Main actors within the extension system are the members of rural communities, extension and other development personnel, researchers, and staff of commercial or public service and support organizations. Empirical evidence shows a variety of forms in which interaction among these groups is institutionalized. The variety of forms suggests a similar variety of goals, and either could be used to classify extension approaches. In practice, however, one finds an almost inseparable mixture of goals inhibiting a clear-cut classification. It seems more appropriate, then, to use a broader category, namely, selectivity with regard to clientele, and treat the respective goals as a continuum. The two end points of this continuum would be marked as technology transfer and human resource development, suggesting either a rather narrow technical or a broader socioeconomic view of development.
In Sub-Saharan Africa agricultural extension work started during the early years of the 20th century by the European colonial powers. The principal objective of the extension services during the colonial era was to generate revenue from and meet the colonial demands for a steady supply of exportable tropical agricultural products, such as coffee, cocoa, tea, cotton, sugar and rubber. Accordingly, the extension system was commodity-oriented. In evaluating the commodity-based extension system, Birmingham (1999), notes that it often neglected the needs of food crops and livestock, and overlooked the effects of cash crop labor demands on household members particularly women and children. Moreover, it paid little attention to the production constraints faced by subsistence farmers. In the early 1960s, when many of the countries of Sub-Saharan Africa gained their independence, the financing and management of agricultural extension work became a largely national, often government responsibility.
The main focus of agricultural extension work in the post-independence period was to increase agricultural (mainly food) production and spread the benefits of improved farming techniques more widely (Picciotto and Anderson, 1997; Anderson, 2007). According to Nagel (1997), in many developing countries the transfer of technology (ToT) model has been the prevalent practice for developing and spreading innovations. Chambers (1993) defines the ToT model as the basic paradigm of agricultural research and extension in which priorities are decided by scientists and funding agencies, and new technologies are developed on research stations and in laboratories and then handed over to extension to be transferred to farmers. The ToT model was based on the assumption that new agricultural technologies and knowledge are typically developed and validated by research scientists, and that the task of extension agencies is to promote the adoption of these technologies by farmers, thereby increasing agricultural productivity.
In terms of institutional arrangements and relationships, the ToT model creates a rigid hierarchy, which discourages feedback from the users of the technology. More precisely, the ToT model was predominantly a one way communication in which messages channeled from top to bottom through the hierarchical structure, most often from government sources to the farmers. Message intent was to "inform" and "persuade" farmers to adopt technologies and practices, which had been developed by the "experts" (Haug, 1998; Botha and Stevens, 1999; Anderson, 2007). The traditional technology generation and transfer model fostered a commodity-focused and discipline-oriented research approach and it relied heavily on researchers’ ability to identify relevant problems and generate widely-useable technologies. This type of research is mostly conducted at research stations under the conditions that are mot representative of farmers’ fields, with little or no farmer involvement. More precisely, researchers work independently of farmers and extension workers, resulting in a poor understanding of farmers and the opportunities and constraints they face. The ToT approach is fragmented, both institutionally and in terms of disciplines. Research concentrates on technology and researchers and extensionists are seen as technical agents. Social competence is not required as complex socio-organizational issues (eg. land-use regulations, power structures, conflict resolution mechanisms) are neglected or reduced to a technical level (Hagmann et al., 1999). Under the ToT model, it is believed that the most innovative farmers adopt the technology first and the remaining farmers eventually follow. The approach reduces the farmers to simple adopters of technologies developed by others. The role of the extension agents is limited to teaching farmers, via various extension methods such as farm visits, demonstrations, group training sessions etc., and put the ready-made technologies into practice
The ToT model has been the official approach to agricultural research and extension in most African countries. However, available evidence shows that in many of the Sub-Saharan Africa countries smallholders are characterized by poor adoption of technologies. Earlier studies attribute the low rate of adoption of agricultural technologies to the absence of "smallholder-friendly" research findings (Wiggins, 1986; Röling, 1988; Jiggins, 1989; Chambers, 1993; Christoplos, 1996; Pretty and Chambers, 1993; Hagmann et al., 1999). However, in the ToT model, the non adoption of recommendations has very often been attributed first to "farmers' ignorance" or "resistance to change", to be overcome through more and better extension services, and later to farm level constraints, with solutions in easing constraints to make the farm more like the research station (Chambers and Ghildyal, 1985; Chambers, 1993). The fact that such technologies either do not fit to the production systems and ecological conditions or that they require resources that small farmers do not have access to have rarely been recognised by researchers.
In analyzing the roles and challenges of agricultural extension in Africa, Opion-Odongo (2000) argues that extension workers in Sub-Saharan Africa have behaved as if the farmers can only benefit from innovations that are external to their farming systems. He further notes that extension workers have tended to treat farmers as if they were empty vessels to be filled with knowledge and expertise. Similarly, Wiggins (1986) argued that research stations in Africa have tended to develop ideas with too little attention to smallholder labor supplies, to the riskiness of the innovations, to the likely availability of inputs, or to the presence of markets and to the economic attractiveness of recommendations. In the same vein, Röling (1988) suggests that the technology generation process generally suffers from lack of social expertise and accountability to rural clients. As a result, agricultural technologies have, in some cases, accelerated the twin processes of resource degradation and selective impoverishment of women and the poor.
The conventional extension system that is heavily influenced by the transfer of technology paradigm considered farmers as a homogeneous mass and thus failed to categorize them into different groups with different resources, problems, opportunities and requirements. As a result, it could not select appropriate technologies and tailor them to the specific groups of farming populations (Röling, 1988; Jiggins, 1989; Chambers, 1993; Christoplos, 1996; Van de Fliert, 2003). Rather, extension agents promoted blank recommendations that may at best be worthy but in need of adaptation to local conditions, and at worst are useless. Moreover, the ToT model fails to recognize farmers’ indigenous knowledge as one of the most important elements in designing and providing effective and efficient agricultural extension services (Vanclay and Lawrence, 1995; Haug, 1998; Black, 2000; Murray, 2000; Van de Fliert, 2003). In this respect, Vanclay and Lawrence (1995) argue that in the traditional extension model indigenous technical knowledge is marginalized, trivialized, subordinated or ignored; there is a ‘hegemony of technocratic discourse’.
