of MOA The ate f pblicinvstmet inagrcultral extension deliver - a hierarchical, extension as a percentage of overall MOAs etnindlvry-hircial extenion a a pecentae of veral MOA5unidirectional process of technology transfer resources is directly related to the proportion of backed by advances in mass media. the labor force employed in agriculture in The by were in mass media. different countries. Where less than 20 percent The 1960s were the era of interpersonal of the population is employed in agriculture, communication and community development, extension receives about 2 percent of ministry and also the beginning of the green revolution. extesionrecevesabou 2 prcet ofminitryThe technolog transfer orientation was still resources, compared with slightly more than 20 Tetcnlgy tase retto a tl percentof resources where more than 60 percent strong. From the mid-1970s, public sector percenthe pources in more. extension again limited advice to technical agricultural matters, mainly major annual food crops, and the multipurpose agents began to be replaced by systems that focused more closely on extension and its management (World Bank, 1990). Fifty percent of extension agencies in FAO' s 1989 survey became organized or 4. Food and Agriculture Organization of the United roaz from 1970 on Nations. reorgamzed from 1970 onward (Swanson et al., 5. MOA is used throughout this text to denote government 1990). The main features of the 1970s included Ministries of Agriculture. 4 integrated rural development approaches, and 33 percent with an intermediate diploma or the rise of the training and visit (T&V) certificate. extension system. The 'diffusion' model of extension gave way to the 'get the technology Purposes of public extension investments right' model, where farm-level constraints FAO reported that around 58 percent of explained non-adoption of technology, with a extension resources were directed toward prescription to ease the constraints through commercial farmers, including specialized integrated packages of services (Axinn, 1988). The transformative decade of the 1980s producers of cash and export commodities. The brouhetransincrmative demphasi of participatory vast majority of smaller, marginal farmers in the brought an increasing emphasis on participatory world receive slightly more than one-third of all approaches. Concemns with increasing the extension resources. productivity of women and preserving Not all extension is directly related to ecosystems were added, along with attempted agricultural knowledge transfer. FAO found that cost recovery and privatization schemes (World extension personnel in developing countries Bank, 1990). Now, the 1990s is the era of spension onner of thelopime ontries alternatives where new approaches are being spend about one-quarter of their time on non- piloted in an environment of fiscal stringency. educational activities, representing "about Democratization has resulted in a dramatic fall 14,0 fultm eqiaetyasovxeso in pulcr sectio p r.esthdoicalirct personnel time." FAO regarded this as a "major in public sector power. Methodologically, direct loss of educational resources, especially in farm-level links are stressed between lossno conaresoure especial in researchers and farmers. More sustainable developing countries where extension coverage approaches to extension funding involve greater is still grossly inadequate, both in quantitative flexibility and multiple partners (Gustafson, and qualitative terms."n 1 991). In commodity programs, extension workers 1991- nmay spend considerable time helping farmers to Extension staffing6 acquire inputs, credit, and marketing services. In contrast, general government extension staff About 95 percent of extension staff work in may be engaged in a variety of local government public agricultural extension systems (Umali servicing activities, as well as responding to and Schwartz, 1994), and 90 percent of information requests, such as crop forecasting extension workers in the world are located in and census taking, from extension or ministry developing countries, over 70 percent in Asia headquarters. alone. Roling (1986) interprets the range of Extension coverage (the ratio of extension extension functions in terms of two traditions in personnel to farmer population) by public extension - technical innovation (TI), and extension services in developing countries human resource development (HRD). From his varies from 1:1,800 to 1:3,000. Developed perspective, most of the world's extension countries of Europe, North America, and Asia agencies are engaged in pure TI financed by tax have ratios averaging about 1:400. revenues to make the production of food, raw Although staff numbers are high in many materials, and export commodities as effective developing countries, staff quality is often low. and efficient as possible. The focus of HRD is Fiscal constraints face managers of public on rural people themselves and on the social extension systems, forcing them to hire staff systems in which they function, and deals with with few skills to cut operational costs. FAO such processes as community and leadership found 40 percent of extension personnel had development, building institutions, and farmer only a secondary school education, and another mobilization and organization. 6. Data cited are from the FAO study by Swanson et al. (1990), unless stated otherwise. 5 6 III Generic Challenges Make Extension Difficult The generic problems of agricultural extension In Sections I and II, we mentioned the are bound to its diverse functions, as well as the 800,000 extension agents and their 1,200 bureaucratic, political, and social operating million clients, many of whom are poor and environments within which extension systems have limited resources. The success of extension operate. We believe that focusing on these depends on the individual farm management generic problems - regardless of the decisions of these millions of people. Many management system or approach to extension- populations are widely dispersed and hard to highlights the areas that should form the agenda reach (Chambers, 1983). In developing for future directions in extension. countries, clients generally have a low level of literacy and formal education, and live far from Eight generic problems information sources. They have specific needs depending on their natural habitat, culture, exteresar genericns- problemspernts in thefarming/production system, and gender (Zijp, extension functions -many aspects of the 1998). services performed by extension are public t i e t goods or toll goods, require collective action, famer itu information addvce, th an ofe inovroemn. edfn ih farmers' agricultural information and advice, the aendeofte invoblvemgovernment.sWe:defineeight multiple stakeholders and partners in the generic problems: agricultural development effort, and the range in * scale and complexity; extension mandate. The variety of * dependence of extension on the wider policy communication forms - such as individual environment and other agency functions; farm visits, farmer-to-farmer extension, use of * inability to trace cause and effect; mass media, and most recently, computerized * commitment and political support; information technology - adds to the * accountability; complexity facing extension decisionmakers. * liability to public service functions beyond The main manifestation of the magnitude agricultural knowledge and information problem is coverage. Many public services are transfer; reaching only 10 percent of potential clientele, a * operating resources and fiscal sustainability; minority of which are women. Magnitude and interacts with quality of governance and * interaction with knowledge generation. adequate operating resources and technology to determine coverage and focus (Nagel, 1997). Scale and complexity The most common response to the coverage problem has been to establish relatively large, These issues are at the top of the list because in hierarchical, centralized, public-sector most respects they represent a 'given' situation. structures, with large numbers of extension The scale and complexity of the extension task agents widely dispersed in the rural areas, where can be understood in terms of numbers, their work is not easily observed or checked. distribution, and diversity of staff, farmers, and The top-down managerial style characteristic of other clients and stakeholders, and in terms of large bureaucracies tends not to be amenable to mandate and methodology. participatory, bottom-up approaches, and the 7 many layers in the hierarchy remove considered adequately, compromising the decisionmakers from the field action. usefulness of extension investment. This exacerbates the often obscured tendency of extension agents to favor more Inability to trace cause and effect responsive clients who are typically better The difficulty of tracing the relationship endowed and more capable of undertaking risks between extension input and its impact is (Axinn, 1988). Such biases were observed in another generic issue faced by extension. This countries as diverse as the Netherlands Estonia, difficulty leads to several other inherent and Tanzania (Zijp, 1988; Chambers, 1983). An problems, including political support, budget inherent tension exists between creating a strict provision, and accountability. hierarchy capable of managing such a large Evising accon'sbiity. system and being accountable to all strata of ealuaing extionshipact invol farmers. ~~~~~~~~~measuring the relationship between extension fanmers. activity and changes in: Depend eof extension on the wderpolicy * farmer awareness, knowledge, and adoption Deendiroenceand other agency functions of particular technologies or practices; and * farm productivity and efficiency, and An inherent problem of agricultural extension, profitability, input demand, and output especially in low-income countries, is that it has supply. to be combined with other policy instruments to These same indicators are also influenced achieve agricultural development (Van den Ban, by many other factors that have confounding 1986). Thus the effectiveness of extension effects. In their review of World Bank investment is highly contingent on relaxing agricultural research and extension projects, wider barriers to the successful development of Purcell and Anderson (1997) observed serious the agricultural sector as a whole, including data constraints and inability to include all the such potentially limiting factors as credit, contributing variables affecting production technology stock, input supplies, price outcomes. Numerous factors contribute to a incentives, institutions, and human resource specific production response and to farmers' constraints (Purcell and Anderson, 1997). decisions about the use of available resources. Specifically, institutional frameworks and Sophisticated econometric studies are agricultural policies may discriminate against needed to glean insights into these relationships. the rural sector, underinvest in technology Feder and Slade (1986) looked at the problems development and maintain inappropriate of observing and measuring changes in agrarian structures, lock-up arable land in low- operational efficiency of extension agents, farm productivity ranching, undervalue and waste husbandry knowledge, and agricultural natural resources, underinvest in health and productivity induced by more intensive systems education of the rural population, discriminate introduced by extension. They noted insufficient against private sector initiatives in food time-series data to cover the post-project marketing, and fail to maintain existing or invest situation, and an inability to compare subject in new rural infrastructure (World Bank, 1997). and control areas. Lack of access to resources and the Birkhaeuser et al. (1991) note that an ideal inefficient operation of complementary simulated experimental framework (before/after agricultural services thus limit the impact of and with/without an intervention) is rarely extension on production. Coordination and links available. Second-best approaches involve with complementary agricultural services are various biases depending on the level of key problems for extension organizations, analysis. Farm-level studies are vulnerable to especially the links with research, input supply problems of self-selection and the prevalence of systems, credit, and marketing organizations inter-farmer communication. Aggregate effects (Axinn, 1988). Often this dependence is not of extension measured in a region are subject to 8 estimation problems that relate to confounding providing required services and appropriate factors. information; Axinn (1988) summarizes the * accountability of dispersed, relatively methodological challenge by noting how much unsupervised field staff to supervisors; and easier and cheaper it is to monitor inputs than * public sector staff accountability to farmers. outputs. When extension indicators are more We are primarily interested in the second and sophisticated and higher level, the cost of third aspects. collecting information is also higher, and it is The third accountability problem arises more difficult to prove causality between the especially in the public sector environment of a selected extension activity and changes in farm 'top-down', supply-driven extension hierarchy income or welfare. The inherent problem, in which agents feel accountable to their however, is that this is precisely the information ministry supervisors rather than to farmers. needed to improve extension effectiveness and However, because even their supervisors cannot establish its justification. easily monitor and evaluate their performance (because of the difficulty of relating cause and Commitment and political support effect), agents become accountable to no one. Lack of commitment by senior government On the contrary, rent-seeking opportunities officials has been cited as a factor adversely provide a counter incentive for extension agents affecting implementation and funding support in to focus on non-extension tasks with more easily nearly half of World Bank-assisted free-standing observable results, such as credit and input extension projects (Purcell and Anderson, invoicing. 