Connecting marginal rice farmers to agri

Journal of Agricultural Education and Extension Vol 12, No 2, 109Á125, June 2006 Connecting Marginal Rice Farmers to Agricultural Knowledge and Information Systems in Vietnam Uplands JEAN-CHRISTOPHE CASTELLA$, JOEP SLAATS%, DANG DINH QUANG, FRANC OIS GEAYĐ, NGUYEN VAN LINH and PHAM THI HANH THO’ $ Institut de Recherche pour le De´veloppement (IRD), France, and International Rice Research Institute (IRRI), Philippines, %Independent consultant for agriculture and capacity building, Vietnam, ’Vietnam Agricultural Sciences Institute (VASI), Vietnam, §French Ministry of Foreign Affairs, French Embassy in the Socialist Republic of Vietnam, Vietnam ABSTRACT In Vietnam, agricultural extension has contributed to rural development and poverty alleviation over the past two decades of agricultural decollectivization, but it was not very effective in reducing disparities within farmer communities The study examined how better interactions of extension services with other agencies and information sources may help marginal farmers in catching up with the general improvement of living conditions in a mountainous area in northern Vietnam It combined three complementary viewpoints on this issue: that of the agricultural extension staff, that of farmers and that of development experts with a long working experience in the mountains of Vietnam The analysis of existing structures and functions of the extension system revealed a number of obstacles to the participation of marginal farmers in extension programmes and helped to identify relevant domains of intervention KEY WORDS: Agricultural extension, Social networks, Innovation, Marginal farmers, Vietnam Introduction Since the onset of the economic liberalization process in 1986, Vietnam has shown tremendous overall economic growth and a sharp decrease in poverty Agricultural production increased exponentially as allocation of land-use rights and better access to inputs and information enabled farmers to react to market incentives (Kerkvliet and Porter, 1995; Obaidullah-Khan, 2000) In the northern mountainous area of the country, agricultural development has been less spectacular because of its isolation and the marked heterogeneity of socio-economic and agro-ecological conditions (Jamieson et al., 1998; Le and Rambo, 2001) People are difficult to reach but they also have limited livelihood options because they are isolated from infrastructures, markets and administrations (Alther et al., 2002) The Vietnamese government’s concern for equal access to national development, coupled with international donors’ Correspondence address : Jean-Christophe Castella, IRD, B.P 64501, 34394 Montpellier Cedex 5, France Email: j.castella@ird.fr 1389-224X Print/1750-8622 Online/06/020109-17 # 2006 Wageningen University DOI: 10.1080/13892240600861625 110 J.-C Castella et al concern with poverty alleviation, has focused attention on agricultural development in the northern mountains (World Bank, 2001) Recent government investments to help overcome the isolation of this region by upgrading the road infrastructure and subsidizing the supply of agricultural inputs led to a substantial improvement of living conditions (Van de Walle, 2002) Nevertheless, even in the communities where general conditions improved over recent years, a large proportion of the households remains poor (Le and Rambo, 2001; Castella and Dang, 2002) The insight that overall economic growth will not be able to lift all people out of poverty is now generally accepted (Beckman, 2002) The government is becoming progressively more interested in adapting extension methods and structures to the challenge of poverty alleviation In Vietnam, as in many other countries around the world (Rivera and Alex, 2004), a major reform of the services in charge of research, development and extension under the Ministry of Agriculture and Rural Development is underway Several support programmes contribute their field experience to the ongoing policy-making exercise (World Bank, 2001; ADB, 2002) However, a key issue for improving the effectiveness of the extension system concerns the institutionalization in the governmental extension service and up-scaling to large geographic areas of the most promising extension methods (for example, participatory technology development, farmer field schools, community-based management of natural resources, companion modelling) A number of participatory methods have been used only recently in various support programmes, in relatively small areas, and are spreading only very slowly For many years, communication initiatives in support of environmental and natural resources management have mainly focused on the dissemination and adoption of technical packages These efforts have met with some impact under a collectivist agricultural system that was designed in such a way that the ‘users’ were mostly inactive recipients in a technology-driven process The technology received most of the attention, often to the detriment of the ‘non-technological’ factors (that is, social, economic and cultural) which influence farmers’ behaviour Several years after the transition to market economy and family-based farming, the knowledge management and communication practices are still remnant from the former period Weak linkages among technology-oriented research, education and extension institutions limit their effectiveness to contribute to development, especially when farmers are poorly organized, poorly endowed with biophysical resources, and have limited access to transport, storage and processing facilities as it is mostly the case in the mountain areas Therefore, communication with the poor is essential to meet the needs of the isolated and marginalized groups by creating a plural information society, supporting the ongoing decentralization of support services to agriculture However, moving from the mainly vertical and top-down communication strategies promoting externally-generated technology towards the integration of different kinds of knowledge held by various stakeholders is a challenging endeavour, especially in the increasingly complex and horizontal communication environments in which development strategies are currently deployed An integrated approach to rural innovation (Leeuwis and van den Ban, 2004) should emphasize the importance of interactive, mutual learning between formal and informal knowledge/technology Marginal Rice Farmers in Vietnam Uplands 111 systems While in the traditional technology transfer approach more attention is paid to ‘trickle down’ flows of information from research to extension and from extension to farmers, an integrated approach to innovation shifts attention to feedback and upwards communication from farmers and to facilitating research-extension-farmer interactions Beyond the improvements of the vertical communication system, conceptual shifts and