As discussed earlier, the ToT model viewed farmers, extensionists and researchers as three separate strata and the links between them have been weak or non-existent. In this connection, earlier empirical studies in developing countries have identified weak links between research and extension as the major factor limiting the flow of information, knowledge, useful new technologies, and resources among actors in the technology-delivery-utilization system and recommended measures to overcome the widely acknowledged weaknesses (Purcell and Anderson, 1997; Agbamu, 2000; Asiabaka, 2002; Belay, 2002, 2003; Van de Fliert, 2003; Anderson and Feder, 2004). Available empirical evidence shows that the ‘top-down’ model of technology development and transfer has resulted in inequalities in benefits to farmers. In this connection, Röling (1988), notes that the application of ‘top-down’ adoption/ diffusion approach has tended to reinforce existing social inequalities within the farming population, since the producers benefiting most from the adoption process have generally been those better endowed than others in material, intellectual and social resources. Similarly, the ToT model was reported to have primarily benefited those farmers in low-risk, natural and social environments, who happened to fulfil the necessary requirements of adoption, including availability of or access to capital, labour and fertile land (Farrington, 1988; Jiggins, 1989; Tripp, 1991; Norman et al., 1995; Martin and Sherington, 1997). The same authors underline that the model essentially failed to bring about perceptible changes in marginal areas, with low agricultural potential, where farmers are commonly faced with a range of adverse agro-ecological, social and economic conditions, including unreliable rainfall, low-fertility soils, fluctuating market prices for agricultural products, and labour shortage.
Approaches Promoting Farmers' Involvement in Agricultural Research and Extension
In the early 1970s, the ToT model came under heavy criticism. More specifically, the sovereign role of researchers in diagnosing problems, developing hypotheses and designing a research process as well as the notion that extension means "to advise" farmers were seriously challenged. This situation has contributed to the emergence of different approaches to technology generation and transfer, which put emphasis on recognition of local knowledge and farmers as rational decision makers and interactive partners in research and extension activities. This shift from "top-down" to "bottom-up" approach was based on ample empirical evidence that pointed to the fact that non-adoption of technologies by smallholder farmers emanated from the fact that the technologies in question had been either unresponsive or inappropriate to the social, physical and economic setting in which the target population had to operate and as a result had not provided directly measurable results or perceived benefits. Consequently, it was emphasized that the whole process of technology development and dissemination must be based on equal partnership between farmers, researchers and extension agents who learn from each other and contribute their knowledge and skills. The different approaches which promote the involvement of farmers in agricultural research and extension are briefly described below.
The Farming Systems Research and Extension
During the 1970s there was growing recognition that agricultural research had primarily benefited resource-rich farmers and that the main reason why resource-poor farmers had been slow or unable to adopt recommendations was that the technologies were not sufficiently relevant to their circumstances and their environment (Farrington, 1988; Jiggins, 1989; Chambers, 1993; Farrington, 1994; Christoplos, 1996). It is also widely accepted that resource-poor farmers tend to live and operate in complex and diverse environments and that there is a high degree of heterogeneity in their circumstances and farming systems (Jiggins, 1989; Chambers, 1993; Gibbon, 1994; Norman et al., 1995). Earlier empirical studies undertaken in developing countries indicated that because of the complexity of the farming systems, resource-poor farmers’ main constraints varied widely and standard technology packages promoted by national agricultural extension systems had not been adopted by farmers, or where they had been, they exhibited variations in degrees and levels of adoption (Röling, 1988; Jiggins, 1989; Chambers, 1993; Norman et al., 1995; Van de Fliert, 2003). This recognition contributed to the emergence of the Farming Systems Research/Extension (FSR/E) approach in the 1970s. The crux of the FSR/E approach is to develop and transfer appropriate agricultural technologies based upon a clear understanding of the existing farming systems. The adoption, adaptation, and evolution of this approach have been highly variable across countries. Farmer participation is an important concept in FSR/E and this approach identifies homogeneous groups of farmers within specific agro climatic zones as the clients of research. The FSR/E approach focuses on the farm household as the principal unit of analysis and it gives more attention to interactions than to components of the farm system (Gibbon, 1994). Using appropriate methodological tools and working within a holistic framework, an interdisciplinary FSR/E team would identify existing farming systems and seek to understand them and their environment. They would identify problems and opportunities for improving the system, the environment or both, and set priorities for research and implementation (Jiggins, 1989; Gibbon, 1994; Norman et al., 1995). In the FSR/E approach, on-farm trials are carried out and the results of one year’s trials generate hypothesis for testing in the next and should influence on-station research priorities. FSR/E rapidly became popular and was strongly supported by many donor agencies. In the FSR/E approach, extension would not be restricted to dissemination functions, but would have a key role in feeding information on what innovations farmers adopted or rejected (and how and why) back into the research system (Farrington, 2002).
Since its appearance, FSR/E has been taken up in various forms by international and national agricultural systems, and donor supported agricultural development programmes and projects. The fact that the different International Agricultural Research Centres (IARCs) using FSR/E have adopted a completely different labels (e.g. on-farm research by the Centro Internacional de Mejoramiento de Maiz y Trigo or CIMMYT, Farmer-back-to-farmer by the Centro Internacional de la Papa or CIP, cropping systems research by the International Rice Research Institute or IRRI, etc.) that only slightly differ in methods have added to the confusion through a profusion of vocabulary (Tripp, 1991; Gibbon, 1994).
A major contribution of the FSR/E approach to agricultural research and extension work has been its recognition of farmers’ knowledge, circumstances, objectives and constraints as most crucial factors in adopting/ adapting technologies. Furthermore, the FSR/E approach has resulted in widespread understanding that farmers’ participation in the process of research is extremely important to better understand their complex situation and the inter-dependencies among elements of farming systems and thereby adapt technologies to site-specific circumstances (Gibbon, 1994; Scoones and Thompson, 1994; Norman et al., 1995; Haug, 1998; Farrington, 2002). It must also be noted that in addition to the development of specific field research methods, FSR/E has contributed to changes in the management and targeting of agricultural research (Tripp, 1991).