1997). Government failure to allocate necessary Purcell and Anderson (1997) found funds to run extension systems is one key evidence of accountability problems in many indication of such lack of commitment. Umali- World Bank extension projects. There was little Deininger (1996) records how inadequate attention to the farming community's systematic recurrent funding inhibits the field operation of participation in problem definition, problem World Bank-supported extension projects, solving, and extension programming. Staff affecting 87 percent of projects rated as quality and attitudes were major constraints for satisfactory and 100 percent of those rated both farmer contacts and technical support. The unsatisfactory. World Bank's 1994 ex-post evaluation of Roling (1986) relates fiscal problems to extension projects noted 'entrenched top-down bargaining power and terms of trade for the attitudes' in 48 percent of satisfactory projects agriculture sector. Lack of commitment to and 75 percent of those ranked unsatisfactory. agriculture and extension is often fed by urban Accountability is not just an issue in bias and poor understanding of rural information developing countries. Hercus (1991) found that (Zijp, 1998). Another possible explanation is agricultural extension services in New Zealand that there is no immediate payoff to politicians accounted to government for money spent in and policymakers for support and commitment terms of "activities not results, and ... almost to extension, because the attribution of exclusively [in terms of] expenditure and hardly production results to extension is so at all with outputs or efficiencies. The mandate problematic. of extension was derived by the agricultural extension service itself, and in the absence of Accountability any challenge or alternative definition by the taxpayers' representatives, the service regarded Howell (1986) identifies three aspects of its charter as the right to exist on the prevailing accountability: terms and conditions." * extension performance in terms of its effectiveness, impact, or benefit/cost ratio in 9 Liability to public service functions beyond 'big' farmers dominate farmer groups. Even if agricultural knowledge and information separate agencies are organized, extension must transfer still coordinate. The extension service is often the most widely- distributed representative of government at the Operating resources and fiscal sustainability grass roots level in the rural sector, therefore, Purcell and Anderson (1997) cite inadequate there is always the temptation to load it with public funding to operate services properly as a more and more functions. Historically, public common phenomenon in World Bank-assisted extension has been entrusted with various public free-standing extension projects, with 76 percent functions, including collecting statistics, of projects having an uncertain or unlikely conducting surveys, writing reports, erosion sustainability rating. According to Antholt control, and various regulatory functions such as (1994), training and visit (T&V) extension production quotas or pesticide usage. Feder and systems actually exacerbate operational funding Slade (1993) note that in many countries the problems by increasing staffing, and typically agricultural field service was given a range of cost 25-40 percent more than multipurpose additional functions as governments increased extension (Feder and Slade, 1993). their role in the rural economy. In rural The generic problem for extension is the extension, dissemination of agricultural inherent difficulty of cost recovery. Much knowledge is one part of wider government information disseminated by extension is a involvement in changing rural attitudes and 'public good', and dissemination costs cannot be promoting community self-reliance (Oakley, easily recovered from individuals, thus there is a 1997). The village-level worker is considered by dependence on direct public funding. Lack of the government as a relatively low-cost, flexible political support and commitment arises, as well administrative instrument to help disadvantaged as confusion over the 'model controversy' groups with multifunctional services, gaining (Rivera, 1991) and the role of the state - is it wider acceptance in the process. implementer, organizer, financier, or controller? We see both 'push' and 'pull' factors Part of the problem is due to magnitude - a involved in this encroachment on extension large extension service to serve large numbers services. The 'push' is the temptation for other of farmers with a large staff that is inherently agencies to use extension because it is the most expensive to operate. Fixed costs for salaries are widely distributed apparatus for contact with high, and operating costs are then treated as a rural communities. The 'pull' is that agents are residual, which makes them vulnerable in a willing to take on other duties, especially input budget shortage. If a government has to cut the distribution, because such tasks often increase operating budget, there are obvious an extension agent's influence over farmers, as consequences for effectiveness if extension well as providing opportunities to extract rents agents cannot get to the field, causing which compensate for low salary. The inability operational as well as morale problems (Axinn, to relate extension cause and effect makes it 1988). feasible for field staff to do this. Howell (1985) sees a cyclical pattern in Because time spent on noneducational tasks which extension agencies hired more staff when reduces potential educational impact by as much funds were available, but when budgets as 25 percent (Swanson et al., 1990), the declined, operating costs were cut. Ameur assignment of input supply duties to extension (1994) also sees the problem as a vicious circle workers tends to be discouraged. Van den Ban of fiscal difficulty, curtailed services, inefficient (1986) notes, however, that alternatives to operation, poorer results, and reduced staff delivery of these services by extension - such motivation, training, and competence. as by the private sector or farmer organizations - may be problematic if demand is limited or 10 Interaction with knowledge generation scientific papers, and therefore do not generate Many would argue that the 'bottom line' of research relevant to farmers. Extension agents extension is the quality of its message. Yet may be more rewarded by distributing inputs insufficient relevant or new technology and credit. necessary to improve productivity is one of the No pressure. Lack of effectively organized most common constraints in extension, and a outside pressure groups, such as national major constraint in rainfed, resource-poor policymakers, donors, fanner organizations, or environments (Axinn, 1988; Anderson and private companies, may hamper cooperation. Purcell, 1997). The World Bank's 1994 ex-post evaluation of extension projects found inadequate research-extension links to adversely Interactions affect a large proportion of the projects reviewed, and insufficient available technology The generic problems we have identified cannot to be an even more common problem. be seen in isolation from one another. They are The inherent problem is that extension interrelated and often represent conflicting disseminates information and advice generated imperatives. Pressures may exist to maximize by a knowledge-generating system which is coverage while minimizing costs, to emphasize generally not under extension management. This firm management control while ensuring leads to a crucial dependence, which itself is not bottom-up participation, and to increase human necessarily a problem. However, research and resource competence while exploiting the cost- extension often tend to compete for power and saving potential of mass media (World Bank, resources, and fail to see themselves as part of a 1990). broader agricultural technology system. There is We have already emphasized some also a tendency for both extension and research important causal relationships, in particular the to look for solutions within national borders, influence of scale and complexity on size and whereas relevant information and technology top-down management of national extension might be readily available across such artificial organizations, and the consequences for political boundaries. support and the inability to trace cause and Kaimovitz (1991) identifies the following effect. Although poorly trained and motivated obstacles to effective research-extension links: personnel; lack of qualified, competent, and Historical perceptions. Policymakers still trusted staff; and staff quality and low morale fail to recognize research and extension as have been described (Antholt, 1994); Axinn, closely interdependent activities. Extension's 1988; Nagel, 1997), we view such problems as credibility as a research partner remains symptomatic of the generic problems that have damaged by the historical experience. The been identified. generally higher status of researchers tends The interactions can also be related to the toward patronizing behavior that is resented by vicious circle of fiscal difficulty, curtailed extension agents. Both research and extension services, and inefficient operation noted by assign a lower priority to linking activities than Ameur (1994). Kaimovitz (1991) records how to core activities in allocation of time, resources, low salaries, limited operating resources, and an and management attention. unclear mandate led to a general decline in Coordination. Resistance to coordination is morale, and given difficult supervision, lower perceived as limiting autonomy by both sides. quality extension work, higher staff turnover, Goals may differ. The two organizations and lower credibility. These problems make it may not share the same goals. The official more difficult for extension to obtain resources. mandate may be to provide agricultural This crisis in extension directly affected technology to farmers, but in practice, relations with researchers who, perceiving researchers may be more interested in producing 11 extension as ineffective, became reluctant to rewarded, there is little incentive to work cooperate. harder to reach more, especially resource- Other significant interactions to note include poor farmers. Field staff may instead focus those between: on larger farmers and high input technology. 'Top-down' olntation and (because of * Accountability and sustainability. Without limited farmr feedback) potential farmer participation, an extension service irrelevance of technology generation. If gains only limited ideas about how willing biased, it may also distort coverage and users would be to contribute (Ameur, 1994). focus, whereby less powerful target groups * Lack of operating resources and fiscal and their crops, and more marginal, sustainability (as the independent variable), resource-poor areas are neglected (Axinn, and coverage, relevance, responsiveness, 1988). staff morale, and hence accountability, and * Inability to trace cause and effect and (by undermining extension credibility in the coverage. If extra work is not observed or eyes of research) knowledge generation. 