economic constraints have lead to a greater consideration of pluralistic approaches, that is, multi-organizational partnerships including nongovernmental actors in the broadest sense A process of social learning involving interdependent stakeholders at multiple scales is critical to the required institutional changes (Leeuwis and Pyburn, 2002) Participatory approaches can facilitate the dialogue and exchange of knowledge and information on natural resources management, increase the community knowledge base, promote agricultural practices which are compatible with the environment, and develop awareness in policy-makers, authorities and service providers (Roăling and Wagemakers, 1998) It is one thing to observe that an integrative and systems approach should be taken to innovation facilitation, but quite another thing to translate the observation into action (Sayer and Campbell, 2004) Many organizationally elegant models of communication can be described, that never quite seem to materialize because they build upon theoretical considerations before taking into account local realities Learning processes and stakeholder interactions are necessarily embedded in a local institutional and political context that must be understood prior to external intervention (Castella et al., 2004) Social learning hinges on context sensitivity, tactical flexibility, and intuitive skill to harness the creative energy of collective engagement in problem solving Then, the challenge is to scale out and up and institutionalize platforms and processes of social learning Leeuwis and Pyburn (2002) stress the role of communication within and in between knowledge systems and policy networks to transcend localized interventions towards system-wide multiple-scale impact The purpose of the study presented in this paper was to define mechanisms for linking the two dimensions of communication for development as presented above, that is, vertical hierarchical structure of the public extension system with the horizontal local networks, which support the active participation of farmers in knowledge sharing and social learning (Balasubramanian and Thamizoli, 2003) It built upon the hypothesis that participatory approaches to technology development, adaptation or adoption by farmers can be sustained and can have a real impact on the marginal farmers who are the most in need only under the condition of reconciling vertical and horizontal communication systems in agricultural extension This paper first presents an institutional analysis of the extension system in Bac Kan province It describes the organizational aspects and structures involved in the innovation process and technological changes Then, the results of the diagnostic study are confronted with the visions and expectations of different groups of stakeholders involved in agricultural development Finally, the discussion section introduces concrete recommendations to reconcile communication channels that were initially designed to support widely practised technology transfer methods with participatory approaches for the benefit of marginalized farming communities in the uplands 112 J.-C Castella et al Methodology Conceptual Framework In Vietnam, interaction between extension agents and farmers still often consists of a top-down flow of information, delivering a fixed and ready-to-use technical solution, not specific to farmers’ needs and wishes (Hoang and Nguyen, 2003) This linear transfer mechanism inherited from the collectivist period offers little opportunity for feedback on content and method of communication and monitoring of achievements and is mainly focused on attaining production targets formulated in provincial and national extension programmes This extension approach may be considered as archaic in the present days despite its successful contribution to the dissemination of technical innovations in large homogenous areas such as the plains of the Red River Delta (Dao, 1995) In the mountain areas where people farm under agro-ecological conditions that differ from one place to another, this approach often results in messages that are not relevant to farmers’ conditions or to their needs, and the result is a low rate of innovation adoption (Shanks, 2002) This approach contrasts with the philosophy of the development projects intervening in the same area, which involve farmers in technology development and dissemination through a large range of participatory methods (BCG, 2002; Castella et al., 2004) As the extension practices are deeply anchored into the structure and functioning of the institution, major institutional changes are needed in the official extension system to support the generalization of participatory methods (Neef, 2005) However, the mechanisms needed to reform the system are still to be designed and debated among multiple stakeholder groups before implementation The study reported in this paper aimed at feeding the ongoing debates with empirical data and input from key stakeholders: extension agents, farmers and development advisers Three research methodologies have been combined in order to take stock of the existing extension system from the three different perspectives Through an institutional analysis we studied the set-up and functioning of the public sector extension service as well as its links with other actors who contribute to technological changes in rural areas Through a farmer survey we analyzed the relationships between communication networks, knowledge acquisition and innovation adoption in farmer communities Finally, we explored the interplay between both the individual components of the innovation process and the validity of the findings beyond the study area through individual and collective exchanges with key informants on their working experiences Data Collection Institutional analysis of the agricultural extension system First, we reviewed reports from major organizations involved in agricultural extension such as the local administration (that is, People’s Committees at province, district and commune levels), mass organizations such as the Women’s Union (WU) and the Farmers’ Association (FA), and non-governmental organizations (domestic and international NGOs), to get an overall picture of the information channels on agricultural technologies in Bac Kan province We described the structure of the major organizations involved, including their mandate, activities, staff, equipment, and financial endowment In 2002, once key stakeholders had been identified, we Marginal Rice Farmers in Vietnam Uplands 113 conducted individual interviews (n 0/25) with extension staff based on a semistructured questionnaire The interviewees were selected according to their position and responsibilities in their institution (that is, administrative staff at provincial, district and commune levels, team manager, field staff, etc.) and their domain of intervention (for