Recently, however, FSR/E is being challenged by the growing interest in sustainable agriculture, and high hopes by some donors on the private sector and the NGOs as important forces for development. In addition some criticise FSR/E arguing that in FSR/E policy issues are put aside, participation of farmers is low, the costs of placing FSR/E teams in the field for long periods are high, gender issues are marginally addressed and despite years of work there are less evidence of success in the hands of farmers (Tripp, 1991; Gibbon, 1994; Biggs, 1995). As noted earlier, the successful implementation of FSR/E programs demands an effective linkage among the research, extension and farmers, based on mutual trust and respect. However, experience from different parts of the world where FSR/E programs have been implemented points to the fact that the research-extension-farmers linkage has been generally weak. In this respect, the relevant empirical literature indicates that even though FSR/E was a good attempt to understand farmers' knowledge and constraints, it essentially remained extractive as researchers are in a dominant position and farmers are only peripherally involved in problem diagnosis, setting the agenda and priorities for research and extension, testing of on-station research findings and providing feedback (Gibbon, 1994; Scoones and Thompson, 1994; Biggs, 1995; Purcell and Anderson, 1997). Likewise, Farrington and Martin, (1987), argue that FSR/E is still dominated by the ToT approach, does not focus on the problems and potentials of the resource poor farmers nor seeks to find unconventional technological alternatives taking into account the complexity and diversity of prevailing agricultural systems. Another shortcoming of the FSR/E programs has been the tendency for researchers and their immediate collaborators to imagine that they can quickly develop relevant research outputs that will positively benefit the lives of many rural people. In this respect, Gibbon (1994), notes that researchers are only one group of actors within rural areas and it is unrealistic to think that by their actions alone they can have a profound effect on rural livelihoods. Finally, it is worth noting that while much of the philosophy of the FSR/E approach remains relevant, in practice, it has been plagued by the inability to shift "recommendation domains" onto a large scale, and resistance by many scientists to be drawn out of conventional modes of operation (Farrington, 2002)  .
Critics of FSR/E stressed the need for more involvement of farmers to better address the needs of the disadvantaged, to build on farmers' knowledge and experimentation, and to easily adapt technologies to complex farming situations, and ultimately for rapid uptake and diffusion (Martin and Sherington, 1997). The suggested models for more involvement of farmers include farmer-back-to-farmer (Rhoades and Booth, 1982), farmer-first-and-last (Chambers and Ghildyal, 1985), farmer participatory research (Farrington and Martin, 1987), and Participatory Technology Development (ILEIA, 1989). These are collectively referred to as Farmer First approaches. The concept of "Farmer First" has been popularized since the international workshop on "Farmers and Agricultural Research: Complementary Methods" held at the Institute of Development Studies in Sussex, UK, between 26 and 31 July, 1987. Out of the workshop was produced a popular book, entitled Farmer First: Farmer Innovation and Agricultural Research in 1989, which was co-edited by Robert Chambers, Arnold Pacey, and Lori Ann Thrupp and published by Intermediate Technology Publications, London.
Farmer First seems to have been widely used to designate concepts and approaches that focus more to alternative methods for ensuring effective farmer participation in research process (Norman et al., 1988). Farmer first approaches stress the abilities of resource-poor farmers to experiment, adapt and innovate as well as the importance of giving priority to farmers’ agendas and knowledge and suggest the need for generating baskets of choices to enable farmers to vary, complicate and diversify their farming systems (Chambers, 1994). The farmer first paradigm demands major changes in the attitudes, approaches and roles of researchers and extensionists (Chambers and Ghildyal, 1985). In the late 1980s, farmer first methods evolved and the need for going beyond field participation to influence decisions and methods of on-station research were stressed. With this development in the early 1990s, came the concept of empowerment and the frequent use of the term Farmer Participatory Research.
Farmer Participatory Research
Since the late 1980s, policy makers, academia and development agencies have recognized the central role of farmers in the technology development and transfer process. As a result, they have been substantiating that the whole process of technology identification, development and transfer must shift from a ‘top-down’ conveyor belt system towards one in which research and extension systems become more demand-driven, customized to local conditions and needs and responsive to farmers’ pressing problems. Consistent with this view and drawing on past experience in agricultural development work, in recent years, in many developing countries, research and extension systems have been embracing participatory approaches in which farmers’ priorities and capacities are key inputs. In this regard, the relevant empirical literature shows that, in many developing countries, the relation between research and extension systems has become increasingly a two-way process and farmers, who are key stakeholders in the development and dissemination of agricultural technologies, have become the target and the hub around which researchers and extensionists focus their actions (Chambers, 1994; Purcell and Anderson, 1997; Agbamu, 2000; Asiabaka, 2002; Anderson and Feder, 2004; Anderson, 2007). More precisely, farmers have found their place in this link-chain mechanism through which they can articulate their problems and needs and influence research and extension priorities.