12 IV Overcoming Generic Problems Experience and Promise A range of institutional innovations emerged as * interconnecting rural people and the use of policymakers have confronted the generic appropriate media. problems identified in the previous section. Figure 1 summarizes our analysis of how the Over time, however, the emphasis of these innovations address the generic problems. innovations has changed. Earlier, the emphasis was more on communication, then farming Improving extension management systems and limiting factors. Later, the emphasis The importance of improving extension shifted toward improving the organization and Thement is recogni eltestO r . . 1 @ * ~~~~management iS recognized in the latest FAG management of existing public extension reference manual on extension (FAO, 1997),. More recently, attempts to resolve the The training and visit (T&V) system of problem of fiscal sustainability have come into extension is used as an example of improving focus. Many innovations make use of the notion management because it is a well-documented, that not all collective action needs to be widely-implemented field model of an attempted thatnorganiz orleecutedibygoveraconmnensto aensystematic and comprehensive improvement in organized or executed by government agencies. h raiainadmngmn fpbi Thus, collective action at the community level the organization and management of public Thus,mn datgstatoecm eei extension. Other formulations of management has many advantages that overcome generic principles are plentiful, but their implementation problems. Furthermore, not all aspects of has been more piecemeal and less well extension's role are pure public goods, and thus documented. some innovations focus on separating elements Starting in the late 1960s, T&V has been that can be privatized and subjected to the implemented or tried in national systems in at discipline of the market, thus overcoming the least 76 countries in all major regions of the incentive problems inherent in public service world (Umali and Schwartz, 1994, citing FAO, delivery. 1990). It was the system employed in 90 percent of World Bank agricultural extension projects reviewed by Purcell and Anderson in 1997. We selected eight innovations and attempted Kaimovitz (1991) interprets T&V as a response modifications for our analysis: to an historical crisis in effectiveness, credibility, and * improving extension management; morale - all results of the generic problems -that * decentralization; public extension suffered in the 1960s and 1970s. To * single commodity focus; be effective, the designers of T&V stressed that * fee-for-service public provision; certain key features had to be preserved - . institutional pluralism (mobilizing other professionalism, a single line of command, players); concentration of effort, time-bound work, field and • empowerment and participatory approaches; farmer orientation, regular and continuous training, and close links with research (Benor et al., 1984). * privatization; and Over the years, many adaptations have evolved, 13 however, the reference here is to the original institutional innovation. design, which we view as an induced Figure 1. Matrix of generic problems and innovations. A l+] indicates a positive effect, [0] indicates little or no effect, and [-] indicates a negative consequence. Areas in the matrix left blank indicate inadequate data reported in the literature we studied from which to draw conclusions. Innovations and/or attempted modifications, and key ingredients Improving extension management Decentralization Single commodity focus Single line of command; exclusive Varying degrees of deconcentration Focus on a single crop, commodity, devotion to extension work; strict and/or devolution of extension or input; vertical integration of scheduling; regular, continuous program and funding decisions and extension and complementary training; field & farmer orientation; staff accountability to local units functions throughout production- links with research marketing system; public, private, or social sector organization Generic problems Scale and [+-] larger organization, more [+] more municipalities and small- [+] focus on a selected commodity complexity, hence staff, higher staff/farmer ratio; but scale farmers reached; simplified simplifies extension task coverage and focus low HRD skill in selecting contact management other crops, areas neglected farmers [-] rainfed, marginal areas Dependence on [0] [+] better potential for integration with [+] through vertical integration of wider policy and other local initiatives complementary services other agency functions Ability to trace [0] [0] [+] through vertical integration cause and effect Commitment and [0] frequently cited problem in [+] if local government democratic, but [+1 powerful interests and lobbies political support World Bank project evaluations local interference more of a problem Accountability [+] tight, hierarchical supervision [+] responsiveness may improve if [+] ease of structuring staff improved accountability within the local system democratic incentives service [-] if commodity organization [-] inflexible & non-participatory in interests diverge from farmers' or relation to clients society's interests Liability to other [+] increased focus on [-] Local govt. may be tempted to use [+] narrow commodity concern public service dissemination function extension agents for other rural functions programs Operating [-] significantly higher costs of [+-]lower cost per beneficiary but [+] through levies and other cost- resources and increased staffing, supervisory increased staff numbers and gross recovery mechanisms fiscal sustainability visits, & training cost [-] uncertain feasibility of local government resource mobilization Interaction with [-] In practice coordination still a [-] research-extension links may be [+] for the single crop knowledge problem more difficult for rest of the farming system generation Other aspects Quality control more difficult Involving private and social sector (farmer organizations) may improve 14 accountability Modification efforts and key ingredients Empowerment, farmer organizing, Fee-for-service public provision Institutional pluralism: mobilizing and participatory approaches other players User charge/fee-based Unlinked funding and delivery; Varying participation forms, from commercialization of public decentralization; client stratification; passive to self-mobilized farmer-to- extension service; farmer public-private partnership & cost- farmer extension and organization; stratification; often transition stage sharing; contracting out to private & decentralization; range of toward privatization. nonprofit sectors; AKIS; enabling participatory appraisal MOA role methodologies; extension agent training in HRD/ group interaction methods Generic problems Scale and [-] tendency for general [+] if mobilizes NGO field presence [+] farmer-to-farmer 'networking complexity participation in extension to drop, effect' effective in reaching small- and poorer, less market-oriented scale farmers farmers to be excluded [+] if accompanied by stratification and "extension safety net" features Dependence on wider policy and other agency functions Ability to trace [+] if contracting process involves [+] farmers are the closest to the real cause and effect design of indicators with farmer situation involvement, and if mobilizes NGO participatory skills Commitment and [-] may be opposition to [+-] farmer organizations can political support reclassifying extension from free mobilize support; centralized to purchased good governments may fear loss of control Accountability [+] stronger client orientation and [+] if mobilizes NGO responsiveness [+] farmer ownership and heightened more professional relationship; and HRD skills; and through provider trust and camraderie of participatory staff commissions and job competition, client orientation, and methods; farmers may hire or satisfaction stakeholder involvement themselves serve as the field agents Liability to other [+] Rationalizes public sector role [+] farmer control ensures service public service within broader range of institutions delivery functions Operating [+] increased efficiency reduces [+] Private and voluntary organizations [+] Group approaches less staff resources and overheads; income from fees contribute resources and roles; may intensive; farmer organizations fiscal sustainability charged; although may be high involve beneficiary co-payment mobilize local resources; sustainable administrative costs of collecting [-] costs associated with initial agriculture paradigm is lower cost charges increased complexity Interaction with [-] less interaction and reduced [+] if mobilizes NGO and nonprofit [-] May still be a problem knowledge feedback in the broader AKIS diagnostic & feedback skills generation Other aspects Positive outcomes depend largely on Outcomes less positive in more partnership arrangements capitalizing 'passive' forms of participation on partner strengths 15 Modification efforts and key ingredients Interconnecting rural people, and Privatization use of appropriate media Transfer of ownership of extension To enhance other innovations, various service to private entity; non-public information/ communication good-type information; support for technologies and applications, including infant advisory industries; public mass media; documentation and relations & media support; policy exchange of local knowledge. reform; stakeholder participation Generic problems Scale and complexity [0] [+] if supports participation, farmer organization, and efforts to reach target groups [-] if reinforces existing biases against coverage Dependence on wider policy [0] [+] if helps strengthen farmer and other agency functions organization negotiating ability, or reinforce policy reform efforts Ability to trace cause and [+] potentially, feedback through [-] Feedback mechanism more difficult effect improved private sector relationship with client farmers; concern for quality, efficiency; imperative to provide data to justify reforms Commitment and political [0] initial resistance to reforms from [0] support vested interests but, once established, little dependence on public budget Accountability [+] priority to farmer-customer [0] interests in long-term relationship; competition; farmer representation on board Liability to other public [+] by definition [+-] Depends on how used. service functions Operating resources and [+] cost efficiencies; mobilization of [+] May economize on cost of fiscal sustainability private and community resources to accessing information; mass media complement public funds more cost-effective than face-to-face; self-financing in the case of telecottages. [-] most IT applications involve significant initial investment and operating costs Interaction with knowledge [+] Private entities have incentives to [+] If developed jointly with knowledge generation link with knowledge generating generating agencies agencies Other aspects Outcome highly dependent on appropriateness of use in conjunction with other efforts. 