Volume 15, Number 3

Extension in Sub-Saharan Africa: Overview and Assessment of Past and Current Models, and Future Prospects

Kristin E. Davis International Food Policy Research Institute Addis Ababa, Ethiopia [email protected]

Abstract This paper describes the role of agricultural extension in sub-Saharan Africa, and gives a typology for types of extension, which includes the basic forms of public top-down, participatory, and private. An overview of the evidence base for successes or failures of various models is given, which shows that evidence has been mixed on some of the major extension models in SSA, and that it is difficult to show impact for extension. There is also a lack of evidence on some of the newer models, extension reforms, and pluralistic models that involve many different extension providers. In general, though, problems in extension systems were due to a combination of a lack of relevant technology, failure by research and extension to understand and involve clientele in problem definition and solving, lack of incentives for extension agents, and weak linkages between extension, research, and farmers. The current status of extension in various sub-Saharan African countries is assessed, and new models are discussed. A framework for designing and analyzing extension systems is briefly described. Finally, future prospects for extension in sub-Saharan Africa are discussed.

Keywords: Extension, Sub-Saharan Africa, advisory services, impact, models

Fall 2008 15 Volume 15, Number 3

Introduction There are many definitions, including input suppliers, purchasers of philosophies, and approaches to agricultural agricultural products, training organizations, extension, and the views of what extension and media groups (Neuchâtel Group, 1999). is all about have changed over time. Extension originally was conceived as a Purpose and Objectives service to “extend” research-based The purpose of this paper is to give knowledge to the rural sector to improve the an overview of extension experience in Sub- lives of farmers. It thus included Saharan Africa (SSA), including a typology components of technology transfer, broader of extension models used in SSA and rural development goals, management skills, worldwide, evidence of the success or and non-formal education. The traditional failure of various models or extension in view of extension in Africa was very much general, assessment of the current status of focused on increasing production, improving extension and number of agents in selected yields, training farmers, and transferring African countries, and examples of technology. Today’s understanding of innovative models in use in SSA extension extension goes beyond technology transfer today. A framework for analyzing extension to facilitation; beyond training to learning, is discussed, as well as prospects for the and includes assisting farmer groups to future of extension in SSA. form, dealing with marketing issues, and partnering with a broad range of service Methods and Data Sources providers and other agencies. Thus many Data were obtained from document people are now using the phrase, analysis, the Internet, and from a simple “agricultural advisory services,” instead of survey on extension reforms sent out to extension (which can imply a top-down selected members of the Sub-Saharan Africa approach and may ignore multiple sources Network for Agricultural Advisory Services of knowledge). This paper will continue to (SSANAAS) (now known as the African use the term extension with the Forum for Agricultural Advisory Services). understanding that it encompasses the broader definition explained above. Results Agricultural extension can be defined as the entire set of organizations that Extension Typologies support and facilitate people engaged in There are many models and types of agricultural production to solve problems extension activities around the world, and and to obtain information, skills, and several authors have given typologies of technologies to improve their livelihoods extension, shown here for this illustrative and well-being (Birner, Davis, Pender, review (Table 1). This paper views Nkonya, Anandajayasekeram, Ekboir, et al., extension as generally, but not always, 2006). This can include different falling into three broad categories: diffusion governmental agencies (formerly the main or government-driven; participatory or actors in extension), non-governmental demand-driven; and private or supply- organizations (NGOs), producer driven, with the different systems or models organizations and other farmer falling under these three overall types. Many organizations, and private sector actors extension systems in SSA today are combinations of these broad categories.

16 Journal of International Agricultural and Extension Education Volume 15, Number 3 Table 1

Typologies of Extension by Various Scholars Rivera (1988) Axinn (1998) Gêmo et al. (2005) Top-down 1. Conventional 1. General agriculture 1. Public 2. Training and visit 2. Commodity 2. Commodity (T&V) 3. T&V 3. T&V 3. University 4. Agricultural 4. NGO 4. Technical innovation participatory approach 5. Private sector 5. Integrated agricultural 5. Project approach 6. Farmer field development program 6. Farming systems schools (FFS) Participatory 1. Farmer information research and extension dissemination system (FSR/E) 2. Farming system 7. Cost-sharing research-extension 8. Educational institute Contract 1. Commodity approach farming development 2. Commodity focused Rural 1. Community development development 2. Integrated rural development programs 3. Animation rurale

