5. PILOT FIELD OPERATIONS

5.1 Selection of Sites

5.1.1 Concept

127. The SPFS guidelines stipulate initially focusing on high potential areas defined as having good potential for increasing production and good economic potential in terms of accessibility to markets etc. Generally, these have been interpreted as areas having good irrigation possibilities, and preference has been given to these. Recently some attention has also been paid to peri-urban and urban agriculture.

128. Many of the LIFDCs have emphasised that while focus in high potential areas represents a coherent development strategy, it does not provide an early response to the needs of the majority of rural food insecure households, who are poor and predominantly found in more marginal areas. More recent SPFS documentation has recognised the need for some orientation towards lower potential areas and the intention to pilot potential improvements in all agro-ecological zones of the country (i.e. known as Expanded Phase I).

5.1.2 Assessment of Achievements

129. The initial focus on quick results in the higher potential areas has obviously influenced the process of selecting SPFS sites within the case study countries. The Evaluation Team found that one or more of the following selection criteria were used by the participating countries, particularly with respect to the water and intensification components:

• The presence of an irrigation scheme with existing irrigation infrastructure. This appeared to have been the major selection criterion in most of the countries visited;
• Accessibility - e.g. accessibility of sites to roads which could easily be visited by extension agents and farmers (China), or accessibility to markets (Ecuador and Bolivia);
• Coverage of different agro-ecological zones within the country (e.g. Ecuador and Senegal), although the selected sites by and large were still confined to the more productive areas; and
• Some sites selected were recipients of earlier development projects, leading the Evaluation Team to wonder whether this might on occasion have been used as a criterion for site selection.

130. Food insecurity at the household level did not appear to be used as a criterion for site selection in most of the case study countries, because the degree of household food insecurity in the sites selected, although it existed, was unquestionably lower than in more marginal areas (see Section 7.1).

131. The sites for the diversification component in the case-study countries often corresponded with the sites for the irrigation and intensification components (e.g. Bolivia and Ecuador). However, rather than specific site criteria being used, selection criteria often appeared to focus on specific groups of the rural population (e.g. women's groups for horticulture, small livestock or value-added post-harvest activities). This meant they were sometimes located in other sites, not necessarily congruent with the sites selected for the irrigation and intensification activities.

132. The Evaluation Team concluded that most of the areas selected for SPFS activities are likely to be the best as far as potentially increasing national food sufficiency but in terms of improving individual household food security the impact of SPFS would likely have been higher in the lower potential areas (see Section 7.1).

133. The Evaluation Team believes that the more recent reorientation of the SPFS to include some lower potential areas is welcome, not only because the needs of more vulnerable households have a better chance of being addressed by the SPFS (see Section 7.1), but because the likelihood of successfully transferring approaches and especially results from high potential areas to more marginal production environments is likely to be highly problematical.

134. In terms of transferability across agro-ecological zones the diversification component probably has more flexibility and adaptability than the irrigation and intensification activities. In this sense there has been merit in diversification activities not always having been congruent (i.e. in the sense of being on the same sites) with the irrigation and intensification activities. However, this has a downside in two senses, namely:

1. This must complicate the management and coordination of activities by the beneficiary committees (i.e. farmer groups) although, on the other hand, such activities were likely to address the needs of more needy groups; and
2. It reduces the potential complementarity and synergy explicit in the multi-faceted strategy required to address food security, particularly at the individual household level (see Section 7.3.3).

5.2 Selection of Target Beneficiaries

5.2.1 Concept

135. In the SPFS, the principal beneficiaries have been perceived as `small and emergent' farmers and their households. The SPFS concept also stresses that the programme is demand driven and is to be adjusted to each country's needs and experience. The idea is that the SPFS is a continuing learning process, which evolves and adjusts according to the experience and results achieved during implementation.

5.2.2 Assessment of Achievements

136. As the SPFS recognises the demand-driven approach and the desirability of adjusting to local needs and experiences in targeting potential beneficiaries, it is to be expected that the ways they are selected are likely to vary. In the case study countries, three different approaches have been used in selecting direct beneficiaries to engage in the demonstration activities. These are:

1. The participatory approach, when the farmers themselves in their groups or committees have nominated the demonstration farmers, generally being those they view as being capable, trustworthy and well respected (e.g. Ecuador, Haiti and Senegal), as well as having the time to attend Farmers' Field Schools (e.g. Cambodia);
2. Selection by the SPFS project team (e.g. Bolivia and Bangladesh), consisting of those they feel are able to shoulder the risks (i.e. particularly with respect to innovations that required some capital investment such as irrigation channels and equipment), and who were thought to be potentially good at disseminating the experience acquired; and
3. A more mechanistic approach, for example: in Bangladesh, where once villages and irrigated areas were selected, all landowners and sharecroppers in those areas were included in the pilot operations; and in China where, in two water control villages, all farmers were included as demonstration farmers.

137. The Evaluation Team considers that, all other things being equal, the farmer participatory selection of demonstration farmers is the best in terms of incentives offered to beneficiaries themselves, and believes it perhaps could be used even more widely than has been the case to date. However, the issue still arises as to how other farmers, attending the demonstration activities, and participating in credit groups and training sessions, became involved. It appears that in most cases, but not always, they were members of the community where the demonstration farms were, and joined because of their personal motivation and interest. Thus, more farmers have been targeted indirectly via training activities in most of the project sites, but the Evaluation Team has some doubt as to the effectiveness of such activities when they are not directly linked with implementation initiatives (see Section 5.5.1).

138. The Evaluation Team strongly endorses the extensive use of farmer groups and organisations not only for participating in SPFS activities but also in disseminating information on the results. However, it is important to understand the modus operandi for successful formation, operation, and sustainability of groups, and to understand under what conditions such groups might or might not work (see Section 5.4.2.1).

139. In the case study countries, many of the demonstration farmers are market-oriented, selling their products to wholesalers or directly in the market and purchasing a considerable part of the foodstuffs they consume. They were generally clearly above average in terms of not only economic but also in terms of political, educational and social status, implying that more disadvantaged farmers were less likely to be selected by the farmers themselves or by the project team. Targeting small but emergent farmers, especially in the higher potential areas, gives a quick and more visible result in terms of production and hence income increases. However, marginal farmers, having less potential in terms of productivity improvement, could have been more involved in the promotion of other income-generating activities (i.e. diversification).