The Farmer Participatory Research (FPR) approach includes recognition of farmers as experimenters who continuously conduct their own trials, and who partially adopt and adapt technologies to their own particular circumstances and spread innovations through their own networks (Haug, 1998). Moreover, this approach implies that researchers and extension professionals accord credence to farmers’ knowledge and advice and that farmers are recognized as partners in research and extension, and key players in the innovation process (Farrington, 1994; Cornwall and Jewkes, 1995; Hagmann et al., 1999; Birner et al., 2006). In other words, farmers become less like objects to be studied and more like active subjects influencing the direction and methods of the research. The basic premise underlying the FPR approach is that greater involvement of farmers in identifying needs-related research priorities and in exploring solutions to problems they have prioritized ensures that the ensuing interventions are appropriate, more effective and sustainable. According to Okali et al. (1994), FPR is understood by many to be one element of a larger 'participatory' development agenda which aims not only to generate, test and disseminate technologies, but also to change the orientation of existing research and development structures, develop a sustainable, community-based research capability, and create new social and political institutions. In the same vein, Martin and Sherington (1997), note that participatory research approaches are characterized by a concern with the need for empowerment of resource-poor farmers and a focus on local knowledge and management capacity. Likewise, Cornwall and Jewkes (1995), suggest that participatory research is about developing a realization that local people are knowledgeable and that they, together with researchers, can work towards analyses and solutions. It involves recognizing the rights of those whom research concerns, enabling people to set their own agendas for research and development and so giving them ownership over the process. FPR is not a substitute for station-based research or researcher-managed on-farm trials. It is a complementary process that involves linking the power and capacities of agricultural science to the priorities and capacities of farming communities, to develop productive and sustainable farming systems (Douglas and Lawrence, 1997). FPR not only seeks to generate technologies adapted to local environments, but also seeks to develop the local capacities, socio-cultural structures and organizational linkages necessary to sustain the process.
FPR calls for a more radical change in relations among researchers, extension workers and farmers. More precisely, interdependence among all these actors prevents isolation, which impedes technology transfer. Close bonding among the three key players also promotes development of relevant technologies that provide directly measurable results or perceived benefits to the target population and adapted to local conditions. One of the salient features of the FPR approach is the reversal learning, where researchers and extension workers are learning from farmers (Asiabaka, 2002). This implies that changes in values, attitudes and behavior in order to ensure that significant learning takes place among all actors (Röling and Pretty, 1997). For extension workers, this means learning to interact closely with farmers and becoming learners, catalysts, and facilitators who help farmers achieve their defined and perceived goals. In FPR, farmers and extension workers develop a research agenda together. The role of researchers is to take up the questions identified by farmers and extension workers and work from there (Hagmann et al., 1999). With the exception of some basic research, most studies are carried out on-farm in an interactive way in order to find appropriate solutions to farmers’ problems.
According to Cornwall and Jewkes (1995), the key element of participatory research lies not in the methods but in the attitudes of the researchers, which in turn determine how, by and for whom research is conceptualized, and conducted. In participatory research approaches, locations and roles are reversed, with farms and farmers as being central, instead of research stations, laboratories and scientists (Asiabaka, 2002). More specifically, with the exception of some basic research, most studies can be carried out on-farm in an interactive way in order to find appropriate solutions to farmers’ problems. Cornwall and Jewkes (1995), argue that participatory research offers ways of making conventional science more relevant, by creating an environment in which new knowledge can be synthesized through a dialogue between scientific and local knowledge. However, a review of the relevant literature indicates the existence of fundamental differences between participatory research and conventional research approaches. For instance, Okali et al. (1994), note that the most important difference between FSR/E and FPR is the focus of the latter on the value and development potential of farmers and its deliberate attempt to actively involve farmers in designing, monitoring and evaluating experiments, in collaboration with researchers, carried out in their own fields. FPR is reported to have lessened the lead time for introducing new technologies to their subsequent learning and ultimate adoption (Adalla and Hoque, 1989). It is also noted that FPR calls for a more holistic "food systems" perspective as opposed to the narrower "farming systems" approach for understanding rural people's capacities and constraints (Scoones and Thompson, 1994). One of the key strengths of participatory research methodologies is seen to reside in exploring local knowledge and perceptions. Moreover, participatory research methodologies are often characterized as being reflexive, flexible and iterative, in contrast with the rigid linear designs of most conventional sciences (Cornwall and Jewkes, 1995).
As an approach, FPR is becoming steadily well established. While it is generally accepted that user participation at all stages of a research process provides useful feedback to researchers that improve the relevance and appropriateness of research results and contributes to actual or potential impact of the research, the application of FPR approach in developing countries had little impact on the ground. For instance, Okali et al. (1994), underline that there exists a gap between intellectual discussion and field reality. Despite the rhetoric of participation, much of the activity of FPR programmes and projects comes down to controlled experimental or test plots on which crop varieties or production techniques are demonstrated and evaluated, which resemble the on-farm trials that are the hallmarks of FSR. Therefore, as Cornwall and Jewkes (1995), suggest in FPR programmes the practical implementation of the empowerment agenda is very limited and the control over research is rarely devolved completely onto users or beneficiaries of the research results. In this respect, Chambers (1994), argues that it is only when practitioners of participatory approaches develop and spread new methods and approaches that create environments for an innovative and co-learning process that they will manage to gain full participation of farmers. When farmers are involved as equal partners, they play a pivotal role that makes agricultural research and extension effective so that it would significantly contribute to the development of agriculture at large. It is, therefore, imperative that researchers move away from reductionism, linear thinking and standard solutions in favour of more inclusive holism, open systems thinking and methodological pluralism that encourage participation and better serve the needy farming population. This move towards participatory approaches is, however, faced with some attitudinal, institutional and methodological challenges.
Participatory Extension Approaches
In many countries, agricultural extension is being reoriented to provide more demand-based and sustainable services, taking into account the diversity, perceptions, knowledge, and resources of users. The new farmer-centered approach to extension, participatory extension approach (PEA), calls for a bottom up approach of planning, implementation and evaluation of extension activities. PEA is broadly defined as a multi-directional communication process between and among extension staff and farmers, involving the sharing, sourcing and development of knowledge and skills, in order to meet farmers’ needs and develop innovative capacity among all actors, in which farmers have a controlling interest (Chandrapatya, 2002).
PEA is based on the premise that the effectiveness of agricultural extension work can be improved if local knowledge and resources are tapped to both diagnose problems and experiment with solutions (when researchers, extension staff and farmers become like partners in technology generation and dissemination). In participatory extension, it is assumed that farming people have much wisdom regarding their environment, but their living standards could be improved by learning more of what is known outside (which they do not normally know), that effective extension cannot be achieved without the active participation of the farmers themselves as well as of research and related services, that there is a reinforcing effect in group learning and group action, and that extension efficiency is gained by focusing on important points based on expressed needs of farmers through their groups or organizations instead of through individualized approaches (World Bank, 1995; Röling and Pretty,1997; Hagmann, et al., 1999; Van den Ban, 2005).