16 T&V attempts to tackle several of the There remained a tendency to neglect generic problems identified in section HI. Using participatory aspects, and accountability to T&V as a reform tool for the problem of scale, farmers was not established (Axinn, 1988; many governments have covered most regions Nagel, 1997). T&V's most obvious of their countries with denser extension disadvantage has been its "highly questionable agent/farmer ratios by focusing on contact sustainability" (Antholt, 1994) as a nationwide farmers who are expected to pass information on system due to the substantially increased staff to fellow farners with similar problems (Nagel, allocations and recurrent and operational costs 1997). The attempt to cover many farmers, (Axinn, 1988). however, led to more dependence on annual public budget allocations for recurrent costs. Decentralization While dependence on external factors and Decentralization has been described as "the first other agency functions was not eliminated, the step on the long road to privatization" because design called for village extension workers diversity becomes more tangible and different (VEWs) to advise farmers on prices, availabilitydiestbcosmreanbladdfeet of t neesaryvinptse fands marietonditions, avasl y approaches to extension can be explored as the of necessary inputs, and market conditions, as local level becomes accessible (Ameur, 1994). well as report on actual availability and farmer Decentralization includes administrative and responses to supervisors. This would have political-fiscal devolution of program and allowed for adjustments to extension funding decisions and staff accountability to instructions. Accountability was to be provided local units. Its impact depends on the extent of through the tight, line-of-command supervision political and societal democratization at the system and the strict timetable of contact farmer local level. Specifically, the major factors group visits. Exclusive devotion to information affecting the effectiveness of decentralization dissemination tasks relieved staff from the 6push' and 'pull' to do tasks that are not related are: to agriculture or extension. Finally, T&V * the existence of an elected, representative designers attempted to resolve the problem of local govemment and a central govemment interaction with technology generation by willing to actually decentralize; structuring research-extension links involving * the ability of local government to raise regular training and continuous feedback of revenues; and farmers' problems. a MOA capacity for efficient quality control In practice, T&V could not escape some and monitoring (Garfield et al., 1996). generic problems, and aggravated a couple of If these requirements are met, various actions them. The contact farmer coverage method may be taken and functions decentralized, such frequently faced problems because contact as building local capacity for farmer farmers were not representative of the farming involvement in extension programming, housing community (Nagel, 1997). Strict scheduling extension agents locally and making them enabled closer checking on what field staff were responsible to farmers' associations, and doing and improved monitoring, but it has not designing resource mobilization and funding resolved the problem of relating input to impact. mechanisms. The dependence on other rural development If implemented effectively, and if local factors could not be eliminated, and the government is reasonably well funded, independent status that the extension system decentralization can transform the top-down received under T&V initiatives did not help to structure and operation of a public service resolve coordination problems with other bureaucracy, and positively affect several of the programs. generic problems of extension. The scale and In spite of some successes, there were quite complexity problem is reduced in proportion to a few countries where T&V has not generated the number of local government units that take sustained political support and commitment. on extension functions in a country, and how 'local' they are. Extension's dependence on 17 other agency functions is reduced because of the and total costs roughly in proportion to the potential for better interaction with other local increase in coverage. initiatives. Commitment and political support, Overall, staff numbers have increased by a along with responsiveness (an aspect of factor of 3.5, more than doubling the total cost. accountability) are enhanced if the local According to Garfield et al. (1996), assuring government is democratic. To some extent, this service quality is more complex now that there circumvents the inability to relate cause and are more than 1,000 decentralized units. effect because client satisfaction is in the Technology generation is cited as another interest of the locally elected government. On weakness of the decentralized arrangement. the other hand, the scope for local political A recent World Bank-financed agricultural interference in technical matters is increased. extension project in Venezuela began with pilot Experiments in decentralizing public innovations in decentralization that may extension services have been carried out in counteract some of Colombia's potential several Latin American countries, building on problems. The pilot activities included major national decentralization initiatives that subcontracting a university or NGO7 to provide are common in the region (World Bank, 1997). an extension team in each of the five pilot Between 1989 and 1993, the government of municipalities. The main project decentralizes Colombia decentralized extension to all but 10 planning and implementation of extension to the of the country's 1,050 municipalities. Each was municipality level; forms farmers' associations required to create its own extension office, to administer the municipal extension service; Unidad Municipal de Asistencia Tecnica contracts private extension consultants, NGOs, Agropecuaria (UMATA), which provides and universities to manage the extension service technical assistance to small farmers on a full provision; and provides for cost-sharing range of issues. The 1,040 UMATAs employ between national, state, and municipal levels of 3,500 technical and professional staff, and government and the beneficiaries themselves. provide 450,000 small farmers with free Field visits have confirmed client (farmer) extension services (28 percent of all small satisfaction with the service they now receive farmers in Colombia). Thus far, almost all costs from extension agents. are sustained by the central government through a complex system of budget transfers to Single commodity-focused extension municipalities and matching grants. Commodity-specific extension has been To date, the Colombian decentralization has practiced across the public, parastatal, private improved coverage (through larger staff inputs) and social sectors, including by agroprocessing and possibly responsiveness. The number of and marketing firms and farmers' associations municipalities served is up over 300 percent, (Nagel, 1997; Umali and Schwartz, 1994). The while beneficiaries are up over 250 percent, focus is often on one commercial or export crop with the beneficiary/technical staff ratio falling linked to established marketing or processing 17 percent to 129, and cost per fanner outlets, or on one aspect of farming, such as diminishing by 10 percent (in line with staff livestock or dairying (Purcell and Anderson, increases). Municipality staff are reportedly 1997l Axinn, 1988)d more accessible to small farmers than staff who The distinctive feature of commodity- were formerly under the MOA-directed specific extension lies in vertically integrating Colombian Institute for Agriculture and most of the components of the production and Livestock (ICA). At least in these early stages, marketing system, including research, input however, the decentralization has complicated staff and program quality control (monitoring and accountability), exacerbated political interference at the local level, interrupted 7. We use the term 'NGO' throughout the text to denote research-extension links, and increased staffing nongovernment organizations or private voluntary organizations. 18 supply, product marketing, credit, extension, interaction of the commodity with other and sometimes price assurance. It thus deals components in the smallholder production most effectively with the generic problem of system are neglected (Purcell and Anderson, dependence on wider policy and other agency 1997). Accountability to farmers is often functions by internalizing the complementary questionable unless the commodity organization services. Complexity and scale are simplified by is controlled by a farmers' association (Axinn, a narrow focus on one commodity. As a 1988). relatively small organization that includes all Fee-for-service public provision aspects of the commodity process, this organizational format to a large extent resolves Introducng user charges or fees for services iS the problem of relating cause and effect. For the have adopted, primarily as a cost-recovery same reasons, staff accountability is readily strategy. Farmers pay a portion of the fees, but assured by uniform salary, training, and staff the govemrment also pays on a contract basis. In conditions of service. tegvrmn lopy nacnrc ai.I conditiocs of narrvice. commdityconcrnssome cases, commercialization is a transitional The focus on narrow coimodity concemrs stage toward privatization. Issues of fiscal reduces liability to other public service sustainability and accountability are the primary functions. The small-and-focused approach is generic problems addressed by this approach. relatively cost-effective, and through levies on Apart from recovering costs from farmer product sales or by factoring cost-recovery into clients who can afford to pay for information product or input prices, fiscal sustainability is and advice, the fundamental effect of achieved. Finally, almost by definition the introducing fees for services, along with vertically-integrated structure assures a performance-related staff commissions, is to Agroprocessing and marketing firms bring about a more professional, client-oriented provide extension services to their farmer relationship between extension agent and suppliers to reduce input supply risks, reduce farmer, thus improving both accountability and spost-harvesto reduces,and inputpplyro nis , reduefficiency. In the New Zealand example post-harvest losses, and improve quantity, described below, there is evidence of some quality, consistency, and timeliness of output. In positive outcome by integrating extension with the agroprocessing and marketing operations, other functions, and tracing cause and effect. extension services are typically an integral Positive impact on the scale (coverage) problem component of contract growing schemes is only obtained if the introduction of user involved in producing high-value commodities. charges is accompanied by stratifying the client Umali and Schwartz (1994) provide numerous harket and arranging special serevices for less examples from around the world of the broad commercial farmers. In the absence of such range of commodities promoted in this manner. protection, fee-for-service extension would They also document many examples of farmers' likely exacerbate the generic problem of associations and cooperative commodity coverage in lower-income countries, and as with ventures, most reforms directed toward privatization, The commodity approach has inherent present problems of commitment and political limitations. As Nagel (1977) observes, the support. Other weaknesses are discussed below. advantages are largely defined from the The advisory services of New Zealand and perspective of the commodity organization. the UK were both public extension providers Obviously, in situations where fare ing is not a that expanded the proportion of their services monoculture, the approach does not fit as well available for a fee. New Zealand began cost- because the narrow commodity focus tends to averyblefor a . New leave the rest of the agriculture sector and the Zealand (AgNZ), formerly the ministry advisory extension system with many unaddressed needs. service, was sold by the govemrment to As a result, scale and coverage, other public Wrightson's, a farm service and stock trading service functions, and issues relating to the 19 company, in 1995 (Mavromatis, pers. comm.). been filled by the creation of a number of Government contracts and training contribute 50 organizations funded largely through producer percent, and the balance is derived from levies. extension contracts with individual farmers and Citing Hercus (1991), Rivera and Cary farmer organizations, research institutes, and (1997) note that commercialization in New agribusiness companies. Zealand, besides reducing the public fiscal The UK's advisory service, ADAS, initiated burden, improved accountability and ability to a system of charges in 1987. It became a more trace cause and effect by involving extension autonomous, government-owned agency in staff in the entire production-processing- 1992, received no government subsidies in transporting-marketing chain. It also shifted 1996, and as of April 1997, became a private- toward a stronger client orientation and a sector company purchased by management and concern to identify and produce results rather staff (Griffis, 1996 and pers. comm.). When than simply to engage in activities. charges were introduced in 1987, cost recovery For ADAS, Griffis describes increased was the objective, with farmers required to efficiency and reduced cost