Evidence Base for Successes and Failures designs with control groups have been used of Extension Models in a few studies; however, such designs can Extension impacts per se are very be problematic. Other problems include the difficult to show, especially in terms of lack of baseline data and the inability to dealing with attribution issues and linking include all contributing variables in cause and effect quantitatively (Purcell & production equations. Studies on the training Anderson, 1997). Many infrastructural and visit system (T&V) in Burkina Faso and variables and other factors affect agricultural Kenya—which showed high returns to performance in complex and contradictory extension (Bindlish & Evenson, 1997)— ways and benefits are difficult to quantify were later criticized for data errors and (Anderson, 2007; Birkhaeuser, Evenson, & limitations due to the use of cross-sectional Feder, 1991). Measurement challenges of data (Gautam & Anderson, 1999). Thus, several types contribute to the difficulty, and some researchers are skeptical of rates of questions of representativeness occur in any return studies in Sub-Saharan Africa. attempt at grouping. In general, extension has been shown Other issues that may confound to have significant and positive effects on studies include endogeneity in program knowledge, adoption, and productivity. placement and extension-farmer Birkhaeuser et al. (1991) reviewed 48 interactions, selection bias, farmer-to-farmer studies of extension, and found that the information flow, and policies that affect majority (36) showed significantly positive various measures such as productivity. results. Studies of rates of return to Extension as an input is also difficult to extension (also in the same paper) generally measure, and usually proxies are used showed very high numbers (between 13- (Birkhaeuser et al., 1991). Experimental 500%). (But see the above critique of rates