According to Hagmann, et al. (1998), if participatory extension approaches are institutionalized in extension organizations, they can help to improve organizational performance at the interface between the service providers (the extensionists) and the clients (the farmers). The same authors summarized the characteristics of participatory extension approaches as follows:
• they integrate community mobilization for planning and action with rural development, agricultural extension and research;
• they are based on an equal partnership between farmers, researchers and extension agents who can all learn from each other and contribute their knowledge and skills;
• they aim to strengthen rural people's problem-solving, planning and management abilities;
• they promote farmers' capacity to adapt and develop new and appropriate technologies / innovations;
• they encourage smallholder farmers to learn through experimentation, building on their own knowledge and practices and blending them with new ideas. This takes place in a cycle of action and reflection which is called "action learning"  ;
• they recognize that communities are not homogenous but consist of various social groups with conflicts and differences in interests, power and capabilities. The goal is to achieve equitable and sustainable development through the negotiation of interests among these groups and by providing space for the poor and marginalized in collective decision-making.
The role of the extension agent is to facilitate this process. While farmers become influential and responsible clients rather than passive beneficiaries of the extension service in the PEA, researchers assist farmers and extension agents in the joint experimentation and learning process and contribute their knowledge of technical options to find solutions to the problems identified by farmers (Nagel, 1997; Hagmann, et al., 1998; Anderson, 2007). In other words, when PEA is used, researchers, extension workers and farmers have the opportunity to work together, exchange knowledge and experiences and reach some consensus on what is most needed. In sum, by making extension more demand driven, and more accountable to farmers, participatory approaches can help to ensure that services are relevant and responsive to local conditions and meet the real needs of users (World Bank, 1995).
PEA has profound implications for agricultural development institutions in that it defines new roles for key actors engaged in agriculture, leading to a whole new professionalism with new values, methods, and behavior. For instance, PEA requires a major shift in roles of agricultural extension workers from teacher to facilitator. In other words, the role of the extension worker changes from the old stereotype of connecting researchers and farmers - to being innovative, providing the methodology for the process, facilitating communication and information flow, and providing the technical backup and options (Christoplos, 1996; Röling and Pretty,1997; Farrington, 2002; Birner et al., 2006; Kahan, 2007).
Most extension field staff in Africa are ill-prepared to assume the emerging roles of extension workers under the PEA, which include, among others: developing farmers’ own capacity to think for themselves and develop their own solution; coordinating and organizing knowledge acquisition from several sources; serving as a link to the world outside the village; reaching marginalized, resource-poor and women farmers; applying technical knowledge to site-specific socioeconomic and agronomic conditions; and feeding back information on farmers’ constraints and potentials to encourage relevant research. These new roles call for a fundamental reorientation of the existing field staff to enable them acquire skills of facilitation, technical know-how, negotiation, conflict resolution and nurturing community organizations. Reorientation programs are extremely important to turn the existing inadequately trained, mostly underpaid and overworked frontline field staff, who prescribe "prepackaged messages", into technically competent, highly qualified, effective and more client-oriented extension workers. In the long run, however, there is a need to look much more to agricultural colleges/faculties where future extensionists are trained. The agricultural education system in many African countries is dominated by the reductionist production oriented science with its deep rooted positivistic philosophy. The attitudes and practices of research and extension workers are largely shaped by the training they get from the education system (Belay, 2008). The reality on the ground requires that agricultural professionals must be trained in areas beyond technical agriculture that enable them to perform a wide-ranging role expected of them. This clearly indicates that both the curricula and training approaches in agricultural colleges/faculties need to be reformed. In this regard, among the most important reform measures that must be undertaken by agricultural colleges/faculties are: updating the current curricula; transforming teaching practices; developing communication and facilitation skills; promoting dynamic interplay between theoretical and practical components; and enabling learners to be reflective practitioners. In this connection, the experience of Haramaya University (Ethiopia) in successfully running an innovative Bachelors Degree program for mid-career Agricultural Extension professionals, in close collaboration with the Sasakawa Africa Fund for Extension Education, since 1997 is a move in the right direction.
Participatory extension aims at giving farmers a maximum role in developing technologies that work and in spreading successfully tested technologies to other interested farmers. PEA uses participatory methods that are based on flexible use and continual adaptation to the situation of participatory tools and techniques to initiate and guide the process of joint learning as well as communal planning and execution of extension activities. Plans and methods are semi-structured and are revised, adapted, and modified as fieldwork in participatory extension proceeds. To this end, participatory extension requires conducive institutionalization of extension that permits participation to be put in practice (Farrington, 1994; Carney, 1998). PEA takes different forms to fit the varying and often specific situations. For instance, Black (2000) listed 32 participatory approaches practiced in the 1980s and 1990s.
Public agricultural research and extension systems in Africa are being challenged to reexamine their raison d’être in an era when food security, improved nutrition, equity and poverty alleviation have become top priorities of many governments in the continent. For instance, in the 1980s and 1990s most countries in Sub-Saharan Africa had been committed to implement economic reform programs, create a trade and investment environment and remove the obstacles to free market operation. These policy reforms had resulted in, among others, removal of subsidies on agricultural inputs and reduced public sector funding. In this connection, some authors underlined the fact that reduced state budgets curtailed the activity of government-run extension service in some African countries (Farrington, 1994; Woodhouse, 1994; Jones and Garforth, 1997; Nagel, 1997; Haug, 1998; Rivera et al., 2001; Eicher, 2007).