example, cropping systems, animal husbandry, credit) Based on this panel, representative of the different extension activities in the province, we studied how priorities are set and we investigated the local communication channels between farmers within the local communities (Hoang et al., in press), as well as between extension services and farmers Last, three focus group discussions provided a better understanding of the spread of, access to and control over agricultural information and local knowledge The first group included representatives of the administrative offices at provincial, district and commune levels, the second group included representatives from national and international development projects, managed by domestic and international NGOs, bilateral and multilateral development agencies, and the third group included field staff from the extension system and their local partners Farmer surveys The first component of the survey focused on the links between farmers’ conditions and their access to information and communication means, their acquaintance with the extension service and their participation in extension activities The second component investigated the relations between farmers’ sources of information and the quality of knowledge on agricultural innovations, and their decision to adopt or not Our questions concerned three specific innovations in paddy rice production that had been recently disseminated in the study area: hybrid rice (HybR), on-farm rice seed production technology (SeedP) and rice transplanting techniques (RTransp) (Box 1) Innovations on paddy rice production were chosen specifically because this cropping system plays a crucial role in farmers’ livelihoods (Castella and Dang, 2002) The diffusion of hybrid rice is promoted by the government through the extension service, which provides technical support, as well as through other stakeholders such as seed suppliers; the second innovation is mainly supported by an international project underway in the area (Le, 2003), while the last-mentioned technology is disseminated by the official extension service only In 2001, we conducted the survey in nine villages located in three communes of Cho Moi district (Box 2) We chose this district because it is relatively accessible and the extension service intervenes on a regular basis In each village 15 households were chosen at random from the village census, making up a sample of 135 households (Table 1) The survey was conducted by means of a structured questionnaire with pre-defined answers (multiple choices) in order to obtain identical information from all respondents Exchange with key informants The key informants we interviewed were managers of development projects (supported by international NGOs, bi- and multi-lateral donors) dealing with agricultural extension in the mountains of northern Vietnam A representative panel of six persons was selected among the 12 projects dealing with rural development in Bac Kan province at the time of the study (BCG, 2002) Discussions were guided by a list of open questions about the functioning of the extension service, the effectiveness of extension methods and the use of incentives for 114 J.-C Castella et al Box Rice innovations introduced in Bac Kan province HybridR Hybrid rice seed varieties, most of them of Chinese origin, are strongly promoted by MARD as part of a policy for food security in Vietnam uplands These varieties require adequate crop management (irrigation, fertilizer, pest and disease control), the taste of these varieties is only moderately appreciated Their massive introduction may lead to risk of erosion in genetic diversity, an issue that is beyond the scope of this paper They are grown in the spring and summer season and yield can reach t/ha The technical guidelines issued by AFEC describe timing and methods of field operations: seed and seedbed preparation, seeding, transplantation, fertilizer application, weeding, irrigation and pest management SeedP This is on-farm seed production of improved open-pollinated rice varieties (Bao Thai Lun), specific to the northern mountainous region These varieties are well adapted to local ecological conditions, are appreciated for their taste and yield stability and have little tendency to degenerate when saved as next year’s sowing seed These varieties are specific for the summer season and yields can reach 3.5 t/ha The guidelines issued by an international NGO describe plot requirements for seed production, crop management techniques and specific considerations and practices for selecting and harvesting the plants as well as storing and handling the harvested seed TransP The innovation consists of raising rice seedlings in a tray consisting of soilfilled holes instead of in a seedbed For transplanting, the seedlings and the ball of soil attached are lifted from the tray and broadcasted by hand into the field Advantages of this technique are the reduction in time requirement for transplanting and better establishment of seedlings after transplantation The technical guidelines issued by AFEC describe the techniques for sowing in the tray and for transplanting adoption With the information gained, the results from the institutional analysis and households’ survey were cross-checked for consistency At a seminar in January 2003 that brought together representatives of the main institutions and groups of stakeholders involved in agricultural extension in Bac Kan province, we presented the results of our study and received feedback from them The Agricultural Extension System in Bac Kan Province: A Pluralistic Model A Vertical Administrative Structure with Horizontal Communication Networks Agricultural extension is officially managed by the Department for Agro-Forestry Extension (DAFE), which was set up in 1993 within the Ministry of Agriculture and Rural Development (MARD) It is closely linked to the administrative structure, the People’s Committee (PC), which has a representative unit at each hierarchical level At the district level, it consists of a specific unit, the Agriculture and Forestry Extension Centre (AFEC), under the provincial Department of Agriculture and Rural Development (DARD) District extension officers support agents at lower levels in the implementation of the programme and perform monitoring and evaluation In 2000, the AFEC appointed 84 officers for 122 communes, each assigned to work in one or two communes They include civil servants and temporary workers and can be considered as Communal Extension Agents (CEA) About half of them are women and they are paid by the central government (MARD) or the province They serve 44,500 farming families (530 farms per CEA on average), which is far above the average rate in Vietnam of one agent for 2000 Á4000 families (ADB, Marginal Rice Farmers in Vietnam Uplands 115 Box Characteristics of the on-farm survey Cho Moi district is located