to government, contribute to the cost of advice that directly higher job satisfaction for the majority, more benefited their businesses. Privatization was not professional relationships with customers, and a stated objective at that time, but became one in better focus. On the negative side in New 1994/95 when it became clear that full cost Zealand, there remained some concerns about recovery was possible (Griffis, pers. comm.). interaction with knowledge generation and The UK government's Ministry of Agriculture, coverage for small farmers. Research agencies Fisheries, and Food (MAFF) remains the largest and advisory consulting work became sharpened customer for ADAS - undertaidng a range of in focus and efficiency. AgNZ engaged in some activities concerned with policy development, specific 'public good' technology transfer implementation, and monitoring, which includes projects on a contract basis to commodity contracted extension. research agencies and the national Foundation As the commercialized ADAS looked for for Research, Science, and Technology. But other business, its client structure changed, interaction among organizations diminished and focusing on farmers willing to pay, and feedback from farmers to science providers corporate business clients associated with declined. agriculture. Commercial pressures meant that In the view of Mavromatis, AgNZ's general less time became available for extension, and manager, less well-off farmers tended not to this was soon limited to the activities that attend field days when AgNZ was a MAFF required and for which they were willing government-funded extension service, but today to pay. As a result, the total number of farmers farmers are exposed to more diverse opinions. reached dropped slightly. In the UK context, this Also in the UK, services to low-income farmers was not viewed as a significant problem because have been reduced because some of these a great deal of information is available to farmers have not been convinced that benefits farmers from other sources. outweigh fees for ADAS services, given the While no formal assessment of the impact of availability in the UK of information from other the New Zealand changes has been undertaken, sources. the increased activity of consultants has Howell (1986) and Rivera and Cary (1997) improved their performance and job satisfaction note the limited scope for funding public (Mavromatis, pers. comm.). Commercialization extension services by user fees in developing may also have had a positive overall effect on countries. Obvious difficulties may be in coverage through he_growth in the number of collecting user fees, establishing cost- farm consultants throughout the country. accounting procedures, and reorienting and Mavromatis reports that the gap in public good retraining extension staff (Griffis, pers. comm.). extension resulting from privatizing AgNZ has 20 In primarily subsistence economies, user coupon schemes, cofinancing, and collaborative charges for 'common good' general extension arrangements with NGOs and farmer information would be difficult to enforce and organizations. possibly reduce general participation in By involving a variety of stakeholders in extension (Howell, 1986). Umali-Deininger forging contracts and collaborative partnerships, (1996) notes that the demand for fee-based pluralistic arrangements have the potential to extension services will almost exclusively come help resolve two fundamental generic problems from market-oriented farming operations, - linking cause and effect, and accountability particularly from areas dominated by medium- or incentive to deliver quality service. In to large-scale farmers. Accordingly, she subcontracting arrangements, the provider's proposed that partial cost recovery may be a client orientation is strengthened through the means of fostering a more demand-driven contracting process, and the farmer's influence system and serving as an important transitional as a fee-paying customer increases. phase toward developing a market for fee-for- Accountability in these arrangements tends to be service extension. Cost recovery and user co- multifaceted, with several stakeholders involved financing components have been incorporated in developing contract terms of reference, into World Bank-funded projects in Chile, competitive bidding, and direct input from Mexico, and Venezuela, and Nicaragua. farmers in the design of indicators. Wilson (1991) describes how cost reduction The motive behind subcontracting may be to in Mexico is achieved partly by stratifying the "get around the institutional inefficiencies client market by income level, and either associated with public delivery" (Umali- progressively graduating higher-income Deininger, 1997). These usually include various producers to private extension services or tenurial problems affecting staff performance requiring greater cost-sharing. Stratifying for that apply less in the private sector. Involving cost recovery reduces both generic fiscal and nonprofit NGOs may further improve liability problems, releasing public resources for responsiveness, cost-effectiveness, and equity in an 'extension safety net' targeted at low- to coverage. middle-income producers in priority areas The difficulty of institutional pluralism is (Umali-Deininger, 1996). for central government to adjust to a position of reduced direct control over either program or Institutionalpluralism - staffing. Additionally, financial and mobilizing other players adrninistrative management may increase in Innovations within this category are designed to complexity, at least initially, as new systems are create a more pluralistic system of developed. Additional resources and efforts may complementary extension services that would be required to monitor service quality. Where reach and respond to diverse farmers and these challenges can be overcome, and where farming systems (World Bank, 1997). A fine complementarity can be achieved through line separates some of these arrangements, such rationalizing public and private sector roles, as cofinancing and subcontracting, from fee-for- mobilizing other players through these service extension. The emphasis in fee-for- approaches can resolve the problems of service extension was (with some exceptions) coverage, ability to relate cause and effect, more on cofinancing public provision, while accountability, fiscal sustainability, and here the emphasis is on moving more toward interaction with knowledge generation. private provision. After discussing the broad Several principles underpin innovations in ways in which innovations in this category this category. First iS unlinking public funding overcome some generic problems, we offer from public delivery (Zijp, 1998). Second, examples of subcontracting arrangements with pluralism implies changes in governance. New private sector firms, including voucher and institutions and institutional arrangements, such as public-private partnerships, are involved. A 21 key governance principle is to open and MOA which individual farmers should graduate democratize extension control so that all from the program. stakeholders may express their perspectives and There is still insufficient information to interests, and play appropriate roles in extension judge whether these coupon and voucher design, implementation, and evaluation. systems entail significant costs. Experience of Third, with pluralism the government voucher systems in the education field (West, recognizes that to meet diverse needs and 1996) suggests that administrative costs might conditions in the farming sector, it should invest be substantial, threatening the sustainability of more broadly in the whole agricultural the voucher system as a whole. knowledge and information system (AKIS), Cofinancing in Ecuador and Honduras aims rather than in public sector extension services to replace the public extension services with a alone. Implied in each of the above principles 'technology transfer market' financed jointly by are significant role changes for ministries of government and the beneficiaries, in which agriculture as they move away from service private agents will compete to provide services. delivery toward providing an enabling policy Both cases involve stratification of farmers into environment, coordinating and facilitating the small-scale producers and medium- and large- work of other players, and assisting farmers in scale producers, with separate programs and negotiating terms of contracts, monitoring graduated scales of copayment designed for quality, and exercising financial control. each farmer category. Recognizing that complete privatization of Collaborative arrangements with the NGO agricultural extension services is often not and nonprofit sector include cooperative feasible, several Latin American countries are arrangements with universities, commodity attempting to address such problems as fiscal boards, and commodity foundations (Umali and sustainability and poor client orientation by Schwartz, 1994; Nagel, 1997). Non-government integrating the private sector into extension organizations represent a highly prevalent systems (Umali-Deininger, 1996). Contracting partner in agricultural extension in developing only specific functions, for example staff countries, frequently focusing on areas training and video production, is one alternative, inadequately served by the government. Many as in Morocco. More comprehensive NGOs strive to be participatory, democratic, subcontracting of extension services has been responsive, cost-effective, community-based, attempted in countries as diverse as Estonia, and focused on the needs of hard-to-reach target Turkey, Madagascar, Costa Rica, and Mexico. groups. Coupons attached to agricultural bank loans, However, NGOs do not always live up to committing a certain percentage of the loan for their articles of faith. Some NGOs push their extension services, have been used in Colombia own agenda and are more accountable to (Rivera and Cary, 1997). Extension vouchers external funding sources than to the clientele are a subcontracting innovation launched in they aim to serve (Farrington, 1997). Hence, Costa Rica and Nicaragua. In Nicaragua, care must be exercised to ensure that partnership farmers issued with vouchers were able to arrangements with NGOs capitalize on their choose their extension supplier, either public or strengths and avoid their weaknesses. private (Umali-Deininger 1996). The government retains a role not only in financing, Empowerment and participatory approaches but also in regulating extension providers. In Evolving control by beneficiaries may be the Costa Rica, vouchers vary according to the types single most important initiative to make of farmers and levels of technology requiring extension accountable to clients for high- or low-intenslty technical assistance. At project completion, beneficiaries are expected to perfon ankeytgene wAness of continue with solely private technical assistance. eesin on ions tho 1994). The private extensionist is to indicate to the 22 emanating from inability to trace cause and backgrounds, including women, who are able to effect. We have already alluded to farmers' perform many extension agent roles in a cost groups as beneficiary organizations required by effective manner (Axinn, 1988; Russell, 1986). some decentralized, cost-recovery, Participatory approaches have been found to subcontracting, and cofinancing arrangements. adjust complementary services more closely to Farmer organizing is a key element in the client- farmer needs (Axinn, 1988), as well as reduce based extension strategies of NGOs (Nagel, farmer dependence on external inputs (Roling 1997). We also mentioned in a previous and Pretty, 1997). subsection that farmers' associations organized Farmers' union formation in Europe on commodity lines actually provide extension improved the integration of complementary services to their members (Umali-Deininger, services and raised political support (Roling, 1996). Some of these commodity-based farmers' 1986)8. Participatory approaches also have a organizations have been highly successful, for positive effect in terms of tracing cause and instance in the dairy industry in India (Chamala effect through farmer-led experimentation and and Shingi, 1997). analysis, and farmer feedback (Axinn, 1988). A A wealth of experience and lessons on key positive impact of participation is participation and farmer control testifies to accountability. Axinn goes as far as to state that potentially positive effects on generic problems. 