140. Women were represented to some extent in demonstrations, although this was more likely to be the case in the diversification rather than the intensification component. While not represented to the same degree as men, the Evaluation Team felt that taking into account the social, cultural and economic context, there was reasonable representation of women. One possible exception was the case of Eritrea, where up to 30 percent of the households are, de facto, headed by women (i.e. because of the war and the border-related issue) but women only constituted about 15 percent of the demonstration farmers.

141. There was little evidence in the case study countries of specific efforts at identifying vulnerable groups (i.e. the more food insecure households) or targeting initiatives at them. To the extent that women are perceived to constitute such a group, they have been specifically targeted. As indicated earlier, women in fact tended to be particularly well represented in the diversification or value-added type activities supported by SPFS directly or indirectly through TeleFood projects which, however, are of very small size.

5.3 Selection of Technologies for Testing

5.3.1 Concept

142. The idea was from the outset that demonstrations would rely on methods and techniques that could be sustained in the absence of a project, and would not require such factors as inputs that could not be mobilised through the market, or require unreasonable additional amounts of labour.

143. Also, according to the SPFS guidelines, technologies demonstrated should not only be acceptable/attractive to farm households (i.e. technically feasible, economically viable, and socially acceptable) but should also have no negative and preferably a positive environmental impact. With reference to ecological considerations, themes have included eco-friendly intensification approaches such as Integrated Pest Management (IPM), Integrated Plant Nutrition Systems (IPNS), and the integration of trees and other crops for soil and water protection, fodder and fuel, into the farming system.

5.3.2 Assessment of Achievements

5.3.2.1 Irrigation and Water Management

144. The demonstration of improved water intake structures, gates and water distribution boxes, compacted earth or, alternatively, concrete lined secondary canals, as opposed to traditional temporary ditches, was widely appreciated and requested by the farmers in most of the case study countries. However, the Evaluation Team found that due to very limited project resources, such improvements could be evaluated only on a very limited scale and then only with in-kind contributions from the beneficiaries themselves, or with help from other larger projects. Very often, the technologies, even if relatively low-cost, were still beyond the financial capacities of farmers in the absence of credit affordable to them.

145. Demonstration of improved irrigation techniques on farm (e.g. furrow, border and basin irrigation, use of siphons of different dimensions for different crops, and of gated pipes which are particularly suitable for sloping land) in contrast to the currently used traditional flood irrigation, was more likely to be associated with Technical Cooperation Programme (TCP) activities. The Evaluation Team strongly opposes the strategy of providing equipment and materials (e.g. gated pipes, cement) to the demonstration farmers in the form of a grant, since the object of demonstrations is to encourage adoption by other farmers (see Section 5.4.2.3). Obviously grants are not feasible for all farmers and thus successful dissemination of such techniques is only likely to occur in conjunction with a credit programme.

146. In the Sahelian countries visited (i.e. Mauritania, Niger and Senegal), irrigation and water management or intensification on irrigated sites, components were not dealing with the most crucial issue: the frequent breakdown of ageing pumps, which were purchased years ago at heavily subsidized prices and for which no provision for maintenance and replacement was set aside. This issue should be addressed by the SPFS through policy dialogue at the macro- and meso-levels with credit institutions, farmers' organizations and private enterpreneurs, as well as with donors having experience with this recurring problem and the means to address it.

5.3.2.2 Intensification

147. In most of the case study countries, complete technological packages, including land preparation, appropriate water management for irrigation, crop varieties, crop spacing, fertiliser, and pest/disease control measures for the targeted crops, have been evaluated/demonstrated. The Evaluation Team found evidence that sometimes some emphasis has been placed on the rationalisation of input use, specifically through reducing the use of chemicals, which sometimes are over used (i.e. particularly on vegetables), with the help of IPM. Some work has also been done with respect to crop rotations although often the existing rotations were considered to be satisfactory.

148. The crops demonstrated have usually been those commonly grown in the area, although occasionally there were examples of new crops from outside the area being introduced (e.g. vegetables in Bolivia, fruit trees in Mauritania). Obviously, in such cases it is important if widespread adoption is to be encouraged, for such `production' oriented demonstrations to be complemented by a consumer education/sensitisation campaign indicating the merits of the new production and information on how they can be prepared for consumption.

149. Most technologies selected for field testing by the SPFS are reasonably appropriate. However, it is evident to the Evaluation Team that some technologies chosen for demonstration should have been better screened prior to their inclusion. For example, in Zambia, it was fairly obvious, given the price of maize and fertiliser, that maize hybrid in the Southern Province and improved maize in the Western Province were unlikely to yield an economic return. Liming was also most unlikely to be economic in view of transport costs, and bird problems associated with millet production were well known. In Niger, monocrop rotations introduced in one site went against existing practice and led to poor results. The technique was already known but not adopted for good reasons. A more participatory approach would have avoided such a situation. In Haiti, some varieties did not prove to be suitable (e.g. rice in terms of fertiliser levels and composition, and maize varieties) and some adjustments were required. Thus, the Evaluation Team feels that sometimes not enough care was taken in the initial selection of the technologies for demonstration.

150. The demonstration of organic farming practices was appreciated in some countries (e.g. in Ecuador and Haiti). However, there was no deliberate attempt to address the `organic' produce market because there is no form of organic product certification in either Ecuador or Haiti or indeed an explicit `organic' market where a price premium would be paid for such products. Thus, the attractiveness of `organic' production as far as producers are concerned is confined to ideological convictions and to possible cost savings in terms of not purchasing inorganic fertiliser and other chemicals.

5.3.2.3 Diversification

151. A wide range of activities has been undertaken including:

• Home gardens (i.e. containing food and vegetable crops, fruit trees and medicinal plants, and using humus/compost with worms);
• Value added activities including processing (i.e. threshing and shelling) and marketing (i.e. jam making, cleaning, polishing and packaging beans);
• Egg, broiler, goat, sheep, pig and guinea pig production; and
• Aquaculture and fish and rice farming.