in the southern part of Bac Kan province The three communes where the farm survey was carried out are located in an area of low mountains and hills with a relatively large number of lowland valleys All three communes are connected to the national road linking the provincial capital to the neighbouring provinces Mostly the household head was interviewed; in some cases it was another household member directly involved in extension activities such as training or demonstration Farmers’ knowledge on the studied innovations was assessed based on their capacity to cite the main topics of the guidelines that had been handed out as support during the introduction of these innovations by the extension service A pre-defined scoring system was applied to evaluate farmers’ understanding of the technical topics in these guidelines in order to obtain uniform results that could be compared As far as the farmers’ conditions are concerned, the composition of the farming family, the level of education, sources of family income and land use system were taken into account (Table 1) The farm households involved in the survey shared some common characteristics They were all active in agriculture and very few had any significant secondary economic activity Most families belonged to the Ta`y ethnic group In general, the household heads had a relatively high level of education, at least secondary school, which was a common feature of people living in easily accessible areas This characteristic is important for this study since formal education facilitates access to information and favours awareness of potential benefits from innovation adoption As for crop production systems, almost all farmers grow paddy rice About 20% lacked irrigation to grow a spring rice crop The proportion of income from rice was relatively high, in 50% of the farmer families it represented between one and two thirds of total income Almost a quarter of the farm households did not produce enough rice to cover annual food needs 2002) This reflects the priority the government gives to improving living standards in the poorest parts of the country Mass organizations such as the Women’s Union (WU) and the Farmers’ Association (FA) have branches at district, commune and village levels They also provide training and participate in governmental programmes and development projects During the cooperatives period, these government-led structures were used mainly for disseminating directives, but they are now increasingly becoming forums for local development planning In Bac Kan province, radio and TV is run by an agency under the provincial PC that operates two broadcasting stations Both broadcast programmes on agriculture with special attention to technological innovations Development projects operated by national agencies in partnership with international NGOs and donor organizations often also include agricultural extension In 2001, 12 projects had activities related to agricultural extension in Bac Kan province All the projects promoted participatory approaches, most of them dealing with training of both extension staff and farmers, while about half of them were active in the demonstration of innovations, input supply, credit activities, exchange through workshops and visits, and organizational strengthening (BCG, 2002) In Bac Kan province, farmers may gain information on agricultural technologies by taking part in meetings organized by development agencies presented above, through direct contact with the local extension agent, and through informal contact with neighbours or other farmers (Table 2) About 70% of surveyed families have 116 J.-C Castella et al Table General characteristics of farms in the survey Characteristic Average Range 42 4.8 2.9 23 Á72 Á9 Á6 Age HHa head (years) Family size (# persons) Active HH members (# persons) Education Ethnic identityb Economic activities (# farms) Acreage paddy land (1000 m2)c Paddy land in spring (1000 m2)d Income from rice (%) Self-sufficiencye in rice (# farms) Classes (% of farms in survey) B/35 (25%) 5/4 (50%) 5/2 (52%) 35 Á44 (39%) /4 (50%) /2 (48%) 2.7 primary secondary (16%) (72%) Tay (84%) Dao (11%) primary: secondary: agriculture (100%) none (92%) 0.2 Á10 B/2 (31%) Á B/3 (33%) 2.0 0.2 Á6 B/45 (36%) high (40%) ]/45 (36%) higher (12%) Kinh (3%) ]/3 (36%) 45 Á B/65 (35%) ]/65 (29%) global (37%) low (23%) a HH: household other ethnic groups: 2% c 130 households have paddy fields d 106 households have paddy fields where spring crop can be grown (two cropping cycles) e Self-sufficiency is measured as total annual rice production per household member divided by annual consumption (300 kg/person/year); high self-sufficiency: summer crop alone meets needs, global: summer plus spring crop meet needs, low: summer plus spring crop fall short of needs b access to radio and TV Surprisingly, farmers did not mention the mass media as a source of extension information However, in a nationwide assessment, farmers considered TV programmes on local agriculture to be useful (MARD, 2003) Village meetings, training courses and technical brochures appeared to be the major sources of extension information used by farmers On-farm demonstrations of innovations are important for about 50% of the farmers Farmers were well aware of the involvement of the extension service in their community About 88% of them knew the extension agent and 78% were able to list the activities he or she was undertaking However, only 8% of the farmers had been visited by the extension agent, suggesting the limited interaction between these two actors, as reported elsewhere in the Table Farmers’ use of sources of extension information Source Village meeting Training Brochures Mass organization Neighbour Demonstration Interest groups Mass media a Proportion of farming households using the source (n0/135) Usea 93 75 71 60 56 56 11 Marginal Rice Farmers in Vietnam Uplands 117 province (Hoang et al., in press) and in Vietnam (MARD, 2003) Farmers’ acquaintance with the extension agent did not depend on the characteristics of the farm or household The channels through which farmers obtained information did not depend on their acquaintance with the extension agent and his activities, suggesting that direct interaction with the extension agent was not essential in order to be informed about agricultural innovations In carrying out extension activities the agent is assisted by the local administration, the Communal People’s Committee (CPC), which channels all the information to the village authorities before it reaches the farmers This mode of operation gives the local authorities (people holding a position of responsibility in communal or village institutions such as members of the Communal People’s Committee, village heads, representatives of mass organizations) a key position in brokering information about innovations The Key Role of Commune Extension Agents in Brokering