Fall 2008 17 Volume 15, Number 3

of return studies.) This review covered a weaknesses were that they were supply- number of different types of extension driven, inflexible, disregarded many models, including T&V, but is somewhat institutions (including NGOs), were multi- dated. sectoral but not holistic, disregarded cost- Evenson (1997) reviewed 57 recovery or privatization, had an enclave economic impact studies, including seven mentality, and had limited sustainability African countries (the models used were (Anderson, 2002). either T&V or not noted). There was a wide According to a review by the World range of impact, from no significant Bank’s Operations Evaluation Department difference to highly significant differences of research and extension investments in the with regard to awareness, adoption, and 1980s and 1990s, three out of five extension productivity. The variability in results shows projects in Africa were “satisfactory,” which that some programs have been highly alluded to how fully the stated objectives effective while others have not been. were achieved (Purcell & Anderson, 1997). Finally, results showed that the highest T&V was seen as somewhat satisfactory in payoffs to extension occurred in developing Kenya and Somalia. However, Gautam countries that are catching up with (2000) later instituted a study in Kenya as a industrialized countries and with farmers result of disagreement on the initially who have access to schooling, technology, unsatisfactory rating (Gautam & Anderson and extension. (1999). In Zimbabwe, a national extension Taking into consideration some of and research project (not T&V) to improve the measurement concerns raised by productivity in rural areas via improved Birkhaeuser and colleagues, Owens, incentives for extension staff was as viewed Hoddinott, and Kinsey (2003) estimated the as satisfactory. T&V in Rwanda and Côte impact of access to extension services on d’Ivoire was deemed unsatisfactory. productivity in Zimbabwe by using T&V in particular, and public longitudinal data and controlling for innate extension systems in general, came under productivity using locality dummies, farm attack in the 1980s due to the cost of plot characteristics, and farmers’ ability. financing coupled with criticisms of They found that access to one or two visits irrelevance, inefficiency, ineffectiveness, per year from extension agents raised the and lack of equity (Rivera, 2001). In value of crop production by about 15%, a Ethiopia, Dejene (1989) found that the statistically significant parameter. communication system from contact farmers The Integrated Rural Development to the rest of the community did not work as Project (IRDP) approach was used in many expected, and up to 25% of contact farmers countries, including Kenya and Malawi, did not have the necessary knowledge and starting in the 1970s. IDRP projects (with skills. In Cameroon, Tchouama and Steele World Bank support) implemented an (1997) found that only 30% of respondents integrated extension approach. The IRDP’s had contact with the extension agent, and goals were to address constraints of furthermore had difficulty applying the smallholders by working synergistically in recommendations. In Nigeria, extension health, nutrition, agriculture, and education. agents lacked communication skills, In agriculture, this included inputs such as transportation, and faced cultural barriers extension, research, irrigation, credit, roads, (Asiabaka & Bamisile, 1992). water, electricity, and sometimes schools The Bindlish and Evenson (1997) and health centers. The focus was all study showed that the T&V management technical, however, and left out crucial system made extension more effective, led issues such as training, linkages with to agricultural growth, and realized high research, and management. IRDPs’ rates of return. However, in Kenya, Gautam 18 Journal of International Agricultural and Extension Education Volume 15, Number 3 (2000) later found that although T&V had benefits of the FFS approach, some studies some benefits in terms of staff training, have called into question their overall increased geographical coverage, and impact and financial sustainability. Farmer improved linkages with research, overall the field schools have shown remarkable impact system was inefficient, ineffective, and not in terms of pesticide reduction, increases in financially sustainable. productivity, knowledge gain among Initially promoted by the World farmers, and empowerment. However, these Bank, T&V has since been criticized within effects have been generally confined to the the Bank (Anderson, Feder, & Ganguly, most directly-engaged farmers, rather than 2006) and others, and is referred to tongue- demonstrating adequate capacity for scaling in-cheek by terms such as “talk and vanish” up for greater impact. The FFS themselves and “tragic and vain” (Axinn, 1988). Despite are undergoing reforms to address these the yield increases, the program was not issues, such as becoming self-financed sustainable, and left many countries saddled (Khisa, 2007). However, some studies do with huge debts. T&V has shown to be more show that FFS has limited or no effect on successful in Asia, where there is more economic performance, the environment and homogeneity within farming systems and human health, and farmer-to-farmer higher capacity among agents and farmers. dissemination of information and The T&V system was also more successful technologies. For a review of impact in promoting very specific packages (where evaluations of FFS, please see van den Berg they were suitable). However, the problem is and Jiggins (2007), who conclude that there that a tight management system and close was substantial immediate and control of “messages” does not constitute developmental impact for participation in relevance for the clients. FFS. These problems noted above have Several reviews have been conducted led to the reforms of many extension of Ethiopia’s Participatory Demonstration systems, including privatization, and Training Extension System decentralization, outsourcing, and (PADETES), based on Sasakawa Global participatory or demand-driven aspects. 2000’s (SG-2000) approach to extension Certain reforms have been formally and that uses demonstration plots and links informally evaluated. Problems encountered technologies to inputs through a package with decentralized extension included use of deal. Although 55% of respondents used the extension agents for non-extension package, a good number of farmers later purposes, lack of financial sustainability, abandoned package components such as and difficulties in linking to research fertilizer or improved seed (Bekele, (Anderson & Feder, 2004). In general, Anandajayasekeram, & Kisamba-Mugerwa, participatory reforms have been seen as 2006). Extension workers saw their role promising, while fee-for service has not mostly as distributors of fertilizer and credit been well taken up in the few low-income rather than technical advisors. Other countries where it has been attempted. researchers found that agricultural However, in general, most extension extension, as well as other rural services, reforms have yet to be evaluated as to their contributed significantly to agricultural effectiveness. productivity in Ethiopia (Ayele, Alemu, & Farmer field schools (FFS) (a model Kelemework, 2005). or an approach of extension education) have After initial (and favorable) been a recent topic of debate as to their qualitative reviews, Benin, Nkonya, Okecho, impact in SSA and elsewhere (Davis, 2006; Pender, Nahdy, Mugarura, et al. (2005) Gallagher, Braun, & Duveskog, in press). conducted a quantitative assessment of Although many positive reports exist on the Uganda’s decentralized, market-oriented,