Training, sometimes including, but unfortunately not always, economic analysis, has been an important activity in facilitating their introduction and helping their acceptance.

152. Although perhaps not strictly a diversification type activity, an activity that appears to be particularly promising in certain countries in facilitating such activities is setting up, under some form of farmer group ownership and management, input supply stores and product marketing outlets. The potential advantages of these activities are as follows:

• With reference to inputs, improving their availability and accessibility and enabling them to be purchased at a price discount as a result of buying in bulk;
• Improving accessibility to expensive equipment that would not be economically feasible to purchase; and
• Providing a means for temporary storage of products to combat and take advantage of seasonal price variations.

153. Examples of the above in the case study countries include:

• Input supply stores in Haiti, which are under the control of the central farmers' committees;
• Machinery and equipment stores in Bolivia under the control of farmers' groups and housed at the farms of demonstration farmers; and
• Input supply stores, combined with food grain marketing outlets in Tanzania, which are under the control of farmers' groups.

154. Some of the diversification type activities have been funded from other sources. Examples are Telefood projects relating to small livestock (e.g. Eritrea), stores (e.g. Tanzania) and production group activities (e.g. broiler and jam making activities in Ecuador). Donors have also occasionally been involved (e.g. DANIDA with reference to egg production in Eritrea).

5.3.3 General Issues

155. The issue of the financial viability of the demonstration packages appears generally to have been inadequately addressed in most countries. The Evaluation Team questions the utility of demonstrating technologies that are not likely to be profitable and/or are delivered to the farmers in the form of grants or at costs that are subsidised implicitly or explicitly by the SPFS. Obviously, the probability of spontaneous adoption by farmers is likely to be very low.

156. The Evaluation Team got the impression that there was a general paucity in terms of innovative improved technologies being demonstrated. That and the problem enunciated in the preceding paragraph imply that SPFS should liaise better not only with extension and other development-oriented projects, but also with academic and research institutions within the countries. This would not only ensure that all potentially available technologies are considered but also that such institutions would provide adequate assistance in assessing ex ante their potential economic viability. In addition, expertise in those institutions could be enlisted in helping to screen and approve any dissemination materials/leaflets prepared by the SPFS project team.

157. As discussed later in the report (see Section 7.3.3), the Evaluation Team believes that because the SPFS focuses on food security which, particularly at the household level, has a significant seasonal dimension, it is important in selecting technologies for demonstration, that specific attention is focussed on smoothing out the income/production stream throughout the year. The Evaluation Team believes this aspect has received insufficient emphasis to date.

158. The Evaluation Team believes, to the extent possible, the emphasis of the SPFS should be on demonstrating options from which farmers can choose rather than trying to deliver a complete technological package of improved technology in which one or more of the components might not be viable.

159. Although some attention has been paid to environmental related issues, the Evaluation Team believes that this has been very deficient (see Section 7.3.3). More of a systems, rather than just a commodity, approach would be of value in helping to tackle, more explicitly, environmentally related issues.

5.4 SPFS Approach

5.4.1 Concept

160. The SPFS was designed to use demonstrations and training sessions as a means not only of providing farmers and entrepreneurs with insights into possible desirable technological changes, but also to provide the opportunity to explore in a participatory mode with farmers, the constraints relating to the shortcomings of the technologies being demonstrated and how these might be overcome. This highly participatory process, which recognises the central role of the farmer as the decision-maker, is intended to empower farmers and improve the relevancy and attractiveness of the technologies being demonstrated.

5.4.2 Assessment of Achievements

5.4.2.1 Enhanced Participation and Empowerment of Local Stakeholders

161. Participation is now generally recognized as a key element in any development policy, not only as a means to foster democratic debate and encourage creativity, but also on efficiency grounds since the stakeholders need to be involved in the decisions that affect them, thus encouraging them to commit themselves to attaining objectives that have been jointly defined. In spite of the rigid design characteristic in the early days of the SPFS, national authorities were able to modify it when conditions for prioritisation of food security issues existed in the country. At the local level, the extent of participation, and particularly of women's empowerment, has been dependent on the general political orientation of the country, on the administrative culture of the organizations involved, and on local sociocultural circumstances. In a few countries, CA allowed some local participation in designing, but mainly in redesigning, activities that had already been implemented to some extent. In Ecuador, this exercise went beyond a simple list of constraints through the use of the SWOT (Strengths, Weaknesses, Opportunities, Threats) techniques in redesigning/implementing SPFS initiatives.

162. Four important examples of ways in which farmers were able to become more involved (i.e. participatory) and empowered in SPFS activities in the case study countries were through the use of:

• Farmers' organizations;
• Farmers' groups;
• Farmer Field Schools (FFSs); and
• Water Users' Associations

163. Farmers' Organizations. The involvement of farmers' organizations in the Senegal SPFS should be the model of reference for SPFS even in countries with less powerful organizations. Discussions of this model in the sub-region and across sub-regions should be encouraged and supported by FAO. From the start of the programme, full operational responsibility was entrusted to two sub-regional farmers' organizations: UJAK (L'Union des Jeunes Agriculteurs de Koli Wirndé) in the Senegal river valley and CADEF (Comité d'Action pour le Département du Fogny) in the lower Casamance, both belonging to the same federation - FONGS (Fédération Nationale des Organizations Non-Gouvernementales du Senegal). With the extension of the programme to additional sites, responsibility for its operation and financial management was given to the operational and technical group ASPRODEB/AGEP (Association Sénégalaise pour la Promotion de Petits Projets à la Base/Agence d'Exécution des Projets) of the apex organization CNCR (Conseil National de Concertation et de Coopération des Ruraux) which currently consists of 19 Federations. Micro-projects and activities have been formulated by the organizations themselves at the grassroots level in consultation with local authorities and according to the SPFS general guidelines, then sorted out at the regional level before being submitted to a national approval committee. The latter includes representatives of the Ministry of Agriculture, Ministry of Finance, National Council on Food Security (CNSA) in the Prime Minister's Office), ASPRODEB and FAO. To have entrusted farmers' organizations with the formulation and execution of projects is to be highly commended. It is to be replicated elsewhere whenever possible, provided the temptation of interference with the needs expressed is resisted by FAO, as discussed earlier.