Technical Information between Vertical and Horizontal Communication Networks The official mandate of the CEA is clearly inspired by a linear model of technology transfer to farmers, which consists in providing information on innovations in the form of technology packages or models to be implemented for demonstration purposes (Peters, 2001) The transfer is mainly accomplished by providing training to farmers, distributing technical brochures formulated by AFEC, establishing demonstration models with contact farmers, and monitoring innovation adoption (MARD, 2002) In addition, the CEA have to carry out extension and rural development activities for NGOs and donor organizations operating projects in their specific commune In carrying out extension activities, the CEA work closely together with the person in charge of agriculture within the Commune People’s Committee (CPC), normally the vice-chairman This person acts as a contact person who informs village heads and other authorities on the activities and these people in turn pass this information to farmers The contact person also keeps a record of farmers interested in a training session or in receiving the inputs needed to apply a technology In order to improve the outreach of the formal extension system, MARD has introduced extension clubs in several provinces, including Bac Kan province (Shanks, 2002) These clubs consist of farmers within a village or commune who meet each month to discuss technical information or undergo training Local authorities assisted in establishing these clubs and are themselves members (Nguyen et al., 2002) The achievements of the formal extension system in Bac Kan province are limited for the same reasons as those reported in other provinces (ADB, 2002; UNDP, 2000) The most common arguments used to explain the poor results are the small number of agents, their low field allowance, and their insufficient means of transport to reach farmers who are widely dispersed and often live in remote areas Extension agents have limited capacities to deliver their message to farmers because they are poorly trained and often not speak the languages of the minorities However, it is also important to note that staff who originate from the geographic area they serve have better knowledge of the local agricultural systems and are highly committed to respond to farmers’ needs In the mountain areas however, their level of education is generally lower than people from outside Nevertheless, despite their low salaries (equivalent to 20 USD per month in 2002), most CEAs are highly motivated and try 118 J.-C Castella et al to respond to farmer’s demands even if these fall outside their formal agenda In addition, extension agents occupy a unique position as broker of information between the different governmental agencies, international development projects and local authorities Unfortunately, some extension agents who contribute to international projects and are trained by these projects in various extension methods including participatory approaches, tend to compartmentalize their knowledge They apply linear technology transfer methods when working in their official function but also apply successfully participatory approaches when working on international projects However, training is required to help them increase their capacities in rationalizing the different flows of information and change their attitudes towards partnerships (Foărster, 2003) In this way, summary information on innovations could be exchanged and subsequently passed on to each structure In the end, this information may reach farmers through different channels, evidence of a huge potential to diffuse general information of a low level of complexity throughout the system Farmers’ Participation in Extension Programmes and Technological Innovation Knowledge Transfer Knowledge transfer on rice-related innovations has been studied by exploring where farmers who adopted the technology obtained the guidelines for these innovations and how they understood these guidelines (Box 1) The main sources of these guidelines were training, neighbours and village meetings (Table 3) For HybR and RTransp, more than half of the farmers got the guidelines by attending a training course Since SeedP is an innovation introduced through a project and is solidly based on collective action by farmers in seed producers groups, most farmers got the guidelines through their neighbours Obtaining guidelines through training was also correlated with relatively large paddy fields and a moderate to high degree of selfsufficiency Conversely, farmers with no position in local administration or those who earn most of their income from other sources than rice (for example, maize, cassava, or livestock products) tended to obtain the guidelines through their neighbours These findings underscore the key role of training in extension Other sources of extension information, such as demonstrations, appeared not to be the main occasions that allow people to obtain specific technical information This result contradicts the widespread belief within the formal extension system that the spread of knowledge and technologies occurs through demonstration because farmers are Table Main sources of the technical guidelines for specific innovations Source Training course Neighbours Village meeting Seed supplier Others a HybR SeedP RTransp 50a 40 32 13 31 49 17 14 63 23 35 Percentage of farmers mentioning the source (HybR: n0/98, SeedP: n0/35, RTransp: n0/38) Marginal Rice Farmers in Vietnam Uplands 119 more likely to learn from farmers like themselves who already practise a given technology In reality, it is mostly the better-off farmers and people well connected to the local authorities who set up demonstration with the help of extension staff and receive credit and subsidies (Peters, 2001) Poor farmers have less chance to benefit from these model demonstrations, especially if they have a marginal position in the village communication network (Hoang et al., in press) Innovation Adoption Of the three innovations monitored in the study, HybR was the most widely adopted, by 75% of the farmers This is likely to be the result of a nationwide campaign launched by the government to promote the use of HybR in order to guarantee food security and increase rice exports SeedP and RTransp were adopted on a limited scale only, by 27% and 29% of the farmers respectively Impact of farmers’ knowledge on adoption Farmers who obtained the guidelines through training adopted HybR and RTransp more than those who got the guidelines through neighbours and village meetings (Table 4) With regard to HybR and SeedP, a better understanding of guidelines also resulted in earlier adoption These findings confirm that training more than other sources of information contributed to farmers’ understanding of the innovations and indicate the importance of the farmer’s position