Fall 2008 19 Volume 15, Number 3 farmer-centered National Agricultural was the need to significantly extend Advisory Services (NAADS) in 2005. The coverage (and quality) of extension. study showed that NAADS had positive In summary, this review shows that impacts on farm income and availability and evidence has been mixed on some of the quality of services. However, there was no major extension models in SSA, and that it significant difference in yield growth is difficult to show impact for extension. between NAADS and non-NAADS areas for There is also a lack of evidence of impact on most crops. At the same time, farmers in the some of the newer models, extension NAADS areas did show less decline in reforms, and pluralistic models that involve income than in other areas due to adverse many different extension providers. In climactic conditions during that time. general, though, problems in extension Shortage and timeliness of inputs were other systems were due to a combination of a lack problems in NAADS. A forthcoming piece of relevant technology, failure by research on NAADS shows that there is a strong and extension to understand and involve knowledge effect but that it does not clientele in problem definition and solving, translate into measurable productivity and lack of incentives for extension agents, and income effects (reviewed in Anderson, weak linkages among extension, research, 2007). and farmers. Kenya’s National Agriculture and Due to these complications, Livestock Extension Programme (NALEP) researchers at the International Food Policy was started in 2000. The NALEP approach, Research Institute put together a framework supported in part by Sida, focuses on for designing and analyzing extension stakeholder inclusion, bottom-up planning, (Birner, et al. 2006). The framework focuses and farmer common interest groups (CIGs) on (a) the design elements of a system of in focal areas. The first phase ran five years, extension—governance structures, capacity and was evaluated in 2006 (Cueller, and management, and advisory methods— Hedland, Mbai, & Mwangi, 2006). Data and their comparative advantages and were collected from project documents and disadvantages under different frame interviews with farmers. The analysis shows conditions; (b) performance measurement that 80% of respondents said that the and quality management in the provision of program offered new opportunities, and 70% agricultural advisory services; and (c) said that they viewed farming as a business impact assessment with regard to multiple as a result of NALEP. Regarding goals as well as assessment of the costs and sustainability, 70% of respondents claimed the benefits of different ways of providing that NALEP assisted them to gain profits and financing extension. The framework— from their farms. and indeed, this paper—call for a move from Finally, a study on the impact of “best practice”—imported standardized extension in Mozambique showed that models—to “best fit”—where location- public and private extension had a specific, participatory, sustainable, “smart” statistically significant positive effect on models are used. rural livelihoods (ECON Analysis, 2005). Extension in Mozambique mainly focuses Current Status of Extension in SSA on introducing new varieties, promoting The failure of many of these natural pesticides, and promoting extension models to meet their goals commercialization. The study showed that effectively, coupled with limited budgets for access to extension increased farm supporting public extension, has led to the production by 8.4%. Because only 13% of implementation of reforms in most SSA the rural population lived in villages with countries. Most African countries today are extension offices, one policy implication thus experimenting with reforms to existing 20 Journal of International Agricultural and Extension Education Volume 15, Number 3 extension systems. For instance, Ghana quality and safety, and environmental modified their extension system in 2003, conservation, among others (Rivera & Alex, based upon a 1997 policy to decentralize 2004). (Anderson, 2007). However, such According to Sasakawa Global 2000, modifications take a long time; hence, there there are about 150,000 extension workers is no information on its success. in Sub-Saharan Africa from the private, Existing models are typically a public, and civil society sector. However, general or modified T&V model housed in Swanson, Farmer, and Bahal (1990) the Ministry of Agriculture, although many reported that there are 58,958 extension countries are using multiple models with workers in Africa, based on a survey done pluralistic service providers (e.g., Nigeria; for FAO in the late 1980s. Data collected by see Oladele, Koyoma, & Sakagami (2004)). the author supplement this information Sasakawa Global 2000) works closely with (Table 2). extension and is currently working in It is apparent that little is known Ethiopia, Mali, Nigeria, and Uganda. The about the capacity, quality of service, and SG-2000 program first searches for a pool of performance of extensions systems in SSA. appropriate technology to be transferred and This type of information is urgently needed then works closely with the government to assess the strengths, weaknesses, and using national extension workers. performance of extension, and to strengthen These reforms are not changing the it to reduce rural poverty and to improve system used so much as the approaches rural livelihoods. Unfortunately, despite within the system. Reforms include use of calls for it, there has not been enough pluralistic extension providers and commitment to collecting these types of data approaches, decentralization/devolution, on a regular basis. privatization, contracting in and out, cost- sharing, demand-driven/participatory Innovative Extension Approaches approaches, fee-for service, and use of Following the above review of the information and communication current status of extension in SSA, we move technologies (ICTs). Qamar (2005) has now into details of specific innovative developed a guide for policy makers for approaches in use in SSA and elsewhere reforming extension systems. today. Several systems provide for Fee-for-service extension is provided competition among extension providers, for by the public (or another sector) and paid allowing for great accountability to farmers for by the farmers (Anderson & Feder, who can at times both hire and fire them. 2005). Small groups of farmers usually The NAADS approach has a certain contract the services. This arrangement competitive element, as do some FFS, where allows clientele to “vote” on programs and farmers have power to “fire” the extension program scale by paying for them (Hanson agent if they are not pleased with the & Just, 2001). Most of the examples of this service. model come from developed countries, such Many countries are committed to as New Zealand, which is completely participatory and pluralistic extension privatized. In addition to providing systems. How that takes place, however, is feedback, fee-for-service also can provide another matter. The increasing number of additional sources of revenue to public players and stakeholders makes the issues of extension. It is suitable for rival and coordination and regulation crucial, and excludable products. Hanson and Just argue underlines the need for the government to that universal paid extension is not in the remain involved in extension. This includes public interest, but that there is an optimal ensuring food security, regulating food mix of public, private, and paid extension. A