164. Farmers' Groups. In addition to the unique case of use of Farmers' Organizations in Senegal, SPFS has made good use of farmers' groups in many countries, not only as a means of improving efficiency in terms of focussing activities on a number of farmers at one time and location, but also as a means of empowering farmers to help themselves, and to elicit their help in improving the impact/multiplier impact of SPFS activities through the dissemination of promising initiatives. In some countries (e.g. Ecuador and Haiti), farmers' groups within communities or related to secondary canal water users' associations have representatives on SPFS initiated central farmers' committees which in essence provide a coordinating and overseeing function in terms of the farmers' committees operating in a specific area, and also represent their interests in contact with outside organizations. Undoubtedly these have, and continue, to play important roles, for example, in reducing conflicts over water accessibility, providing a modus operandi for introducing group owned agricultural input stores, etc. The Evaluation Team strongly endorses the concept of farmers' groups used extensively in organising and implementing SPFS-related activities but recognises that it is important to understand when they are not relevant or appropriate. Two specific situations or examples are the following:

1. It is important to recognise under what circumstances communal group activities are expected to and are not expected to work effectively. For example, in Bolivia, there were efforts to introduce group implemented demonstration plots involving men in an area where men tended to be very individualistic in their approach to farming and marketing. Not surprisingly these communal plots proved not to be popular or feasible and consequently the SPFS eventually resorted to individually operated demonstration plots; and
2. It is important that groups need a cause or reason for staying together (e.g. engaging in an activity that requires or benefits from collective action ensuring equitable access to a common property resource such as water). Thus, when the SPFS encourages group formation, it is important to give some thought to the potential sustainability of such groups in the long run. In the Portoviejo area of Ecuador, for example, farmers' groups were formed under the SPFS that are unlikely to continue now that external funding has ceased. It might have made more sense to have used the existing water users' groups that are likely to continue and, as a result, might help perpetuate some of the SPFS-related activities (e.g. collective learning and group-initiated activities).

165. Farmer Field Schools. Farmer Field Schools (FFSs) in Cambodia and Zambia were excellent examples of farmers selecting among options offered to them as to what best fits their needs and capacities. The extension officers' role then becomes one of facilitator of knowledge building, and instead of relaying centrally contrived extension messages, they relate to locally relevant problems emerging from the FFS field study. The FFS participants (25-30 persons) meet regularly throughout the season from pre-planting to harvest, to learn about the various activities and to decide how to manage the activities chosen. In Cambodia, women actively participated in FFSs and stated clearly they found the knowledge acquired, not only beneficial for their activities, but that it also improved their social status.

166. Water Users' Associations (WUAs). Although in some of the case study countries there was, prior to SPFS, some local control of water use, the SPFS has played an important role in supporting and improving the efficacy of the approach, and in helping to introduce it in other countries. This strategy is congruent with the current developmental philosophy of handing responsibility for managing common property resources back to local communities. In the case of water, thanks to the support of SPFS in management and conflict resolution types of training, farmer responsibility for its allocation/distribution is being encouraged through local Water Users' Associations (WUAs). In several countries, SPFS has introduced WUAs as a way to ensure local ownership of the irrigation and drainage structures built, and to ensure participation in their operation and maintenance. The sustainability of WUAs is predicated on the orderly payment of water fees and the building-up of sufficient funds for maintenance and repairs. Members are being sensitised no longer to expect public funds for such purposes and/or the usual sequence of rehabilitation projects. This is most clearly manifested in the case in Cambodia where farmers, because of the turmoil they have experienced, do not have a dependency culture. However, there are potential issues that need to be appreciated and possibly addressed in encouraging such `collective' type initiatives. Three examples are as follows:

1. Individuality should not be suppressed. For example, the preference for individual water pumps in Zambia, has to be taken into account in WUAs operations;
2. Land tenure issues tend to be kept in the background and not squarely addressed by SPFS. In Bangladesh, for example, where land owners, tenants and share-croppers cultivate side-by-side in the same irrigation blocks, and relatively large landowners and tube-wells owners tend to dominate WUAs, the issues at stake are not only technical and yield-centred in nature, but also relate to profitability, social sharing of costs and benefits, impacts on land tenure, and to power relationships; and
3. The role of women tends to be rather limited in WUAs, which tend to be male-dominated.

5.4.2.2 Extension and Irrigation Services

167. Extension and other support services, largely in the public sector, have until recently, played critically important roles in facilitating agricultural development. However, in the last decade or so there have been two fundamental changes in the approach to development in the form of downsizing of government services and the move to decentralisation of development initiatives accompanied by the encouragement of local responsibility and empowerment. The SPFS has been influenced by these paradigm shifts and, as indicated in the preceding sub-section, responded in a constructive manner. Two government-sponsored institutions that, as a consequence, have been heavily influenced by the changes, have been the extension service and irrigation related departments. Specific examples of the way in which such types of changes have affected the operation and efficacy of SPFS activities are given in the following paragraphs.

168. Extension Systems. Government supported extension staff are having to adopt a less directive and more participatory modus operandi. There is, however, often an in-built resistance among extension staff, particularly as farmers manage, in their own intuitive way, complex production environments, while extension staff are trained in a more fragmented and linear manner, often giving great priority to yield maximization and much less priority to farm management analysis. In one of the countries visited by the Evaluation Team, it was observed that extension workers accepted that farmers did not follow all their advice for very good economic reasons. However, extension staff had a distinctly `top-down' attitude to the application of recommendations and did not themselves adequately question their validity (e.g., fertiliser recommendations) in the context of changing relative prices and subsidies, and in the light of specific local experience. Not surprisingly, the Evaluation Team observed that some of the field level extension staff provided under SSC also had a somewhat `top-down' orientation or approach. Undoubtedly in some cases this tendency was reinforced by language and cultural barriers, which make the application of participatory approaches, even if desired, more difficult.

169. Irrigation Departments. The engineering culture that dominated most government associated irrigation departments has, in the light of the paradigm changes mentioned above, had to change. Departments have had to divest control of schemes to local communities/groups and hence to become much more sensitive to issues relating to local participation. Increasingly Departments are having to pay more attention to water use efficiency related issues rather than just water delivery systems per se. SPFS has, in some of the case study countries, facilitated such a change in focus.