in the local communication network, which largely determines his or her access to training activities The majority of the farmers perceived the innovations to be a benefit, since only 15% to 30% of them considered the impact of adoption as low Informed non-adoption remained limited (Table 5) Impact of farmers’ livelihood systems on adoption The adoption of the three innovations studied here depended on farm and household characteristics (Table 6) The high resource requirements for implementing any of the innovations proved to be the biggest limitation (Table 5) Farmers with only small paddy fields did not fully adopt the innovations This is in particular true for HybR as farmers with smaller paddy fields were fewer to adopt, and also adopted the innovations later Farmers for whom rice was an important source of income are more likely to adopt HybR and SeedP than those for whom rice was a minor source of income Farmers with a high degree of self-sufficiency adopted HybR in larger numbers and earlier than those who were less self-sufficient Table Relationship between innovation adoption and knowledge transfer Dependent variable Year of adoption (from 1999) a Explaining variable Source of guidelines Understanding guidelines (Box 2) Subsidies Probability value of x2 test; ns 0/not significant (p ]/0.10) Innovation HybR SeedP RTransp B/0.05a B/0.10 ns B/0.10 B/0.05 ns B/0.10 B/0.01 ns 120 J.-C Castella et al Table Constraints limiting farmers’ adoption of innovations Type of constraint Innovation HybRa SeedP RTransp 15 20 86 21 46 50 28 46 24 74 Low impact of innovation Limited access to information Guidelines difficult to understand High requirements for adoption a Percentage of respondents mentioning the constraint (HybR: n0/98, SeedP: n 0/35, RTransp: n0/38) Impact of subsidies on adoption Members of the local authorities were more numerous in adopting SeedP and RTransp than farmers with no official position A similar trend was found for the adoption of HybR These findings may reflect the effect of subsidies on adoption The proportion of farmers who received hybrid rice seeds at a subsidized price tended to be larger among local authorities than among farmers with no position, although the difference was not statistically significant All three innovations studied were introduced together with incentives in the form of subsidized or free distribution of seeds (HybR, SeedP), fertilizers (HybR) or equipment (RTransp) Since resource requirements (that is, material inputs, irrigation capacity, etc.) were revealed to be serious limitations (Tables and 7), it is easy to understand that subsidies sped up innovation adoption as observed in the case of SeedP and RTransp The explicit increase in yield of HybR may explain why subsidies did not affect the adoption of this innovation despite the high proportion of farmers who received a subsidy (78%) Discussion: Burning Issues in Extension Reaching the Farmers The present study indicated that farmers moderately to well endowed with resources have better access to extension than their poorly endowed fellows since they tap a greater variety of sources of extension information and participate more in extension Table Link between innovation adoption and farm characteristics Dependent variable Explaining variable Innovation HybR SeedP RTransp B/0.10 B/0.10 ns B/0.05 B/0.05 ns ns B/0.01 Adoption (yes/no) Size paddy fields Proportion income from rice Self-sufficiency in rice Social position household head B/0.01 B/0.10 B/0.05 ns Year of adoption (from 1999) Size paddy fields Proportion income from rice Self-sufficiency in rice Social position household head B/0.10 ns B/0.10 B/0.05 Probability value of x2 test; ns 0/not significant (p ]/0.10) ns ns ns ns ns ns ns ns Marginal Rice Farmers in Vietnam Uplands 121 Table Resource requirements that limit farmers’ adoption of innovations Type of requirements Material inputs Fertile soils Irrigation capacity Labour Land a Innovation HybRa SeedP RTransp 88 46 41 64 57 29 54 40 26 14 14 Percentage of respondents mentioning the constraint (HybR: n0/98, SeedP: n 0/35, RTransp: n0/38) activities As a consequence, local authorities and better-off farmers are more numerous in adopting innovations and adopt them earlier than poorer farmers This finding implies that the formal extension system in Bac Kan province does not reach all farmers to the same extent and some categories of farmers, especially the poorest, are likely not to be reached at all Farmers’ exposure to the formal extension system also depended on their position in local communication networks and thus on their relation with the local authorities at the centre of these networks Through their function, local authorities are likely to have contacts with many people, offering them easy access to specific technical information through training and to material assistance through subsidies However, they interact most with people who are close to them, and these are farmers who are moderately to well endowed with resources They also tend to select these farmers to participate in extension activities such as training and to arrange a demonstration, as they are likely to understand the information disseminated and be able to exploit it Consequently, opportunities and advantages are likely to end up with farmers not the most in need of them Similar results have been reported in other countries where farmers with direct contact with extension agents, informal exchange with peer farmers and favourable socioeconomic conditions (that is, farm assets) also proved to be more inclined to adopt innovations (Glendinning et al., 2001; Wu and Pretty, 2004) While the interpersonal contact contributed to raising awareness and changing attitudes to innovation, greater assets enabled farmers to use formal and mass media information sources Even development projects seeking broad implication of the farmers, especially marginal farmers, collaborate with local authorities, as these cannot be bypassed in the administrative process of project building They alter the way the local authorities are involved in the activities, for instance by employing them to obtain introductions to targeted groups of farmers with whom they will work subsequently, or by having the authorities focus on targeted groups By including marginal farmers in their network, local authorities will increase these farmers’ access to information and extension activities If the methods are appropriate to their conditions, poor farmers may attend and participate in extension activities Therefore it is very important to keep in mind the structure of the village communication networks (Hoang et al., in press) and to sensitize the local authorities to the situation of farmers who are in a marginal position in these networks Furthermore, the mobilization and empowerment of diffuse, inarticulate and poorly organized rural producers into collective actors strong enough to pressure extension institutions will insure that farming communities can tap the available knowledge and institutional resources to define 122 J.