'automatic quality control' is achieved through However, their impact depends where on the raising farmer awareness and confidence. The 'participatory continuum' a particular initiative quality of trust established (Pretty and Volouhe, lies. Pretty and Volouhe (1997) reviewed the 1997) and ownership (Chamala and Shingi, rapid/participatory rural appraisal field, citing 1997) are also emphasized. no less than 30 different terms and names for Fiscal sustainability is improved through alternative systems of participatory learning and mobilizing local resources. Cost-effectiveness action that have sprung up over the last decade and efficiency are achieved by using relevant or so. Many of these involve 'self-mobilizing' methods that focus on expressed farmer needs methods - in group and team dynamics, and local people taking over many extension sampling, interviewing and dialogue, and roles (Axinn, 1988). Participation has positive visualization and diagramming. effects on the interaction with knowledge It is clear from examples they give that the generation by combining indigenous knowledge conscious use of these methods, which pay great with feedback into the agricultural knowledge attention to monitoring and self-evaluation, has system (Axinn, 1988; Chamala and Shingi, a significant impact in raising the level of trust, 1997), and identifying and verifying applicable understanding, and links among the various technology and technology generation needs for actors and agencies involved in a rural situation, farming systems (Purcell and Anderson, 1997). with the farmers at the center of the process. France, Norway, and Taiwan provide Chamala and Shingi (1997) review lessons examples of the prominent role farmers' learned in establishing and strengthening farmer organizations can play in funding and organizations, confirming valid extension roles organizing agricultural extension (see Ameur, in farmer empowerment, community organizing, 1994; Umali and Schwartz, 1994; Haug, 1991; human resource development, and problem Nagel, 1997). solving and education. Axinn's analysis of the In Taiwan, 90 percent of the country's 'participatory approach' also notes its farmers belong to farmers' associations advantages and disadvantages (1988). Participatory approaches have positive effects for most of the generic problems of extension. On the problem of scale and 8. In the Netherlands, 95 percent of farrners operate within provincial organizations linked to national and even coverage, participatory approaches produce European farming unions, and are able to exercise farmer leaders with appropriate local substantial influence on agricultural policies and hold extension accountable. 23 organized at township, county, and provincial countries with a large base of small-scale, levels (Nagel, 1997), with an overall extension subsistence farmers (Umali-Deininger, 1996). In policy defined by the government that includes such circumstances, public sector finance strong, institutionalized links with research and remains essential, mixed with various cost- other services. Extension is carried out by recovery, cofinancing, and other transitional agents employed by the farmers' associations at institutional arrangements that are appropriate to the township level and financed largely by the the pace of structural and commercial changes farmers themselves. In Argentina, the farmer in agriculture. group movement CREA (Agricultural If privatization is appropriate, some generic Experimentation Regional Consortia) has spread problems will be positively affected. All to Brazil, Chile, Paraguay, and Uruguay. It privatization efforts report improvements in exemplifies fully self-supporting extension accountability, usually expressed in terms of organized by farmers (Tobar, 1996). client orientation and satisfaction (thus avoiding Not all participatory attempts produce a the cause and effect problem). Obviously, wholly positive experience. Often private sector initiatives are not affected by decentralization is virtually a prerequisite for other public service liabilities. All privatized effective local participation. On the other hand, efforts claim improved efficiency, cost- in conventional agriculture with more passive effectiveness, and reduced public sector costs by forms of participation, there is a risk that the servicing the needs of farmer clients who can elite will capture delivered benefits (Howell, afford to pay for the information, thus 1986). overcoming the problems of sustainability and dependence on fiscal allocations. Incentives Privatization exist for private providers of extension to Use of the term privatization often tends to be maintain close links with knowledge generation misleading. In its pure sense, privatization agencies in order to have a marketable product. implies full transfer of ownership (usually by For other generic problems, the impact of way of sale) from government to a private privatizing extension is mixed. Overreliance on entity, with that entity meeting all costs and private extension risks neglect of less receiving any profits (Rivera and Cary, 1997). In commercial farmers and lower-value crops. most cases, governments have not actually Stratificaton and separate, publicly-funded privatized their agricultural extension services targeted programs are needed to counter this in this sense. risk. Privatization also does not deal with the The private sector has the incentive to complexity of providing a socially and provide private and toll-good information to environmentally optimal service. 'better-off' commercial farmers and members of Most analysts suggest a cautious, private associations for whom extension service evolutionary approach toward privatization delivery is profitable. In areas dominated by within a clearly formulated mission and commercial farming and farmers with strategy, along with open communication among marketable output, it makes sense to mobilize all stakeholders. Privatization may start with the private sector to provide investment capital more commercial farmers for whom technology and serivices (World Bank, 1997). Input packages already exist and extension is largely a suppliers have strong incentives to provide delivery function. It may begin in a single region advice on a range of crop and livestock and expand over time, bnnging farmers to the activities. With the increased commercialization point where their future extension needs can be of agriculture in many developing countries, this met by private-sector services or provided on a source of technical knowledge may assume a fee-paying basis, leaving the public service to much more prominent role. Fully privatized serve new clientele and cropping systems, extension is not econonically feasible in including more marginal groups (World Bank, 1990). Below we describe a few cases where 24 governments have privatized all or part of their consulting firms to contribute up to 30 percent extension services, usually in stages. There are of the cost of the program. The government plenty of other instances where private entities agency prepares terms of reference for the have found it profitable to provide extension contracts. A recent program for medium- and services. large-scale farmers in Chile is now totally Chile completely privatized its extension privately funded and executed by the National system in the 1970s, forcing commercial Agriculture Society. producers to obtain extension services from In the Netherlands, the government is half- privately-owned consulting firms (Umali- way through a 10-year phased privatization Deininger, 1996), and has since been taking process in which its public extension service, steps to rationalize services for different farmer DLV, became a foundation with farmers' categories. While the large commercial farmers organizations and the government equally were not seriously affected by the shift to full represented on the board (Rivera and Cary, private-sector delivery, small-scale and 1997). In December 1997, agreement was subsistence farmers were left out of the reached with the Dutch government that all extension market. direct bilateral financial contracts with DLV As a result, the Chilean government had to would terminate after three more years.9 From actively target extension services to groups with 2000 onwards, government contracts for smaller marketable output. Two target groups extension programs will be tendered in the open were identified and served by specially designed market with competition among all parties programs with World Bank assistance. One interested in carrying out the programs. The subgroup of farmers considered to have foundation will then change into a stock sufficient resources to eventually attain self- company with shares held by employees, and sufficiency and market surplus production temporarily, by the MOA. received a producer-oriented extension package In some respects, the Dutch privatization through a special program. This beneficiary was an upheaval. Tacken (1996) records how 50 group paid 15 percent of extension costs, a rate percent of DLV's original staff had to be fired which may reach 50 percent. The second or take early retirement in a process subgroup of poorer, subsistence farmers characterized by considerable tension as DLV received a more farm family-oriented, 'basic' changed its organizational culture. service for free, with a proposed eventual Tacken reports a number of positive effects contribution of 15 percent of total costs. of the privatization, including a 50 percent Services for both programs were delivered reduction in overhead. Client satisfaction by contracted private consulting firms. Wilson jumped 40 percent in three years, attributed to (1991) notes the goal is to increase commercial improved client orientation, quality control, and and family farm contributions, and that the more specialization of advice. Extension overall Chilean cost reduction principle is to programs before privatization were closely target free services, limit the period of coverage related to national policy goals, often conflicted to general recipients, and work with groups to with farmers' interests, had low impact, and led reduce costs. To qualify for the Chilean to conflicting roles, low morale and turnover of program, a firm must meet technical and extension staff and lack of farmer confidence. professional staffing criteria, bid for contracts, The service was neither needs-driven nor and agree to have its activities supervised and problem-oriented, and overused mass media for evaluated by a designated public agency. delivery. After privatization, farmer interests Ameur (1994) reports that these programs became a priority as farmers became strongly reach a large number of small-scale farmers, the represented on the DLV board and sectoral typical farner/agent ratio is as low as 48 to 1, consulting firms focus only on technology transfer, and farmers contract with their 9. Personal communication, May 1998. 