5.4.2.3 Use of Subsidies

170. The SPFS has two ways of making extensive use of subsidies, namely through providing inputs and marketing services free to participating farmers, and giving inputs at subsidised prices and interest rates.

171. Provision of Inputs and Marketing Outlets. Governments often used to play significant supportive roles in distributing inputs and marketing of products. Unfortunately, the private sector in general has failed to fill the lacuna left by government no longer providing such services. The SPFS via, for example, TeleFood funding, sometimes has helped fill the void through various types of farmer group initiatives. Free inputs are provided to demonstration farmers generally when the capital inputs required of farmers are substantial. This is, for example, often the case in irrigation structures where capital equipment amounting to 30% - 60% of total investment cost is given free, the rest being the contribution of farmers in kind for locally available items like sand, stone and unskilled labour. Free inputs are also often given in the intensification component for purchase of farm implements and post-harvest equipment, and even for purchased annual inputs such as fertiliser and chemicals, especially in West Africa. The Evaluation Team believes that making such free gifts to farmers is counter-productive to the long run sustainability of the project. Instead, it recommends that all inputs, including capital inputs, should be provided on credit. That way, all farmers, including demonstration farmers, will have to make the decision as to whether to take the risk of trying out the new technology. Of course an appropriate `insurance' mechanism can be built into such pilot activities by compensating farmers who, through no fault of their own, incur losses rather than the benefits anticipated by the SPFS.

172. Provision of Credit. As indicated above, the Evaluation Team is not in favour of giving grants, but a dilemma arises when the cash flow/savings situation of farmers does not permit the necessary investment, and no form of institutional credit exists. Consequently in the case study countries, the SPFS has often extended credit, both on an individual and group basis. It is difficult to come to any firm conclusion as to the efficacy of such efforts, although based on discussions and the limited amount of documentation available, it appears that:

• Group responsibility for repayment was usually superior to individual responsibility - and examples were found where groups were formed specifically to access credit (e.g. Bolivia), and sometimes evolved out of groups that had already come together for other purposes (e.g. demonstration activities in Tanzania); and
• Repayment rates were reportedly often quite good, although there was a tendency to forgive repayment in times of stress (e.g. bad years). Also, interest rates were invariably subsidised by the project. Thus, using a revolving fund credit approach, the usual modus operandi, will result in no funds being available for credit after a few years.

173. Subsidies in SPFS credit schemes result from two elements: a purchase or sale price that is below the market price (e.g. Haiti and Eritrea), and an interest rate that is below the going market rate, which is always the case. Subsidised input prices in Eritrea and Haiti are the result of provision of fertiliser and inputs by foreign donors to the government, which passes them on to all farmers. In such cases, the Evaluation Team recognises that it would be virtually impossible for a project like the SPFS not to provide the inputs to its farmers at the subsidised prices. However, the `true' market price should always be used in sensitivity analyses of results when the economic viability of the tested technologies is being assessed.⁵

174. The use of subsidised interest rates is currently under much debate among rural development practitioners. While classical economists, including most of the multilateral aid agencies, continue to maintain that such subsidies should not be given because of the potential of distorting the private credit markets, there is an increasing number of economists who maintain that such subsidies can be justified because of market failures which tend to drive private market interest rates above equilibrium rates, not to mention the great shortage of such credit, in most rural areas, despite progress with building up micro credit institutions in many developing countries. The Evaluation Team believes that subsidised interest rates can be justified for a project like the SPFS when it is part of national policy.

175. Credit was usually given directly by the project rather than, what the Evaluation Team considers to be a superior approach, namely being channelled through relevant financial agencies. The Evaluation Team supports the notion of group responsibility for credit but is concerned that the SPFS has not generally linked the notion of compulsory savings with micro-credit. The limited success that SPFS has had with reference to credit initiatives constitutes a weakness. This is because it not only inhibits the likely sustainability of the results achieved on the selected sites, but also because it limits the extension of the technological packages and institutional arrangements promoted by the project to less affluent farming households in areas with non-existent or feeble rural financial institutions. In the future, self-help credit groups need to be developed in partnership with national and international partner agencies having both experience and resources in the savings-credit area, which the Evaluation Team views as a critical component in the strategy for alleviating food insecurity, particularly at the household level.

5.5 Results Achieved in Pilot Field Operations

176. Since the different production components of SPFS (i.e. water management, intensification, and diversification) were sometimes executed together, component specific results are difficult to isolate, particularly with respect to the water management and intensification activities. Results are therefore discussed together in this section, although attempts are made to separate them whenever possible, in order to illustrate their differential impact.

5.5.1 Adoption of Improved Technologies Demonstrated

177. For a programme such as the SPFS, which attempts to get farmers to adopt improved technologies by demonstrating them to a small group (i.e. demonstration farmers), one can distinguish four categories of adopters:

1. During project implementation -- farmers who have been informed about the technologies (e.g. in FFSs, and decide to use them on their farms);
2. During project implementation - non-participant farmers who decide to use the technologies;
3. After the period of project implementation - project farmers who continue to use the technologies; and
4. After project implementation - non-project farmers who decide to use the technologies.

178. Systematic evidence of the degree of adoption/uptake of the technologies demonstrated by the SPFS is not available, partly because the SPFS has not systematically collected such information and partly because many of the projects are on-going, or have only recently ended. The Evaluation Team therefore had to form impressions from interviews with stakeholders during its field visits. The SPFS should make an effort to systematically document evidence of the uptake of its technologies according to the four classes of adopters given above.

179. The Evaluation Team was made aware of the first type of adopters in at least three countries, specifically in:

• Cambodia, where farmers who had attended the FFSs purchased inputs and practised what they had learnt;
• Senegal where middle-class urban households are adopting micro-garden technology; and
• Tanzania, where the technologies and methods being demonstrated were being taken up, more or less as promoted, with the exception of those for maize where line planting and use of farmyard manure and improved varieties was adopted, but the application of fertilisers and mono/sole-cropping were not, especially in the dryer Dodoma region.