-C Castella et al their own pathways in problem solving (Sayer and Campbell, 2004) In a rapidly changing environment, farmers’ organizations will have to play a major role in enhancing farmers’ adaptive capacities and production efficiency through improved knowledge management (Van den Ban, 2004) The many agencies supporting agricultural innovation should therefore better coordinate their efforts between them and with the authorities to prepare farmers’ organizations to the challenges ahead, that is, increased economic competition with producers within and outside their region in a globalization trend, development of the private sector and emerging marketing channels, more constraining environmental regulations, diversification towards non-agricultural activities, etc Education will play a key role in reforming the communication systems to cope with ongoing societal changes Farmers will have to incorporate an increasing number of parameters and actors’ viewpoints in their decision-making process They will engage in negotiations with multiple-stakeholders groups for natural resources management All these changes occur in a context of rapid development of information and communication technologies they will have to adapt to Fostering Adoption The findings of the survey highlight the importance of incentives to foster innovation adoption by farmers We found that farmers attending training courses have better access to subsidized inputs (seeds, fertilizers) This may explain why training more than other sources of information contributed to early adoption The question arises how to justify the large-scale provision of subsidies in public extension programmes Some project managers contacted in this study pointed out that farmers need to be compensated for the risk they take in adopting the innovation This is certainly true for the introduction of new technologies on an experimental scale, because in these cases the benefits of the technologies under a given farmer’s conditions remain to be confirmed and proper experimentation often requires farmers to dedicate a large proportion of their resources to the innovation Incentives in the form of subsidized delivery of inputs may provide a necessary compensation to farmers for taking risks by adopting innovation In the present system however, incentives are not likely to reach the farmers most in need of them Moreover, they may not result in sustainable development of the agricultural sector, since they distort the supply and demand of inputs and products It is thus essential to develop alternative ways of reducing farmers’ risks in adopting innovation, such as insurances and revolving funds, and to explore their feasibility and sustainability The large-scale provision of incentives also is considered the remnant of the period of collectivization when farmers were delivered the inputs free of charge so that they would produce the targeted outputs Although the subsidized supply of inputs may foster innovation adoption in the short term it does not necessarily result in sustainable agricultural development in the long run as it may stimulate farmers to implement innovations that not necessarily comply with their real needs (Peters, 2001; Shanks, 2002) At present, about 80% of the funds for the implementation of agricultural extension programmes of the ministry, the provinces and the government are input expenditures In agreement with the Asian Development Bank, subsidized input supply will be phased out in the coming years (Hoang and Nguyen, 2003) This Marginal Rice Farmers in Vietnam Uplands 123 major change in extension strategies will require subsequent changes in the management style of extension organizations Our experience showed that it is possible to enhance the entrepreneurial and managerial abilities of farmers by tapping the existing resources of the vertical and horizontal communication systems (Martin et al., 2004) Pressure must come from outside for these organizations to coordinate and integrate their activities Despite institutional divisions, which are likely to remain, an integrated agricultural knowledge and technology system can potentially address the issues raised by the most marginalized farmers Coordinating mechanisms can be developed under the conditions that a pluralistic approach to knowledge management is maintained, farmers’ organizations are strengthened and empowered, and all partners are well connected to both components of the dual communication systems (vertical and horizontal) This can be achieved by promoting flexible, informal networking at many levels with an incentive system that rewards collaboration, changes in individual attitudes towards partnership and a clear policy mandate supportive to multiple-stakeholders participation in negotiation platforms (Castella et al., 2004, 2005) The main lessons learned from our case study in Bac Kan province may serve as a good example for other provinces in Vietnam and in neighbouring Asian countries Conclusion On-farm surveys showed that the public extension service promotes technology adoption by farmers by providing training and subsidies, access to which depends on their relations with local authorities and their resource endowment It could reach marginal farmers better by focusing its collaboration with local authorities on these beneficiaries, by applying participatory approaches to address these farmers’ needs, and by looking for alternative methods to reduce risks involved in the adoption of technology Therefore, a prerequisite to both the generalization of participatory approaches and inclusion of the marginal farmers in extension programmes is to cease subsidizing technological innovations The overall picture is that the pluralism of extension models and the considerable independence of the different organizations that operate in Bac Kan province has the advantage of reaching farmers through different channels and of adapting national programmes to local specificities However, coordination mechanisms need to be better defined to make all these extension activities accountable to the local People’s Committees and as a result to the end-users: the farmers Institutional capacity building is required to empower extension agents in dealing with this diversity of partners and modes of partnership References ADB (2002) Vietnam: Agricultural Sector Development Programme Hanoi: Asian Development Bank Alther, C., Castella, J.C., Novosad, P., Rousseau, E & Tran, H.T (2002) Impact of Accessibility on the Range of Livelihood Options Available to Farm Households in Mountainous Areas of Northern Viet Nam In: Castella, J.C and Dang, Q.D (Eds), Doi Moi in the Mountains Land Use Changes and Farmers’ Livelihood Strategies in Bac Kan Province, Viet Nam Hanoi: The Agricultural Publishing House, pp 121 Á146 Available online at: B/http://www.knowledgebank.irri.org/sam/sam/pdf/ chap05_E.pdf/ 124 J.