Fall 2008 21 Volume 15, Number 3

problem with this type of extension service farmers, allowing the commercial farmers to is that non-commercialized farmers may purchase services while smaller, poorer purchase fewer services (Anderson & Feder, farmers are serviced by public extension. 2005). One solution to this is to stratify

Table 2

Extension Models and Numbers of Agents in Selected SSA Countries Country Current Model(s) Angola Rural Development and Extension Programme; FFS Benin Participatory management approach; decentralized model; FFS Burkina Faso FFS Cameroon National Agricultural Extension and Research Program Support Project; FFS Ethiopia (65,000a) Model based on SG-2000 approach: Participatory Demonstration and Training Extension System; FFS Ghana Unified Extension System (modified T&V); pluralistic with NGOs and private companies part of the national extension system; decentralized; FFS Kenya Pluralistic system including public, private, NGOs; FFS; stakeholder approach (NALEP): sector-wide, focal area, demand-driven, group- based approach Malawi Pluralistic, demand-driven, decentralized; “one village one product;” FFS Mali Modified T&V; both private and parastatal services for cotton; FFS; SG- 2000 Mozambique (1,068) Government-led pluralistic extension; FFS Nigeria (5,252) FFS; participatory; SG-2000 Rwanda (500) Participative, pluralistic, specialized, bottom-up approach; FFS Senegal FFS; government-led demand-driven and pluralistic system; FFS Tanzania (7,000) FFS; group-based approach; SG-2000; modified FSRE from Sokoine University of Agriculture’s Centre for Sustainable Rural Development; private extension; decentralized Participatory District Extension; pluralism Uganda Pluralistic; National Agricultural Advisory Services (NAADS) is demand-driven, client-oriented, and farmer-led; SG-2000; FFS Zambia Participatory Extension Approach; FFS Note. aEthiopia has dramatically increased the number of extension agents from approximately 15,000 to about 65,000 in the past couple of years as part of a major capacity building effort (Ministry of Agriculture and Rural Development, 2007).

22 Journal of International Agricultural and Extension Education Volume 15, Number 3 In Uganda, the government has been Group members may also cover the cost of implementing the Plan for the travel of the extension staff. Modernization of Agriculture. One As with many models worldwide, component created in 2001 is the NAADS there has not been enough effort to provide program mentioned above, which has the hard evidence on the effectiveness of FFS. goal of increasing market-oriented Most FFS programs rely on ex post production through empowering farmers to evaluations, which are not able to provide demand and control extension services. rigorous results as to how the program NAADS is an innovative public-private compares to alternative programs or to the extension approach. The main components counterfactual situation of having no FFS. If of NAADS include decentralization, there are data, they often remain in project outsourcing, farmer empowerment, market documents and other grey literature, and the orientation, and cost-recovery (Anderson, information is not available to stakeholders 2007). who could provide peer review and Farmer field schools were validation of the methods and results. Future introduced into Sub-Saharan African in the studies would be aided by collecting mid-1990s. They are in place in at least 27 baseline data, obtaining panel data, and SSA countries (Braun, Jiggins, Roling, van using experimental designs. den Berg & Snijders, 2005). FFS came from A related concept to FFS is the Asia, where they were developed to promote farmer study circles. Study circles, much integrated pest management programs. In more informal than FFS, provide Africa, FFS are being used for a variety of opportunities for group exploration and activities, including food security, animal learning, to gain knowledge on whatever husbandry, and soil and water conservation. topic members decide. A group of people They are even moving beyond agriculture meet regularly, with no external “expert” into health (HIV/AIDS) and other relevant (although resource persons may be called in rural topics. or facilitators may guide the groups). Study FFS are a participatory method of circles allow a forum for people to learn and learning, technology development, and solve their own problems. The Swedish dissemination based on adult-learning Cooperative Centre focuses on human principles such as experiential learning. rights, improved livelihoods, and increased Groups of 20-25 farmers typically meet incomes, and has developed at least 68 weekly in an informal setting in their own different study circle guides in SSA for environment with a facilitator. The defining issues ranging from crops to HIV/AIDS characteristics of FFS include discovery (www.sccportal.org). To date, an inventory learning, farmer experimentation, and group has been conducted, but there have been no action. The approach is an interactive and studies looking at the effectiveness or practical method of training, and empowers impact of study circles (Torsten Andersson, farmers to be their own technical experts on personal communication, 19 April 2007). major aspects of their farming systems. Other innovative methods are related Farmers are facilitated to conduct their own to the rapidly-expanding information and research, diagnose and test problems, and communications technology sector. come up with solutions. Both to ensure Although ICTs are used in extension in sustainability and to enhance the sense of countries such as China, India, and Chile, ownership and responsibility, FFS programs Africa has lagged behind in harnessing ICT are encouraging cost sharing. In East Africa, potential for extension and other rural self-financed and semi-self-financed schools development issues. However, some are in place, and schools use commercial examples exist; for instance, in Kenya and plots to repay loans to operate the schools. Uganda, mobile phone services provide