180. In many other countries visited, the farmers indicated that they learnt a lot during training sessions. However, in general the Evaluation Team found that very few farmers who had not received project inputs, were able or willing to put what they had learnt into practice as they could not `afford' to purchase the necessary inputs. This is particularly true where relatively large capital inputs were required (e.g. improvement of irrigation and water management systems and purchase of land preparation or post-harvest equipment). Adoption would also be clearly constrained where market outlets do not exist or are very limited (e.g. for Nuoc nam in Senegal).

181. Although the Evaluation Team was told on a few occasions that there were cases of the second type of adopters, this could not be substantiated. However, it is important to note that this type of uptake is facilitated by the irrigation component in the SPFS. Farmers who have not directly participated in project activities may have benefited from increased water supply as a result of improvements in irrigation systems, including improved water use efficiency among project farmers. They may therefore be more willing to intensify or diversify their production systems in ways similar to those advocated by the SPFS.

182. Evidence of the third and fourth types of uptake was found in Zambia where project activities officially ended in 1998, as well as in West Africa. The Evaluation Team observed that farmers are adopting those technologies that have proved to be viable without project subsidy. For example:

• In Zambia, in Western Province a cassava variety that gave both higher yields and reduced the production period from three years to 18 months. In Southern Province, cowpeas were adopted as a garden vegetable, and open pollinated short season sorghum varieties are continuing to spread slowly. However, farmers did not adopt hybrid maize, and fertiliser was not generally applied, as it was not economic. Also, following project completion farmers went on to purchase 1,900 treadle pumps (US$ 40-50 each) on credit in an IFAD-financed project. They had also extended areas under irrigation with their own resources -- probably at least doubling their irrigated areas; and
• In Senegal, Mauritania and Niger improved rice varieties, such as Sahel 108 and 202 are widely adopted as they have demonstrated their superiority at constant cost. Their spread is further supported by bulk purchase through cooperatives or similar mechanisms. Technologies requiring higher input costs or capital costs have not been adopted, even when they have been shown to have high benefit/cost ratios in demonstrations, because of a lack of cash-flow and affordable credit facilities.

5.5.2 Effect on Household Food Security

183. Generally, households directly affected by SPFS field activities were limited in number, usually a few hundred (see Annex 3). The low numbers are understandable given that in all but one country (i.e. Senegal), the SPFS has not gone beyond the pilot Phase I.

184. In virtually all demonstrations significant increases in crop yields were obtained for the packages tested. However, it should be noted, as indicated earlier, that in many cases these yield increases had been demonstrated earlier within the countries, and that only in a few cases were the technologies being demonstrated particularly innovative. In a few instances (e.g. when incorrect fertiliser combinations were used for irrigated rice in Haiti), yield differences were not significantly different from farmers' existing practices.

185. Apart from their contribution to crop yield/production increases in association with the intensification component, irrigation activities which have involved investment in structures and strengthening of WUAs have contributed to increased water supply to farms in the dry season because of increased water capture and greater water use efficiency by farmers. This has led to an increase in areas under cultivation by participating farmers and increased crop intensity (e.g. from two to three crops per year in northern Haiti), as well as diversification into cultivation of second season vegetables (e.g. Bangladesh and Tanzania), and into fish farming (e.g. Cambodia and China). The effect is thus likely to be a positive boost to household income.

186. Although systematic calculations of actual financial and economic returns to the improved technologies tested are not generally available, it is clear that in most cases the demonstration farmers, at least, obtained increases to their household incomes due to the yield increases as well as to the subsidies involved in receiving the inputs, which were substantial in the case of irrigation demonstration farmers. However, for the latter there are clearly after-project sustainability issues linked to input availability, subsidies, credit and technical support. As already stated the Evaluation Team found very little evidence of uptake by non-demonstration farmers and results of economic and financial returns obtained by them.

187. On occasion, farmers reported that the significant yield increases did not translate into corresponding increases in household income. For example :

• In Ecuador, although the yields of some crops (i.e. beans in Ambuqui, and beans and other crops in Portoveijo) were increased somewhat compared with pre-SPFS levels, the extra return in the market place failed to materialise, because of the depressed markets resulting from implementation of Plan Colombia, thereby closing the Colombian market, and the dollarisation of the Ecuadorian currency. The major benefits of the crop intensification activities are cost reductions in terms of reduced chemical use, particularly with reference to pest/disease control through implementation of scouting techniques, and the introduction of organic farming techniques; and
• In the case of the improved rice technology in the south of Haiti (i.e. improved variety and fertiliser), the increased yields were obtained at a time when market prices were on the decline because of the impact of increased levels of imported rice in the local and regional markets.

188. The diversification component had particularly pervasive effects on household food security. Beneficiary households were able to improve their food security by improving household incomes from sale of produce, and/or through direct improvement of household food nutrition, particularly of vulnerable groups such as children and women, through the increased supply of protein-rich products like milk, eggs and vegetables.

5.5.3 Effects on Gender Equality

189. There are instances in which the Evaluation Team noted that women were benefiting from SPFS activities, especially those relating to the diversification component. For example:

• In Ecuador, women and women groups have been the major and almost exclusive focus of diversification activities. Since women are important in handling the nutritional aspects of households, improving their income potential is likely to improve the nutritional content of the diets of their households. As is shown elsewhere (see Section 5.3.2.3), a wide and impressive range of activities has been undertaken, including home gardens, and health and nutrition related activities. Many of the diversification activities appear to be highly profitable;
• In Eritrea, TeleFood projects have helped women to increase their herd/flock sizes. The package consists of improved veterinary health and feeding practices. The Ministry of Agriculture (MOA) provides some training and a committee for the programme is established in each village. The committee consists not only of beneficiaries but also of representatives from the village council and the women's association;
• In Bangladesh, Cambodia and China, the diversification component impacts women directly. Technologies reported upon favourably by these groups included the home gardens and fish production. In Bangladesh, NGOs contracted by the project for group strengthening activities trained womens' groups in pilot areas on a variety of crop diversification and home gardening techniques;
• In Bolivia, Mothers' Clubs were specifically targeted for intensification related initiatives. However, it is important to note that women do not, in the Bolivian setting, necessarily constitute a vulnerable group, since men and women within the household participate in decision making, activities, and the resulting production/income streams; and
• In Mauritania, Niger and Senegal, the diversification component mainly benefited women, whether as a result of vegetable growing, sheep fattening, poultry raising, or, in the case of Senegal, honey production.