-C Castella et al Balasubramanian, K & Thamizoli, P (2003) Social Differentiation in the Horizontal Transfer of Knowledge: A Case Study from South India Journal of Agricultural Education and Extension, 9(2), pp 51 Á60 BCG (2002) Bac Kan Programmes Directory Vietnam, Bac Kan Town: Bac Kan Coordination Group Available online at: B/http://www.knowledgebank.irri.org/sam/sam/pdf/2002_BacKanPIDS_E.pdf/, accessed December 2005 Beckman, M (2002) Extension, Poverty, and Vulnerability in Vietnam In: Contribution to the Workshop ‘Extension and Rural Development: A Convergence of Views on International Approaches?’, 12 Á15 November 2002 Washington, DC: The World Bank Castella, J.C & Dang, Q.D (Eds) (2002) Doi Moi in the Mountains Land Use Changes and Farmers’ Livelihood Strategies in Bac Kan Province Hanoi: The Agricultural Publishing House Castella, J.C., Dang, Q.D & Thevenot, P (2004) Toward New Modes of Governance of the ResearchDevelopment Continuum to Facilitate the Dissemination of Agricultural Innovations in a Mountainous Province of Northern Vietnam International Journal of Agricultural Resources, Governance and Ecology, 3, pp 77 Á94 Castella, J.C., Trung, T.N & Boissau, S (2005) Participatory Simulation of Land-use Changes in the Northern Mountains of Vietnam: The Combined Use of an Agent-based Model, a Role-playing Game, and a Geographic Information System Ecology and Society, 10(1), article 27 [online], available at: B/http://www.ecologyandsociety.org/vol10/iss1/art27// Dao, T.T (1995) The Peasant Household Economy and Social Change In: Kerkvliet, B.J and Porter, D.J (Eds), Vietnam’s Rural Transformation Boulder, CO: Westview Press, pp 139 163 Foărster, E (2003) Experience with Participatory Agricultural Extension Methods in Son La and Lai Chau In: FAO (Ed.), Success Stories of Mountainous Development in Vietnam International Year of the Mountains 2002 Hanoi: FAO, pp 91 Á97 Glendinning, A., Mahapatra, A & Mitchell, C.P (2001) Modes of Communication and Effectiveness of Agroforestry Extension in Eastern India Human Ecology, 29, pp 283 Á305 Hoang, L.A., Castella, J.C & Novosad, P (in press) Social Networks and Information Access: Implications for Agricultural Extension in a Rice Farming Community in Northern Vietnam Agriculture and Human Values Hoang, T.X & Nguyen, K.V (2003) Agricultural Extension Services for the Poor A Documentation Review Hanoi: VUFO-NGO Resource Centre Jamieson, N., Le, C.T & Rambo, A.T (1998) The Development Crisis in Vietnam’s Mountains Honolulu, HI: East-West Center Kerkvliet, B.J & Porter, D.J (1995) Vietnam’s Rural Transformation Boulder, CO: Westview Press Le, C.T & Rambo, A.T (2001) Bright Peaks, Dark Valleys: A Comparative Analysis of Environmental and Social Conditions and Development Trends in Five Communities in Vietnam’s Northern Mountain Region Hanoi: The National Political Publishing House Le, T.D (2003) Specificities of Rice Seed Multiplication in Mountain Areas In: PAOPA (Ed.), Proceedings of the International Conference on ‘Scaling Up Innovative Approaches in Agricultural Development From Research & Development Projects to Broad-based Impact on Livelihood Systems, Extension and Public Decisions’, 23 Á27 September 2002 Hanoi: French Embassy in Vietnam, pp 311 Á326 Leeuwis, C & Pyburn, R (Eds) (2002) Wheelbarrows Full of Frogs Social Learning in Rural Resource Management Assen: Koninklijke Van Gorcum Leeuwis, C., with van den Ban, A.W (2004) Communication for Rural Innovation Rethinking Agricultural Extension, 3rd edn Oxford: Blackwell MARD (2002) Current Status of Forestry Training and Extension Discussion paper of Forestry Training and Extension Support Workshop, 21 Á22 March, Hanoi, Ministry of Agriculture and Rural Development MARD (2003) Farmer Needs Study Project VIE/98/004/B/01/99 Hanoi: Statistical Publishing House Martin, C., Castella, J.C., Hoang, L.A., Eguienta, Y.K & Hieu, T.T (2004) A Participatory Simulation to Facilitate Farmers’ Adoption of Livestock Feeding Systems Based on Conservation Agriculture in the Vietnam Uplands International Journal of Agricultural Sustainability, 2(2), pp 118 Á132 Neef, A (Ed.) (2005) Participatory Approaches for Sustainable Land Use in Southeast Asia Bangkok: White Lotus Publishers Nguyen, V.K., Nguyen, T.H., Dang, V.M & Vu, T.S (2002) Participatory Agricultural Extension in Thai Nguyen Hanoi: Agricultural Publishing House Marginal Rice Farmers in Vietnam Uplands 125 Obaidullah-Khan, A.Z.M (2000) From Near Famine to Export: Vietnam’s Agriculture in Transition Entwicklung und Laendlicher Raum, 30, pp Á9 Rivera, W & Alex, G (2004) Extension System Reform and the Challenges Ahead Journal of Agricultural Education and Extension, 10(1), pp 23 36 Roăling, N & Wagemakers, A (Eds) (1998) Facilitating Sustainable Agriculture: Participatory Learning and Adaptive Management in Times of Environmental Uncertainty Cambridge: Cambridge University Press Peters, J (2001) Transforming the Model Approach to Upland Rural Development in Vietnam Agriculture and Human Values, 18, pp 403 Á412 Sayer, J & Campbell, B (2004) The Science of Sustainable Development: Local Livelihoods and the Global Environment Cambridge: Cambridge University Press Shanks, E (2002) Vietnam: Agriculture and Forestry Extension and Sustainable Livelihoods in the Uplands Hanoi: Swiss Agency for Development and Cooperation UNDP (2000) Agricultural and Rural Development in Vietnam Project VIE/99/002 Assistance for the Preparation of a Socio-economic Development Strategy for Viet Nam up to the Year 2010 Hanoi: Ministry of Planning and Investment, UNDP Van de Walle, D (2002) Choosing Rural Road Investments to Help Reduce Poverty World Development, 30, pp 575 Á589 Van den Ban, A.W (2004) Changes in Society Require Changes in Agricultural Extension: Some Observations Asian Journal of Extension Education, 23(1), pp Á17 World Bank (2001) Vietnam Development Report 2002: Implementing Reforms for Faster Growth and Poverty Reduction Hanoi: World Bank Wu, B & Pretty, J (2004) Social Connectedness in Marginal Rural China: The Case of Farmer Innovation Circles in Zhidan, North Shaanxi Agriculture and Human Values, 21, pp 81 Á92 View publication stats ... close to them, and these are farmers who are moderately to well endowed with resources They also tend to select these farmers to participate in extension activities such as training and to arrange... 63 23 35 Percentage of farmers mentioning the source (HybR: n0/98, SeedP: n0/35, RTransp: n0/38) Marginal Rice Farmers in Vietnam Uplands 119 more likely to learn from farmers like themselves... more attention is paid to ‘trickle down’ flows of information from research to extension and from extension to farmers, an integrated approach to innovation shifts attention to feedback and upwards