25 councils. Staff attitudes on efficiency, Taiwan. In Brazil around 1988, there were over effectiveness, and job satisfaction improved as 2,000 private consulting firms largely catering DLV monitored individual performance. to the specialized needs of the commercial Albania is another recent case where livestock sector. In Asia, consulting firms tend privatization has been attempted on a national to concentrate on plantation crops, often staffed scale. With World Bank assistance, the by ex-plantation managers and technicians. government created a nationwide competitive private-sector network of 300 dealers to deliver Interconnecting rural people and using inputs and technology to 600,000 newly appropriate media established, small-scale private farmers (Schultz The arrival of the information age has naturally et al., 1996). The initiative explicitly tackled the led to an interest in its potential for innovative generic problem of dependence on wider policy applications of the latest information and other agency functions because the technologies (IT) to enhance extension delivery. government implemented policies in a Considering appropriate media for extension consistent and coordinated manner to support also includes the place of more traditional private-sector development. In addition, a private-sector tradevassocint.In (AFadAi and extension methods, such as mass media, group ptevatelbanian FarerssUnion it sponsord meetings, field days, demonstrations, and theAlbniaFanerUnont sponexchange visits (Campbell and Barker, 1997). lobbied the government to reduce import duties Innovations in this category are most and remove a value-added tax on fertilizer. The reormedMOA rle becme enirelydirectly associated with overcoming the generic The reformed MOA role became entirely .. . . ~~~~problems of scale and complexity and - related to supporting policies, services, and fair trough osciencociatd rules. fo coptto in th.natpivtetr through cost-efficiencies associated with certain rules for competition in the infanttrvt eco, mspmda-fsclssanvabltyeto,. In Abna including an information base, weekly market mas mei iclssanaiiy nAbna pincludinge a informationatnd reabase, wekculy l mart the privatization effort described above involved price information,eand prelblem agpricultura data.extensive use of mass media (television, radio, The other generic problem improved by adawdl-icltdmnhyarbsns privatization was accountability through the and a widely-circulated monthly agribusiness long-term relationships private dealer firms newsletter) to inform, create public awareness, create with customers. To overcome the generic and change attitudes. This media effort problem of political co-mitment, a direct supporting the reformed MOA role helped to approach to the private sector (that bypassed overcome the generic problem of dependence on vested power and rent-seeking interests which wider policy. In a study by Wete (1991), the most delayed and disrupted the reform process) was Inal aproc - prin withe gpsa required. The generic problem of interaction radioe s t appro prate and with knowledge generation remained a problem. radio -emerges as most appropriate and cost- wihulth knowedge generati remained al proble effective in a developing country. Zijp's (1994) Schultz et al. (1996) record that although privatization facilitated transfer and adoption of study provides numerous examples of cost- effective use of different IT applications. Their technology, reduced government funding also poitv imat.oee,ismtmsahee created competition rather than cooperation . . . i within the knowledge system, hampering only at a significant initial and operating cost. communication with research, education, farmer Further, this impact tends to occur when the media are used in combination with other organizations, private consultants, and suppliers. inoaosthtwhveriwd,nicig . m~~~~inovations that we have reviewed, indicating Finally, as we have mentioned, farmers' that IT is best considered not in isolation, but as associations, often along commodity lines, a 'force multiplier' enabling or enhancing the provide their own private extension services to members. ~ Uml-enne (19)Also effectiveness of other innovations and members UmaliDeiningr (199) alsoconventional extension methods (Antholt, documents the role of private consulting firms conventoaliextion metho (AnthofT, providing extension services in Argentina, 1994). Generallzatlon about the effect of IT on Brazil, Colombia, Mexico, Uruguay, Korea, and 26 range of media that may be considered for information, educational, and organizational different extension purposes. needs of the community. AED (1985) found the suitability of In Sweden, 50 telecottages became different media dependent upon the message, operational within five years after the first one target audience, and social environment. Radio opened in 1985. Their aim is to counteract and television are more appropriate for reaching geographical disadvantages, and provide many people quickly with relatively simple isolated villages with access to databases, data ideas, while print media are better suited to processing, fax, e-mail, computer-assisted provide timely reminders of information. services, open university and on-line tutorials, Interpersonal communication, including and village hall facilities. extension agents, group meetings, and Wete (1991) pointed out some limitations: demonstrations are best suited for teaching and * information alone is an insufficient enhancing credibility of information. Based on condition for social change; Lionberger's (1968) model of the adoption * far from being neutral, provision of process, Campbell and Barker recommend: information can actually widen the gap * mass media and popular theater to provide between rich and poor; new or additional information in the * communications technology (CT) does not 'awareness' stage; have produce effects without government * group meetings, radio, and field days to commitment to change, reflected in its increase knowledge in the 'interest' stage; provision of budgetary support and * result and method demonstrations, and conducive policy and complementary farmer exchange to improve skills in the services; and 'evaluation' stage; * most developing countries cannot afford CT e individual visits, farmer exchange, on-farm hardware costs; as a result, the benefit/cost trials, and method demonstrations to induce ratio of some CT applications is doubtful. behavioral change in the 'trial' stage; and In overcoming generic problems, the i recognition programs, competitions, and limitations of IT/CT have much to do with incorporating practices into farming systems innovations in this category not standing alone. to consolidate attitudinal change in the Thus, by itself, IT cannot overcome the problem 'adoption' stage. of relating cause and effect. The principle of 'Interconnectivity' describes the use of 'garbage in, garbage out' applies, and care must appropriate information and communications be taken that the overzealous use of IT does not technology (IT) to enable people to connect with give biased or inaccurate information an aura of other people (Zijp, 1997). To harness its full veracity. In terms of coverage, IT cannot replace potential requires considerable commitment and face-to-face contact between extension agents some radical changes in perspective. One and farmers. Also, there are tendencies toward change is to lessen our reductionist, sectoral urban and other biases (for example, against the orientation in favor of a pluralist, cross-sectoral, disadvantaged and traditional culture) in some systems perspective. Thus, interconnectivity is IT applications. closely interrelated to decentralization and Nevertheless, increased investments in institutional pluralism. Community many IT applications appear to make sound communication centers, or telecottages, economic and social sense, and deserve public exemplify the new partnerships emerging for sector support (Zijp, 1997). local information access, communication, and education in rural areas. They are as diverse as Interpretation and conclusions the communities they serve, but many are self- In designing extension, an approach is less financing after initial start-up, and they all shift important than its ingredients. Using the control and accountability to focus on the framework developed above, it is important to 27 isolate ingredients of success and find ways to release public funds to serve smaller-scale replicate or transfer these characteristics to farmers. Conversely, increased coverage by improve the performance of another approach. enlarging the MOA and attempting to improve A familiar example is to explore ways of the management of its general services is not integrating positive characteristics of private- sustainable in the long run. sector or NGO operations into public-sector Not surprisingly, moving to the right in row management. A less familiar example, but one 7 of the matrix (Fig. 1), the generic problem of which springs from this framework, is to explore funding begins to look more positive - the modifications to the vertical integration feature matrix is arranged in the direction of public to that contributes to the success of the commodity private. approach. Innovations that resolve the generic This implies broadening the historical problems of scale, dependence on other policies tendency of extension to focus on production, and services, and inability to trace cause and and pay more attention to transformation and effect tend to resolve several other generic marketing. Perhaps most important would be to problems as well. Here, we notice some integrate farmer participation and control into significant findings. The conventional wisdom other extension modifications and a means to of the past has tended to look to massive, overcome several critical generic problems of technocratic, and sophisticated efforts - in scale and complexity, relating cause and effect, management and media -to seek impact in and accountability. This requires proportion to the perceived scale and decentralization, and is even more effective complexity of these problems. when institutional pluralism is built in. In contrast, we notice that the most Looking across the rows of the matrix in significant effect on these three problem areas Figure 1, we can assess which innovations, involves concepts and practices falling broadly alone or in combination, offer solutions to a within the approach of sustainable agriculture. given generic problem or set of interrelated Without a full discussion, its advocates10 believe problems, and whether a gain in overcoming one that a sustainable approach - low external problem is to be had at the expense of another inputs, a systems orientation, pluralism, and problem. In some instances, care is required to reliance on arrangements that use the incentives avoid a potential pitfall and achieve a positive of both farmers and those who serve them- trade-off between the pros and cons inherent in releases the local knowledge, organizing ability, an approach. To a large extent, our framework resources, and commonsense of rural people to confirms most current professional wisdom. overcome policy constraints. For example, impact on the coverage Rural people know when what is being done problem is most powerful through participation is relevant and effective. The ingredients of a and control by farmer organizations, mobilizing sustainable approach tend to be inherently low other players (e.g., NGOs), and using cost and build relationships of mutual trust and appropriate media. These ingredients involve reciprocity. From these relationships, using local people as field agents who belong to commitment, political support, accountability, target groups, training extension workers in fiscal sustainability, and effective interaction human resource development skills, and with knowledge generation develop. collaborating with community organizations and their support groups to help them use their own systems of knowledge, experimentation, and communication. Impact on coverage can also be obtained by prioritizing, categorizing, and stratifying farmers into target groups, using cost-recovery schemes 10. Such as Roling and Pretty (1997), Axinn (1988), with more comnmercially-oriented farmers to Antholt (1994), Rivera (1991), Haverkort and de Zeeuw (1991). 28 References AED (Academy for Educational Development). The World Bank Research Observer 1985. Communication Strategiesfor 12(2)183-201. Agriculture: Hybrids of a Different Kind Birkhaeuser, Dean, Robert E. Evenson, and Washington, D.C.: AED. Gershon Feder 1991. The Economic Impact Albrecht, Hartmut. 1986. Extension Research: of Agricultural Extension: A Review. Needs and Uses. In Investing in Rural Economic Development and Cultural Extension: Strategies and Goals (Gwyn E. Change 39(3)607-650. Jones, ed.). London and New York: Bunting, A.H. 1986. "Extension and Technical Elsevier. Change in Agriculture." In Investing in Amanor, Kojo and John Farrington. 1991. Rural Extension: Strategies and Goals NGOs and Agricultural Technology (Gwyn E. Jones, ed.). London and New Development. In Agricultural Extension: York: Elsevier. Worldwide Institutional Innovation and Campbell, Dunstan A. and St. Clair Barker Forcesfor Change (W. M. Rivera and D. J. 1997. Selecting Appropriate Content and Gustafson, eds.). Amsterdam: Elsevier. Methods in Programme Delivery. In Ameur, Charles 1994. Agricultural Extension: A Improving Agricultural Extension: A Step beyond the Next Step. World Bank Reference Manual. FAO: Rome. Technical Paper 247. Washington, D.C.: Chamala, Shankarah and P.M. Shingi. 