At present, there is consensus in the literature that, with reduced public sector funding, the participatory extension approach is proving to be the best means to improve sustainability-both of the benefits of investment in new technology and the extension service itself (Woodhouse, 1994; Hagmann, et al., 1999; Opio-odongo, 2000). Röling and Pretty (1997) argue that PEA was effective in disseminating improved technologies in many developing countries. Similarly, Picciotto and Anderson (1997) note that in both more and less-developed countries, farmer-led approaches to extension are spreading, while farmers’ associations, cooperatives, and self-help agencies are contributing handsomely to the diffusion of modern technologies. A cursory survey of the implementation of participatory extension approach, as a pilot project, in the African continent shows that it was used: in the development and spreading of soil conservation practices in Zimbabwe (Hagmann et al., 1999); in pasture management technology generation and dissemination in South Africa (Botha and Stevens, 1999); in integrated soil fertility management in Kenya (Baltissen, et al., 2000); in irrigation and water use projects in Zambia (Rivera, 2001); in FAO special program for food security in Tanzania (Rivera, 2001).
A review of the relevant literature shows that participatory extension approaches present a number of challenges that must be overcome so as to derive maximum benefit from them. Some of these problems include: relatively high costs of participation (particularly in terms of time) to smallholder farmers who often are too busy securing the basic necessities of life; lack of understanding and awareness of the long term outcomes of participation and the pursuit of short-term goals result in low community participation; community members have a tendency to "identify" problems that they believe are important in terms of encouraging donor participation and relegate the real problems to the second rank; some groups suppress diversity and impose what has been called a ‘tyranny of consensus’, arbitrarily rejecting some potentially good ideas; extension agents, even experienced, often lack technical, communication and facilitation skills to respond to the emerging and rising needs of farmers; there is often a tendency to leave out inaccessible, unorganized and marginalized groups by practitioners (Farrington, 1994; Cornwall and Jewkes, 1995; World Bank, 1995; Martin and Sherington, 1997; Hagmann et al., 1999; Black, 2000).
It is, therefore, important that extension practitioners recognize the multitude of problems that arise while implementing participatory extension approaches and embrace a strategy that is more responsive to local conditions, more accountable, more effective and more sustainable as well as more flexible about whom to work with and how.
Decentralization of Extension Services
In most Sub-Saharan African countries agricultural extension is largely a publicly-funded and publicly-delivered service. However, the history of state-led agricultural extension in Africa is full of stories about ineffective services. Part of the explanation for the poor performance of public agricultural extension institutions has to do with how these institutions are organized and how the services are delivered. In this respect, it is important to note that in most Sub-Saharan African countries the agricultural sector comprises a large number of small producers who generally cultivate relatively small plots of land and live in geographically dispersed communities, where the transport links are often of low quality. Thus, the number of clients that need to be covered by extension is large, and the cost of reaching them is high. The challenge is complicated further by the fact that farmers’ information needs vary even within a given geographical area due to variations in soil, elevation, microclimate and farmers’ means and capabilities (Anderson and Feder, 2004). However, available evidence shows that in Sub-Saharan Africa extension organizations have not been able to respond to the rising need for increased and more varied extension services (Purcell and Anderson, 1997; Anderson, 2007).
In the past several years, the realization that decision making and management in public extension organizations of developing countries have been highly centralized and formalized, while extension fieldwork has been increasingly demanding location-specific, flexible, and often quick decisions/actions, has called for extension reforms. Since the early 1980s, the main direction of reform in public sector agricultural extension service delivery has been towards decentralization of services with the firm belief that the services will be closer to the client, and thus more relevant. Decentralization of extension services often takes place in the context of public sector reform program at country/state level. The decentralization of extension services retains the public delivery and public funding characteristics of traditional centralized extension, but transfers the responsibility for delivery to local governments (district, county, etc.)  . Several Latin American governments undertook decentralization of extension in the 1980s and 1990s, and more recently this approach is being initiated in many Asian and increasingly also African countries (Qamar, 2002; Anderson, 2007; Eicher, 2007). Transferring program governance, administration, and management to the local level facilitates user participation and co-financing, enhances the response to local problems and opportunities, increases accountability to clients, and increases program efficiency (World Bank, 2006; Kahan, 2007; World Bank, 2008). A comprehensive strategy for decentralizing extension services must ensure service quality, develop capacities needed at all levels in the system, and provide clear definition of the respective roles and responsibilities of local and national governments and user groups. However, a review of the relevant literature reveals that decentralization of extension services may have adverse effects on service delivery. Among the most important problems facing decentralized extension agencies are: greater potential for political interference and utilization of extension staff for other local government duties (including election campaign activities); loss of economies of scale in training and the updating of staff skills; marginalization of extension agencies by local governments which prefer to spend their extension budget on other activities which generate quick revenues; and difficulty to establish efficient linkages among agricultural researchers, extension workers and farmers at local level if there is no research organization in the area (Qamar, 2002; Anderson and Feder, 2004; World Bank, 2006; Anderson, 2007; Kahan, 2007).
A review of the decentralization experience over the past two decades shows that there is no blueprint on how to implement decentralization successfully and that the forms of decentralization differ across countries. An important implication is that each country must craft a system of decentralization of agricultural extension services that is compatible with its government structure, needs of the farming population, central and local government capacities, agrarian structure and development strategy.
Pluralistic Agricultural Extension Systems
Even though the provision agricultural extension services was traditionally a public sector activity, new actors have now entered the scene to provide and finance advisory services, including non-governmental organizations, farmer organizations, academic/research institutions, private providers, commercial companies, and community-based organizations  . Some of the reasons that have led to the emergence of pluralistic service providers include: extension service provided by a diversity of organizations gives farmers greater choice of source of information; some organizations are more effective in reaching certain categories of farmers  ; the need for tailoring extension services to the requirements of specific sub-sectors or regions; with escalating budget deficits, the provision of extension services as free pubic services has been increasingly challenged on efficiency ground; and the quest for spreading the reach of extension to areas where a purely public sector service is unlikely to be viable and make it more responsive to local needs and opportunities (Farrington, 2002; Qamar, 2002; Alex et al., 2004; Birner et al., 2006; Anderson, 2007; Kahan, 2007; World Bank, 2008).