190. Despite the examples above, it is the view of the Evaluation Team that the SPFS has not made a deliberate effort to explicitly mainstream gender equality in its country programmes and projects. The recent FAO Gender and Development Plan of Action 2002-2007 presents a suitable framework to mainstream gender equality in the SPFS. In pursuit of FAO's mission to help build a food-secure world, it aims at removing the obstacles to women's and men's equal and active participation in, and enjoyment of the benefits from, agricultural and rural development. It emphasises that a transformed partnership based on equality between women and men is an essential condition for people-centred sustainable agricultural and rural development. Three objectives of the Plan of Action are particularly pertinent to the achievement of the SPFS goals: (a) promote gender equality in access to sufficient, safe and nutritionally adequate food; (b) promote gender equality in the access to, control over and management of natural resources, and agricultural support services; and (c) promote gender equality in policy- and decision-making processes at all levels in the agricultural and rural sector.

5.5.4 Effect on Vulnerable Groups

191. As indicated earlier (see Section 5.1.2), SPFS site selection criteria led the activities to be initiated in areas of participating countries which were certainly not areas with the most vulnerable or food insecure rural households. Even in the high potential areas where the projects operated, SPFS activities did not particularly target vulnerable groups (e.g. landless labourers). Consequently, it is the view of the Evaluation Team, that the SPFS has had a minimal effect on vulnerable groups in the countries in which it has operated. In only one instance (i.e. Bangladesh) did the Evaluation Team identify a situation where there might possibly have been a positive impact on vulnerable groups. Comments by some farmers indicated the daily wage rate during the transplanting season had increased significantly as a result of SPFS intensification activities. This effect could be compounded by the impact of expanded land cultivation within the WUAs (see Sections 5.4.2.1 and 5.5.5). Such wage increases will obviously directly benefit landless labourers.

5.5.5 Effects at the Community, Regional and National Levels

192. As indicated earlier, proportions of rural households benefiting from SPFS activities in all the case-study countries were very small. Consequently, there has been only limited impact on food security at the community level, and virtually none at the regional or national level. The Evaluation Team could discern limited community level effects in a few of the countries visited. For example:

• In Tanzania, community food security improved somewhat because more food and more nutritious foods were available on the local market, including eggs, milk, meat and fish. Also, at least one of the TeleFood supported stores was buying maize at a better price than farmers could obtain in the market (i.e. due to transport costs) but at a price which they believed would enable them to realise a substantial margin, through reselling the maize back in the village market later in the year;
• In Bangladesh and Haiti, where all individuals in the selected WUAs are involved in the irrigation component, the improved canals have increased the irrigation command area within each WUA, so all landowners and sharecroppers are potentially beneficiaries of such extensions, whether or not they are participants in the project;
• In Ecuador, the potential for irrigation has been, or is being, increased, but the benefits relating to rehabilitation and extension of area irrigated cannot be solely attributed to the SPFS since there has been close collaboration with the government sponsored authorities responsible for irrigation related activities (i.e. the CRM in Portoveijo and CORSINOR in Ambuqui). These linkages have been critically important in facilitating the community wide payoff from irrigation related activities under the SPFS;
• In Haiti, WUAs have been promoted in each area with adequate secondary canals (e.g. Laverdure) and influential central farmers' committees have been established. Members of those committees have been trained, and have been successful in reducing the degree of water use related conflicts in the communities; and
• In Senegal, because of expansion of the SPFS to all agro-ecological zones, it is expected that there will be significant favourable effects at the community level, especially as a result of the diversification component.

193. Finally, because SPFS human resource development activities such as FFSs, field days, and other training activities have usually involved other farmers, in addition to those associated/linked with demonstrations, it is likely that there could have been some community, regional and even national effects, however small. Where these activities may become institutionalised in normal extension activities (e.g. FFSs in Zambia and Cambodia), the impact is likely to be greater.

6. OTHER GENERAL EFFECTS/ISSUES OF SPFS

6.1 Effect on National Policies

194. The World Food Summit has certainly raised awareness of the crucial importance of food security. The SPFS carried this momentum further by contributing to the preparation of new instruments of food policies in some of the countries visited, namely in Ecuador and Tanzania and, to a lesser extent, in Senegal and Bangladesh. The varying degrees of the SPFS contribution depend on the commitment of individuals or groups, the state of consensus building on a national food policy, the evolving food situation in the country, and/or political changes which have occurred since 1996.

195. In one group of countries visited, the SPFS has had a major effect on food security policy. For example:

• In Ecuador, the commitment of the SPFS National Coordinator and his team, along with the support of FAORs, have played a pivotal role in the drafting and signing of the Presidential Decree of January 1998 establishing a National Food Security System and making food security a state policy. In 1998, the National Assembly adopted an amendment to the Constitution explicitly declaring that food security is a `basic citizen's right'. The Planning Commission is preparing a Food Security Strategy Towards 2025 and the SPFS Coordinator contributed also to the drafting of the Food Security Law, which is currently awaiting ratification by the Congress;
• In Tanzania, the Technical Advisory Committee of the Comprehensive Food Security Programme supported the SPFS approach which, in turn, had a positive influence on food security policies. The strong national commitment to food security is reflected in the title and functions of the Ministry of Agriculture and Food Security, and in the continuation of SPFS activities after cessation of major FAO support;
• In Senegal, the high visibility of the SPFS has contributed to the setting up of the Interministerial Committee on Food Security which has, however, not met since its first meeting in 1999 under the chairmanship of the Prime Minister. Its Executive Secretariat has, to date, not yet received any funding. Following the political change in the country, the Agricultural Policy Unit of the MoA has been requested by the Prime Minister to evaluate the SPFS and the Group on Strategic Thinking (GRS) in order to contribute to the formulation of a food policy; and
• In Bangladesh, which attained national self-sufficiency in the late 1990s, the evolving food situation, and several political changes, have not contributed to maintaining the high degree of priority given to food security during the World Food Summit, as the idea of a National Plan of Action to attain food security by the year 2010 was the outcome of the Rome Declaration and Plan of Action. During implementation, however, the SPFS became one of many projects addressing agricultural and rural development issues and its small size and fragmentation in three quite separate components did not give it great visibility, including within the Department of Agricultural Extension, a large unit with more than 20,000 staff. No systematic learning about farmers' circumstances, constraints and preferences has been derived from the On-Farm Water Management Pilot Programme and, as a result, the SPFS has contributed little, if at all, to the ongoing national debate on food security issues and policy in the country.