Fall 2008 23 Volume 15, Number 3 cheap messages about crop price information can also be found in the information via text messaging. In Tanzania, Agriculture Investment Sourcebook’s there are “market spies,” farmers who visit Module 3 (World Bank, 2005). local markets and remain in contact with the village using mobile phones. Future Prospects for Extension An Indian decentralized market- in SSA driven extension model that may provide This paper has shown that there is a insights for extension in SSA is the wide variety of extension approaches and Agricultural Technology Management reforms in SSA today. Different programs Agency (ATMA) Model, an attempt to have diverse goals and thus differing increase farm income and rural employment strengths and weaknesses. For instance, (Singh, Swanson, & Singh, 2006). ATMA is T&V, although financially unsustainable, meant to integrate extension programs proved effective in training agents and across line departments, link research and improving the management of the overall extension, and use bottom-up planning system. The PADETES/SG-2000 type of procedures. Many judge it as a successful model in Ethiopia has proved effective in model of extension reform (Anderson, getting large numbers of farmers to adopt 2007). The authors outline four axioms technology packages for maize. Farmer field essential to market-driven extension. These school models have proven very effective at are (a) Don’t encourage farmers to produce strengthening farmers’ capacity and without a market; (b) Use products that are empowering rural people. easily transported; (c) Pay attention to agro- Based on the current status of ecological conditions for crops; and (d) extension in SSA, it appears that pluralism is Diversify crops to avoid saturation. the future of extension in SSA, with a There are also several innovative greater emphasis on demand-driven, approaches in financing extension services. participatory programs. Extension will have The creation of a Trust Fund (Ghana) and a greater focus on facilitation and access to Basket Funding (Tanzania) allows for the markets through farmer group formation and pooling of funds and distribution to end- ICTs. See Table 3 for various approaches users based on demand. In both cases, and where they may work the best. stakeholder forums consisting of farmer Because of the diversity of groups bring together concerns for required approaches (which mirrors the diversity in services from either public or private bodies. rural areas), extension agents in SSA Under the system, farmers are empowered to (public, private, and civil society) will need identify and use selected qualified service special skills that go beyond the basic providers (Government of Kenya, 2005). technical skills. Agents will need skills in Other potential methods include levies on group dynamics, marketing, and ICTs. More export commodities, community-driven than ever, he or she will need to be a skilled development funds (Guinea and Kenya), and technician who also is a broker of sorts, contracting by the government being able to connect farmers in their areas (Mozambique) (Alex, Byerlee, Helene- to markets and other institutions that are Collion, & Rivera, 2004). demanded by farmers. Furthermore, financing can come through decentralization, involvement of Conclusions farmers’ associations and NGOs, It is important to understand the experience contracting-out of extension services, of different countries and different extension public-private partnerships, privatization, models to develop and implement more and embedding advisory services in other effective models for Sub-Saharan Africa types of contracts (Anderson, 2007). More countries. There is very little information in 24 Journal of International Agricultural and Extension Education Volume 15, Number 3 the literature on current models and the performance and positive impact that number of public, private, and third sector policymakers are looking for in Sub-Saharan extension agents on the ground. Data are Africa. Promising models include the farmer also lacking to show the capacity and field school approach; the Indian ATMA performance of various extension systems. market-driven approach; and pluralistic, This paper attempts to document past and demand-driven models that incorporate the present extension models, reforms, and use of information and communication numbers of extension agents. It then gives technologies. prospects for what the future extension This paper has described the role of models and extension agents in SSA will be extension in Sub-Saharan Africa, and given like. a typology for types of extension, which Unfortunately, there is no “best includes the basic forms of public top-down, practice” for modifying extension programs, participatory, and private. An overview of a magic model that can be standardized and the evidence base for successes or failures of implemented anywhere. This had been tried various models was given. The current with Integrated Rural Development status of extension in various Sub-Saharan Programs, training and visit extension, and African countries was assessed, and new to a certain extent, farmer field schools. models were discussed. A framework for However, there are many good models with designing and analyzing extension systems useful features that, when implemented in a was briefly described. Finally, prospects for flexible, participatory, and sustainable the future of extension in SSA were (“smart”) way that meets the unique frame discussed. conditions of different countries and farming systems, can lead to improved extension