Birner et al. (2006), distinguish three sectors that may be involved in financing and providing agricultural extension services: the public sector (public administration, state agencies); the private sector (farm households, agribusiness enterprises, other profit-oriented firms); and the third sector (non-governmental and non-profit organizations, farmers’ organizations, civil society organizations). The same authors further note that institutional structures that are composed of organizations from different sectors may be referred to as "hybrid". Table 1 displays the variety of options that exist for financing and providing advisory services, if one takes the role of different sectors into account.
Source: Birner et al., 2006
In recent years, in most countries of the sub-Saharan Africa, a multi-institutional approach to extension service provision has become common place. For instance, Uganda is experimenting with the privatization of extension through creation of a pool of private extension specialists out of its existing public extension service, which farmers associations could call upon for providing services, and pay for the services from the funds given to them by the donors through central and decentralized government units (Farrington, 2002; Qamar, 2002; Anderson, 2007; World Bank, 2008). In some African countries (e.g. Ethiopia and Mozambique), NGOs have been engaged in delivering extension advice to resource-poor farmers living mostly in areas which are not serviced by public extension organizations. In these countries, NGOs have become "agents of development" that are actively involved in designing and implementing rural development programs and projects in the wake of major disasters (such as droughts and floods). Agricultural extension services provided by NGOs have the principal objective of enhancing agricultural development and improving farmers’ livelihoods. By promoting the participation of key stakeholders in the extension-decision making processes, emphasizing gender roles and relations and including vulnerable/ disadvantaged groups as the most important target beneficiaries in their agricultural extension programs, NGOs have proved themselves successful in terms of empowering beneficiaries and responding to the emerging needs which they express.
Likewise, in many sub-Saharan African countries, community-based organizations, such as farmers’ associations make extension services more accessible to small-scale farmers by providing economies of scale in service delivery and a mechanism for producers to express their demands for services (Farrington, 2002; World Bank, 2006; Anderson, 2007). Working with community-based organizations may enable extension programs to reach more farmers and rural households, facilitate participation in activities, and develop human resources and social capital  . In most countries of sub-Saharan Africa commercial agricultural companies provide embedded advisory services, which are integrated in commercial transactions such as sale of inputs or contract farming (Farrrington, 1994; Kahan, 2007). As already noted, public sector agricultural extension service delivery alone is not sufficient to address the multi-faceted and emerging needs of farmers. For instance, as farming becomes more commercial and high economic value produces are directed towards regional and export markets, more specialized advisory services are required. An important implication is that agricultural extension services need to be tailored to the demands of different types of farmers. Available evidence shows that while most commercial farmers get extension advice from private service providers on payment basis, smallholder farmers rely heavily on public sector advisory services (Alex et al., 2004; Birner et al., 2006; Anderson, 2007).
In a pluralistic agricultural extension system, the challenge for government is how best to coordinate the activities of the many other actors involved in the provision of extension services in order to meet the needs of farmers with varying resource requirements and social arrangements, and who live in diverse locations (Farrington, 2002; Alex et al., 2004; Kahan, 2007)  . In this respect, it is important to note that the absence of co-ordination among various agencies, in pluralistic agricultural extensions systems, has led to unhealthy competition, wasteful overlap and conflicting technical recommendations, which created confusion among farmers and damaged the reputation and credibility of service providers (Qamar, 2002; Alex et al., 2004). This clearly indicates that the shift towards pluralistic advisory services requires new skills, which allow field and administrative staff to manage complex relations among a wide set of partners (Birner et al., 2006).
Recent reforms in the delivery and financing of agricultural extension services in many developing countries have shown that funding extension is a central issue and mechanisms of cost recovery are critical for the sustainability of extension services (Anderson, 2007; Kahan, 2007; World Bank, 2008). Consequently, new funding mechanisms in which clients share costs are increasingly being pilot-tested in many developing countries where direct payment for agricultural extension services by smallholder farmers is not an established practice. However, there is genuine fear that the zeal for cost-recovery would deprive smallholder farmers from benefiting from the services. This is precisely because smallholder farmers either do not believe that extension advice is worth paying for, or they simply cannot afford to pay (Farrrington, 1994; Qamar, 2002; Anderson, 2007; Kahan, 2007). In the short run, the common wisdom would dictate that commercial farmers should pay for extension advice while the government should provide extension services to smallholder farmers free of charge. In the long run, however, an incremental step-wise approach to cost recovery measures could be considered and the level of subsidy reduced over time so as to give farmers ample lead time to adjust to paying for services that previously were provided as a free good. It should also be noted that public sector will continue to be the major provider of agricultural extension services to resource-poor farmers and disadvantaged groups in the foreseeable future.
Agricultural Value Chains
At present, there is ample empirical evidence pointing to the fact that the dissemination of standard packages of inputs and practices in most developing countries has not brought about meaningful improvements in smallholder farmers’ livelihoods and incomes. Part of the explanation for the poor performance of public agricultural extension services in many developing countries has to do with the ‘top-down’ approach to extension work which focuses only on farm level interventions that result in agricultural productivity improvements. However, the reality on the ground reveals that smallholder farmers face the core problem of low productivity combined with, among others, lack of market access, distorted market prices, inadequate market information, high transaction costs, shortage of working capital and underdeveloped and unreliable infrastructure (Bernet et al. 2005; Anderson, 2007; World Bank, 2007). Given this state of affairs, it has become imperative to shift away from the conventional extension system, which focuses only on productivity-enhancing technology promotion, toward an approach that empowers farmers and fosters linkages and alliances from production to consumption.
At present, there is an agreement in the literature that the prospects for rural communities that only receive support for production-based development and subsidized services are bleak. Partly as a preemptive measure to ward off this imminent problem, since the late 1990s, in many developing countries, the provision of agricultural advisory services has been closely associated with the concept of agricultural value chains  . This approach goes beyond the farm and the farm family and looks into common business relationships and interactions between and among farm enterprises and agribusinesses along the pathway from planning for production to the consumption of the final product (Van den Ban, 2005; Bammann, 2007). The principal aim of the value chain approach is to reinforce business linkages and partnerships among the various market chain actors, who normally compete and mistrust each other in their daily business, and thereby improve the performance of the chain and generate direct and/or indirect benefits to all the participants in the chain (Bernet et al. 2005; Bammann, 2007).