1997. World Bank. Establishing and Strengthening Farmer Antholt, Charles H. 1994. Getting Ready for the Organizations. In Improving Agricultural Twenty-First Century: Technical Change Extension: A Reference Manual. FAO: and Institutional Modernization in Rome. Agriculture. World Bank Technical Paper Chambers, Robert 1983. Rural Development: 217. Washington D.C.: World Bank. Putting the Last First. Harlow, UK: Antholt, Charles H. 1991. Agricultural Longman. Extension in the 21st Century: Lessons from Evenson, Robert E. 1986. The Economics of South Asia. In Agricultural Extension: Extension. In Investing in Rural Extension: Worldwide Institutional Innovation and Strategies and Goals (Gwyn E. Jones, ed.). Forces for Change (W. M. Rivera and D. J. London and New York: Elsevier. Gustafson, eds.). Amsterdam: Elsevier. Axinn, George 1988. Guide on Alternative Farrington, John 1997. The Role of Nongovernmental Organizations in Extension Approaches. Rome: FAO. Etnso n.aI I rgAgricuturali Extension. In Improving Agricultural Benor, Daniel, James Q. Harrison, and Michael Extension: A Reference Manual. FAO: Baxter 1984. Agricultural Extension: The Rome. Training and Visit System. Washington, Feder, Gershon and Roger Slade 1993. D.C.: World Bank. Institutional Reform in India: The Case of Bindlish, Vishva and Robert E. Evenson 1997. Agricultural Extension. In The Economics of The Impact of T&V Extension in Africa: Rural Organizations (K. Hoff, A. The Experience of Kenya and Burkina Faso. 29 Braverman, and J. Stiglitz, eds.). Oxford: Hercus, John. 1991. The Commercialization of Oxford University Press. Government Agricultural Extension Feder, Gershon and Roger H. Slade 1986. Services in New Zealand. In Agricultural Methodological Issues in the Evaluation of Extension: Worldwide Institutional Extension. In Investing in Rural Extension: Iva and D. r.Gsfongeds W. Strategies and Goals (Gwyn E. Jones, ed.). Amserdam: E usevoer. London and New York: Elsevier. Amsterdam: Elsevier. FAO (Food and Agriculture Organization of the Howell, John. 1986. Accountability in United Nations). 1997. Improving Extension Work. In Investing in Rural Agricultural Extension: A Reference Extension: Strategies and Goals (Gwyn E. Manual. Romex FAOn Jones, ed.). London and New York: Elsevier. FAO (Food and Agriculture Organization of the Howell, John. 1985. Recurrent costs and United Nations). 1990. Global Consultation agricultural development. London: on Agricultural Extension. Rome: FAO. Overseas Development Institute. Garfield, Elsie, Maurizio Guadagni, and Daniel Jones, Gwyn E. and Chris Garforth. 1997. The Moreau. 1996. Decentralization of History, Development, and Future of Agricultural Extension in Colombia. Agricultural Extension" hi Improving Extension Workshop, Alternative AgriculturalExtension: A Impreng Mechanisms for Funding and Delivering Agricultural Extension: A Reference Extension. Washington, D.C.: World Bank. Manual. FAO: Rome. Kaimovitz, David. 1991. The Evolution of Links Griffis, David. 1996. ADAS (United Kingdom). between Research and Extension in In Extension Workshop, AlternativebewnRsarhndEtsinn M chaensis fork Funin And Deiver Developing Countries. In Agricultural Extension. Washington, D.C.: World Bank. Extension: Worldwide Institutional Innovation and Forces for Change (W. M. Gustafson, Daniel J. 1991. The Challenge of Rivera and D. J. Gustafson, eds.). Connecting Priorities to Performance: One Amsterdam: Elsevier. State's Response to the Forces for Change Keynan, Gabriel, Manuel Olin, and Ariel Dinar. inU.S. Extension. In Agricultural' ExtenU.SioExten:sWor Instricltutal 1997. Co-financed Public Extension in Extension: Worldwide Institutional Nicaragua. The World Bank Research Innovation and Forces for Change (W. M. Observer 12(2)225-248. Rivera and D. J. Gustafson, eds.). Amsterdam: Elsevier. Leonard, D.K. 1985. African Practice and the Haug, Ruth. 1991. Public-Private Cooperation: Theory of User Fees. Pages 130-145 in Farmer-Led Research/Extension Circles in Recurrent costs and agricultural Norway. In Agricultural Extension: development (J. Howell, ed.). London: Worldwide Institutional Innovation and Overseas Development Institute. Forces for Change (W. M. Rivera and D. J. Lionberger, H.F. 1968. Adoption of New Ideas Gustafson, eds.). Amsterdam: Elsevier. and Practices. Ames: Iowa State University Haverkort, Bertus and Henk de Zeeuw. 1991. Press. Development of Technologies towards Manarungsangan. 1989. Consulting Report on Sustainable Agriculture. In Agricultural Thai contract farming operations. Country Extension: Worldwide Institutional Operations Division, East Asia and Pacific Innovation and Forces for Change (W. M. Country Department 1. Washington, D.C.: Rivera and D. J. Gustafson, eds.). World Bank. Amsterdam: Elsevier. 30 Nagel, Uwe Jens. 1997. Alternative Approaches Goals (Gwyn E. Jones, ed.). London and to Organizing Extension. In Improving New York: Elsevier. Agricultural Extension: A Reference Roling, Niels and Jules N. Pretty. 1997. Manual. FAO: Rome. "Extension's Role in Sustainable Oakley, Peter. 1997. The Monitoring and Agricultural Development. In Improving Evaluation of Non-Material Objectives of Agricultural Extension: A Reference Extension. In Investing in Rural Extension: Manual. FAO: Rome. Strategies and Goals (Gwyn E. Jones, ed.). Russell, John F. A. 1986. Extension Strategies London and New York: Elsevier. Involving Local Groups and their Picciotto, Robert and Jock R. Anderson. 1997. Participation, and the Role of this Approach Reconsidering Agricultural Extension. The in Facilitating Local Development. In World Bank Research Observer 12(2)249- Investing in Rural Extension: Strategies and 259. Goals (Gwyn E. Jones, ed.). London and Pretty, Jules N. and Simplice D. Volouhe. 1997. New York: Elsevier. Using Rapid or Participatory Rural Schultz, James, Ray Diamond, Claude Freeman, Appraisal. In Improving Agricultural and Thomas Thompson. 1996. Albanian Extension: A Reference Manual. FAO: Agriculture Adjustment Project. Extension Rome. Workshop, Alternative Mechanisms for Funding and Delivering Extension. Purcell, Dennis L. and Jock R. Anderson. 1997. Washington, D.C.vWrld Ban. Agricultural Extension and Research: W Achievements and Problems in National Songlin, Liu. 1998. Current Situation of Systems. World Bank Operations Evaluation Agricultural Production and Agro- Study. Washington, D.C.: World Bank. Technical Extension in China. National Rivera, William M. 1991. Sustainable Agro-Technical Extension and Service Agriculture: a Unifying Concept for Center. Agriculture and a Central Strategy for Swanson, B.E. (ed.). 1984. Agricultural Extension. In Agricultural Extension: Extension: A Reference Manual. Second Worldwide Institutional Innovation and edition. Rome: FAO. Forces for Change (W. M. Rivera and D. J. Swanson, B.E., B.J. Farner, and R. Bahal. 1990. Gustafson, eds.). Amsterdam: Elsevier. The Current Status of Agricultural Rivera, William M. and John W. Cary. 1997. Extension Worldwide." In Global Privatizing Agricultural Extension. In Consultation on Agricultural Extension. Improving Agricultural Extension: A Rome: FAO. Reference Manual. FAO: Rome. Tacken, Willem. 1996. The Dutch Extension Rivera, William M. and D.J. Gustafson. 1991. Service: The Change from a Publicly New Roles and Responsibilities for Public Funded Extension Service to a Privatized, Sector Agricultural Extension. In Client-Oriented Organization. Extension Agricultural Extension: Worldwide Workshop, Alternative Mechanisms for Institutional Innovation and Forces for Funding and Delivering Extension. Change (W. M. Rivera and D. J. Gustafson, Washington, D.C.: World Bank. eds.). Amsterdam: Elsevier. Tobar, Julio Garcia. 1996. "CREA-AACREA Roling, Niels. 1986. Extension and the Associative Private Agricultural Extension Development of Human Resources: The in Argentina. Extension Workshop, Other Tradition in Extension Education. In Alternative Mechanisms for Funding and Investing in Rural Extension: Strategies and Delivering Extension. Washington, D.C.: World Bank. 31 Umali-Deininger, Dina 1997. Public and Private Change (W. M. Rivera and D. J. Gustafson, Agricultural Extension: Partners or Rivals? eds.). Amsterdam: Elsevier. The World Bank Research Observer 12(2)203-224. Wilson, Mark 1991. Reducing the Costs of 12*2)203-224. Public Extension Services; Initiatives in Umali-Deininger, Dina. 1996. New Approaches Latin America. In Agricultural Extension: to an Old Problem: The Public and Private Worldwide Institutional Innovation and Sector in Extension. Extension Workshop, Forces for Change (W. M. Rivera and D. J. Alternative Mechanisms for Funding and Gustafson, eds.). Amsterdam: Elsevier. Delivering Extension. Washington, D.C.: World Bank. 1997. Rural Development: From World Bank. Vision to Action. Sector Strategy, Umali, Dina L. and Lisa Schwartz. 1994. Public Environmentally and Socially Sustainable and Private Agricultural Extension: Beyond Development Studies and Monographs Traditional Frontiers. World Bank Series 12. Washington, D.C. Discussion Paper 236. Washington, D.C.: World Bank. 1990. Agricultural Extension: The World Bank. Next Step. Policy and Research Series No. Van den Ban, Anne W. 1986. Extension 13. Washington, D.C. Policies, Policy Types, Policy Formulation Zijp, Willem. In press. Unleashing The and Goals. In Investing in Rural Extension: Potential: Changing the Way We Think Strategies and Goals (Gwyn E. Jones, ed.). About and Support Extension. Working London and New York: Elsevier. Paper. Washington, D.C.: World Bank. Van den Ban, Anne W. and H.S. Hawkins. Zijp, Willem 1997. Extension: Empowerment 1996. Agricultural Extension. Second through Communication. Paper presented at edition. Oxford: Blackwell Science. symposium "Rural Knowledge Systems for West, Edwin G. 1996. Education Vouchers in the 21st Century: The future of Rural Practice and Principle: a World Survey. Extension in Western, Central, and Eastern Human Capital Development Working Europe." Reading, Cambridge, and Paper No. 64. Washington, D.C.: World Edinburgh, U.K. Bank. Zijp, Willem. 1994. Improving the Transfer and Wete, Francis N. 1991. New Technology for Use of Agricultural Information: A Guide to Transferring Agricultural Information. In Information Technology. World Bank Agricultural Extension: Worldwide Discussion Paper 247. Washington, D.C.: Institutional Innovation and Forces for World Bank. 32 Policy Research Working Paper Series Contact Title Author Date for paper WPS2110 Life during Growth: International William Easterly May 1999 K. Labrie Evidence on Quality of Life and 31001 Per Capita Income WPS2111 Agricultural Land Reform in Postwar Toshihiko Kawagoe May 1999 P. Kokila Japan: Experiences and Issues 33716 WPS2112 Industrial Policy afterthe EastAsian Ashoka Mody May 1999 S. Kpundeh Crisis: From 'Outward Orientation" 39591 To New Internal Capabilities? WPS2113 Wage Determination and Gender Stefano Paternostro May 1999 N. Nouviale Discrimination in a Transition David E. Sahn 34514 Economy: The Case of Romania WPS2114 Economic Reforms and Total Factor Pablo Fajnzylber May 1999 S. Holt Productivity Growth in Latin America Daniel Lederman 37859 And the Caribbean (1950-95): An Empirical Note WPS2115 Foreign Investment and Productivity Simeon Djankov May 1999 R. Vo Growth in Czech Enterprises Bernard Hoekman 33722 WPS2116 Does Child Labor Displace Martin Ravallion May 1999 P. Sader Schooling? Evidence on Behavioral Quentin Wodon 33901 Responses to an Enrollment Subsidy WPS2117 On the Size and Number of Regional Soamiely Andriamananjara May 1999 L. Tabada Integration Arrangements: A Political 36896 Economy Model WPS2118 Developing Countries and the Next Anne 0. Krueger May 1999 L. Tabada Round of Multilateral Trade 36896 Negotiations WPS2119 Tax Deductions, Environmental Ian Parry May 1999 R. Yazigi Policy, and the 'Double Dividend' Antonio Bento 37176 Hypothesis WPS2120 Can Africa Export Manufactures? Ibrahim A. Elbadawi May 1999 C. Bernardo The Role of Endowment, Exchange 31148 Rates, and Transactions Costs WPS2121 Fiscal Management in Federal William Dillinger May 1999 A. Lara Democracies: Argentina and Brazil Steven B. Webb 88148 WPS2122 Decentralization and Fiscal William Dillinger May 1999 S. Webb Management in Colombia Steven B. Webb 38680 Policy Research Working Paper Series Contact Title Author Date for paper WPS2123 Access to Land in Rural India Robin Mearns May 1999 G. Bumett 82111 WPS2124 Social Exclusion and Land Robin Mearns May 1999 G. Burnett Administration in Orissa, India Saurabh Sinha 82111 WPS2125 Developing Country Agriculture and Bernard Hoekman May 1999 L. Tabada The New Trade Agenda Kym Anderson 36896 WPS2126 Libert6, Egalite, Fraternite: Monica Das Gupta May 1999 M. Das Gupta Exploring the Role of Governance 31983 In Fertility Decline WPS2127 Lifeboat Ethic versus Corporate Monica Das Gupta May 1999 M. Das Gupta Ethic: Social and Demographic 31983 Implications of Stem and Joint Families WPS2128 Learning Outcomes and School Gladys Lopez Acevedo May 1999 M. Geller Cost-Effectiveness in Mexico: 85155 The PARE Program 1/-