Table 3

Extension Approaches and Potentials for Success Approach Where does it work? Fee-for-service High potential areas; capable public, private, and civil society providers T&V Homogeneous areas; hierarchal and structured systems NAADS Available markets and market infrastructure; capable public, private, and civil society providers; decentralized systems; ability of farmers to pay; high social capital FFS High social capital; capable extension agents ATMA Available markets; capable research, extension, and other technical backstoppers; decentralized system; strong links between line departments PADETES Hierarchal and structured systems ICTs ICT infrastructure; enabling policy environment (e.g. low taxes on mobile phone usage) NALEP High social capital; available markets and market infrastructure

Fall 2008 25 Volume 15, Number 3

References Axinn, G. H. (1998). Guide on alternative Anderson, J. R. (2007). Agricultural extension approaches. Agricultural advisory services. Background paper Education and extension Service, for World Development Report 2008, Human Resources, Institutions and Agriculture for Development. Agrarian Reform Division. Rome: Washington, DC: The World Bank. Food and Agriculture Organization Alex, G., Byerlee, D., Helene-Collion, M., of the United Nations. & Rivera, W. (2004). Extension and Axinn, G. H. (1988). T&V (tragic and vain) rural development: Converging extension? INTERPAKS views on institutional approaches. EXCHANGE. International Agriculture and Rural Development Agriculture 5(3). Urbana: College of Discussion Paper 4. Washington, Agriculture, University of Illinois at DC: World Bank. Urbana-Champaign. Anderson, J. R. (2001). Relationship Ayele, G., Alemu, D., & Kelemework, F. between ecosystem health and (2005). The provisions of rural economic development: Rural vision services in Ethiopia: to action, In D. Rapport, W. Lasley, Characterization, impacts, and D. Rolston, C. Qualset and A. farmers’ priorities. Unpublished Damania (eds.), Managing for manuscript, International Food Healthy Ecosystems, 299-311. Boca Policy Research Institute, Raton: CRC/Lewis. Washington, DC. Anderson, J. R. (2002). Ecosystem health Bekele, E., Anandajayasekeram, P., & and economic development: Rural Kisamba-Mugerwa, W. (2006). vision to action. Rural Development Review of agricultural extension Department, World Bank, impacts in Ethiopia. Report prepared Washington, DC, U.S.A. for the World Bank. International Anderson, J. R., & Feder, G. (2004). Food Policy Research Institute: Agricultural extension: Good Addis Ababa, Ethiopia. intentions and hard realities. The Benin, S., Nkonya, E., Okecho, G., Pender, World Bank Research Observer J., Nahdy, S., Mugarura, S., Kato, E., 19(1), 41-60. & Kayobyo, G. (2007). Assessing the Anderson, J. R., Feder, G., & Ganguly, S. impact of the national agricultural (2006). The rise and fall of Training advisory services (NAADS) in the and Visit extension: An Asian mini- Uganda rural livelihoods. IFPRI drama with an African epilogue, in Discussion Paper 00724. A.W. Van den Ban and R. K. Washington, D.C.: International Samanta (eds.) Changing roles of Food Policy Research Institute. agricultural extension in Asian Bindlish, V., & Evenson, R. E. (1997). The nations, pp. 149-174. New Delhi: impact of T&V extension in Africa: B.R. Publishing Corporation. The experience of Kenya and Asiabaka, C. C., & Bamisile, A. I. (1992). Burkina Faso. The World Bank Evaluation of constraints affecting Research Observer 12(2), 183-201. the performance of extension agents Birkhaeuser, D., Evenson, R. E., & Feder, in the Lagos State Agricultural G. (1991). The economic impact of Development Project in Nigeria. agricultural extension: A Review. Tropical Science 32, 421-428. Economic Development and Cultural Change 39(3), 607-640.