According to KIT et al. (2006), for smallholder farmers, value chain promotion by agricultural service providers involves empowering them to identify market opportunities, build their organizational and supply chain management skills, increase their level of competitiveness and diversify into alternative and higher value products. These goals can be achieved through better economic coordination and institutional arrangements that link the various chain actors. In this respect, essential activities and mechanisms include: organizing forums and supporting establishment of producer organizations; promoting information flows; and experimenting with new approaches to facilitate access to knowledge, skills, and services from a wide range of organizations (Rajalahti et al., 2008). For public extension organizations, it means becoming market-oriented and placing more emphasis on developing farmers’ capacity to produce for identified markets, reduce losses, improve the quality and delivery of the product (or range of products), and reduce marketing and/or other transaction costs and risks (Van den Ban, 2005; Anderson, 2007). However, in most developing countries public sector agricultural extension workers are trained in production technologies and have very limited knowledge on relevant marketing issues and linkages. Therefore, for agricultural advisory services to succeed within a value chain system, extension workers at all levels must be trained in areas beyond technical agriculture to build skills in mobilizing farmers, tapping market intelligence and managing farm and non-farm businesses (Van den Ban, 2005; Bammann, 2007; World Bank, 2007; Christoplos, 2008; Rajalahti et al., 2008).
Agricultural advisory services within a value chain system can serve as a bridge among the various market chain actors and facilitate partnerships, building coalitions of different actors, such as between producers and processors or traders, to better respond to market demand and exploit value addition opportunities (Bernet et al., 2005; Rajalahti et al., 2008). The value chain approach to agricultural advisory services recognizes that there are advisory service clients at each tier in the value chain and this implies that advisory services must meet the needs of all the market chain actors. As agricultural markets operate in an ever changing environment, agricultural advisory services must respond effectively to the needs of the various market chain actors as they adapt to factors impacting on agricultural markets. The availability of effective advisory services has consequences for the performance of the market chain in that the efficiency of a market chain is dependent upon how well information flows between chain actors, their level of business linkages, and the ability of advisory services to overcome problems as they arise. In this respect, in order to establish effective service provision and to keep their services updated, attractive, of high quality and in tune with clients’ needs, agricultural advisory services providers need access to ‘back-up services’, including accessing information, training and mentoring in a range of skills (Christoplos, 2008).
It must be noted that the provision of agricultural advisory services is not a matter for public extension organizations alone. In practice, even though public extension might be the main source of information at the initial stage of value chain promotion, over time, the demand for agricultural advisory services suitable to diverse clientele and particular social and market conditions would lead to the emergence of other service providers, such as farmers’ organizations, the private sector and non-governmental organizations. As successful intervention in a value chain enables all the participants along the chain to sustain a profitable business, with increased commercialization, agricultural advisory services tend to become increasingly "demand-driven", with the various chain actors paying for services that respond to their needs. For instance, there are increasing numbers of private production, processing and marketing enterprises in developing countries that work through contracts with farmers or farmer organizations and provide advisory services to ensure timing, quality and quantity of product delivery. Likewise, in some developing countries farmer organizations have started to take over what had previously been the governments’ role in providing training, market information and advisory services to their members.
The existing empirical literature indicates that value chain promotion replaces the traditional supply-driven and ‘top-down’ approach to public agricultural extension delivery by a demand-led approach with multiple providers (Van den Ban, 2005; KIT et al., 2006; Anderson, 2007; World Bank, 2007). However, as the new approach focuses on enhancing the entrepreneurial and managerial abilities of market chain actors, specific attention must be given to empowerment, capacity building and teaching (particularly among smallholders), the organization of producers and the rural poor, and the identification, articulation, and building of demand for advisory services through financing and capacity building (Christoplos, 2008; Rajalahti et al., 2008).
Experience from other parts of the world shows that distrust among different actors discourages the rural poor from taking on more market-oriented approaches (Christoplos, 2008). Thus, interventions aimed at linking smallholder farmers to markets should explicitly include measures designed to promote the building of trust-based linkages. Successful value chain development in developing countries also requires the existence of effective farmers’ organizations. This is precisely because individual smallholder farmers are weak players in the market who are often at a disadvantage vis-à-vis large traders, input suppliers and multi-national companies. Under this increasingly unfavorable condition, farmers do not have to deal with other market chain actors as individuals, but through their own organizations. The reality on the ground requires smallholder farmers to be better organized at both the enterprise and market levels. According to Conroy (2003), some of the benefits of farmers’ organizations include: making agricultural extension services more client-driven and efficient; facilitating processing of agricultural products; strengthening farmers’ bargaining power with traders; reducing transaction costs for input supplies and output buyers; allowing economies of scale (e.g. from bulking up in output marketing, buying inputs or storage); facilitating savings and access to credit; and reducing public-sector extension costs. In value chain promotion interventions, agricultural service providers play crucial roles in building farmers’ organizational and management skills. More precisely, service providers can build the capacity of farmers’ organizations to identify profitable market opportunities, adapt and improve their produce and work with others in a market chain to meet the increasing demands of ever-diversifying consumers. In this respect, effective agricultural market chain development requires good understanding of how to establish and manage successful farmers’ organizations.
The success of value chain promotion interventions depends also partly on government commitment to support the shift toward market-oriented production system. Some of the possible measures that governments could take to enhance market orientation include: designing effective public policies which support market orientation and regulate the downsides of growing commercialization; investing in public goods which promote the development of agricultural markets; creating a business enabling environment for private sector to perform efficiently; preventing powerful actors from manipulating market governance; and building the capacities of chain participants to innovate, diversify or exit as markets change. A key area for government support is improvement of rural market infrastructure, which is critically important for linking up rural production with urban markets and ensuring economic benefits to chain actors. It is also important that governments put in place mechanisms to regulate and coordinate multiple providers of advisory services.