26 Journal of International Agricultural and Extension Education Volume 15, Number 3 Birner, R., Davis, K., Pender, J., Nkonya, E., Feder, G., Willet, A., & Zijp, W. (1999). Anandajayasekeram, P., Ekboir, J., Agricultural extension—generic Mbabu, A., Spielman, D. J., Horna, challenges and some ingredients for D., Benin, S., & Kisamba-Mugerwa, solutions. The World Bank, W. (2006). From best practice to Washington, DC. best fit: A framework for designing Gallagher, K. D., Braun, A. R., & and analyzing agricultural advisory Duveskog, D. (in press). services. ISNAR Discussion Paper Demystifying farmer field school No. 5. Washington, D.C.: IFPRI. concepts. Manuscript submitted for Braun, A., Jiggins, J., Roling, N., van den publication. Berg, H., & Snijders, P. (2005). A Gautam, M. (2000). Agricultural extension: global survey and review of farmer The Kenya experience. An impact field school experiences. Report evaluation. Washington, DC: The prepared for the International Food World Bank. Policy Research Institute. Gautam, M., & Anderson, J. R. (1999). Wageningen: Endelea. Reconsidering the evidence on Cueller, M., Hedland, H., Mbai, J., & returns to T&V Extension in Kenya. Mwangi, J. (2006). The National Policy Research Working Paper Agriculture and Livestock Extension (WPS2098). Programme (NALEP) Phase I impact Gêmo, H., Eicher, C. K., & Teclemariam, S. assessment. Sida Evaluation 06/31. (2005). Mozambique's experience in Stockholm: Swedish International building a national extension system. Development Cooperation Agency. East Lansing, MI: Michigan State Davis, K. (2006). Farmer field schools: A University Press. boon or a bust for extension in Government of Kenya. (2005). National Africa? Journal of International agricultural sector extension policy Agricultural and Extension (NASEP). Nairobi: Ministry of Education 13(1), 91-97. Agriculture; Ministry of Livestock Dejene, A. (1989). The training and visit and Fisheries Development and agricultural extension in rainfed Ministry of Cooperative agriculture: Lessons from Ethiopia. Development and Marketing. World Development 17(10), 1674- Hanson, J. C., & Just, R. E. (2001). The 1659. potential for transition to paid ECON Analysis. (2005). Impacts of extension: Some guiding economic Extension Services in Rural principles. American Journal of Mozambique. Consultant report for Agricultural Economics 83(3), 777- the 2005 Rural Development 784. Strategy. Washington D.C.: World Khisa, G. (2007). Farmer field school Bank. experience in Kenya. Presentation at Evenson, R. (1997). The economic the IFAD TAG Regional Meeting, 8- contributions of agricultural 10 October, Bukoba, Tanzania. extension to agricultural and rural MAAIF/NAADS. (2004). First networking development. In Improving symposium on innovations in agricultural extension: A reference agricultural advisory services in sub- manual, ed. B. E. Swanson, R. P. Saharan Africa. 11-14 October, Bentz, & A. J. Sofranko. Rome: Kampala, Uganda. Food and Agricultural Organization of the United Nations.

Fall 2008 27 Volume 15, Number 3

Ministry of Agriculture and Rural Purcell, D. L., & Anderson, J. R. (1997). Development. (2007). RCBP (Rural Agricultural extension and research: Capacity Building Project). Draft Achievements and problems in Project Implementation Manual. national systems. Washington, DC: Neuchâtel Group. (1999). Common The World Bank. framework on agricultural extension. Qamar, K. M. (2005). Modernizing national Paris: Bureau des Politiques agricultural extension systems: A Agricoles et de la Sécurité practical guide for policy-makers of Alimentaire. developing countries. Rome: Food Odale, I. O., Koyoma, O., & Sakagami, J. I. and Agriculture Organization of the (2004). Africa in search of extension United Nations. system: Experience from Nigeria. Rivera, W. M. (1988). Developing Food, Agriculture & Environment Agricultural Extension Systems 2(1), 276-280. Nationwide: A Structural Approach. Owens, T., Hoddinott, J., & Kinsey, B. Journal of Extension Studies, 4(2). (2003). The impact of agricultural extension on farm production in resettlement areas of Zimbabwe. Economic Development and Cultural Change 51(2), 337-357.

28 Journal of International Agricultural and Extension Education