196. In another group of countries visited by the Evaluation Team, the SPFS did not have an impact on the overall food or agricultural policy as such, but has contributed to some elements of it:

• In China, the SPFS helped support the move of the Sichuan Province authorities towards a less directive and more participatory agricultural extension system. Such an experience could be usefully directed to the most food insecure areas, particularly in Western China;
• In Zambia, FFSs have recently attracted a lot of attention and appear to have the potential for being mainstreamed among participatory approaches being used for extension; and
• In Cambodia, the relatively weak ownership of the SPFS by the national authorities resulting from insufficient flexibility in the design, is making mainstreaming of the potentially valuable FFS approach in the national extension system somewhat difficult. Nevertheless, the Evaluation Team did detect some SPFS influence as several of its elements have been adopted by other projects in specific geographical areas.

197. In the other case study countries, SPFS has had no discernible impact on national policies, which do not include yet a major and/or specific focus on long-term food security.

6.2 Effect on the Donor Community

198. The Evaluation Team considers that the SPFS has had no noticeable effect on donor strategies as such. Some initiatives and approaches might have had a little influence but not to the extent of modifying donor strategies.

199. The most intense interaction so far between FAO and a donor has been the FAO/DFID joint analytical study of the application of sustainable livelihood approaches in the SPFS. The study concluded that the underlying concepts and principles of the DFID Sustainable Livelihoods Approach (SLA) are embedded in the implementation of the SPFS in Tanzania, and recommends further incorporation of SLA within the SPFS. The SPFS is not viewed as having had an effect on the DFID strategy, but rather the reverse is the case. The study perceived a potential for further collaboration, provided SPFS:

• Develops linkages between pilot phase activities and the wider socio-economic, institutional and policy environment;
• Seeks integration/collaboration with other rural development programmes;
• Develops linkages with sectors other than farming; and
• Adopts a more systematic approach that will allow a better appraisal of the impact of programme activities on food security and livelihoods.

These recommendations, and others, if followed by FAO, would have the effect of broadening the SPFS strategy.

200. The DFID approach is congruent with the views of other donors, who generally feel the SPFS approach should be broader. Most of the donors involved in food security issues target their interventions on the most food insecure areas, households, or even individuals, and address income-generating activities or even, if not explicitly, livelihood systems. Their initial perception of the SPFS as rigid in design, `top-down' in orientation, narrowly centred around crop yields, and addressing, not the most food insecure areas, but high potential areas in order to quickly achieve results leading to national food self-sufficiency, has, in most cases, persisted even though a more flexible and broader approach has, in recent years, been adopted by the SPFS. Once perceptions have been created it is difficult to change them, particularly in the absence of a credible reporting/evaluation system. Donors' perceptions are therefore largely dependent on oral communication, with all its strengths and weaknesses, unless they visit the field itself as, for instance, has been the case of DFID, CIDA and the Belgian Cooperation.

201. When they know about the SPFS at all, donor representatives met by the Evaluation Team at the country level tended to perceive it as very small relative to their own programmes. That view contrasts with the ambitious goal of the SPFS to developing new approaches to be taken up by large investment programmes and bankable projects. FAO and Government officials tend to view the financial dimension somewhat differently. For instance, the support of the Government of Japan to the SPFS in Bangladesh of US$ 3.3 million for a period of five years (i.e. 2001-06) is considered a remarkable achievement for the SPFS by FAO. However, this annual support of US$ 0.66 million is, only a tiny part of the JICA annual programme in Bangladesh, which in 1999 amounted to US$ 226 million in grants alone. Another example, in the same country, is the SPFS component of soil testing and fertility management (i.e. US$ 0.33 million for the period March 2000-December 2002), which is part of a very large programme financed by the World Bank/IDA and DFID. Among the six organizations implementing soil testing activities, one NGO (BRAC), has a projected budget for the year 2001 of US$ 152 million. The relative financial contributions of partner agencies should not, in principle, be an impediment to influencing strategies at the qualitative level. However, for this to happen there would need to be remarkable achievements.

202. While SPFS achievements have not appeared to impact donors' strategies, they have, however, generated some donor support in the countries visited, as has been shown earlier (Table 3).

6.3 Cost-Effectiveness of the National SPFS

203. No credible cost-benefit analyses or rates of return at the project level are available. The absence of monitoring and evaluation at the project level, which the Evaluation Team was informed would divert resources from implementation, is viewed by the Evaluation Team as a weakness in a pilot project. Extension and replication of technological packages and institutional arrangements require, in addition to spontaneous diffusion, an active dissemination policy supported by credible economic/financial data, particularly if there is a desire to attract donor funding.

204. In such a situation, there is little that the Evaluation Team can say about the cost-effectiveness of the SPFS. It is not possible to estimate the cost per farmer and it remains to be seen whether the benefits accruing to the participating farmers are sustainable and over what period of time. Furthermore, data on the number of non-participating farmers adopting the techniques are not available, even in relative terms. As shown in Section 5.5.5 the Evaluation Team is of the view that there has been little impact of SPFS at the community level and none at the national level, except as noted above, in some cases at raising awareness of the crucial importance of long-term food security issues.

7. OPTIONS FOR THE FUTURE OF SPFS

7.1 Lessons From the Past

205. When the SPFS started it had what the Evaluation Team feels was a rigid design. It was also required initially that it be implemented in those areas where there was the potential for rapidly increasing production. These areas were characterised as being where there were irrigation possibilities. It was envisioned that the production focus would help solve food security problems both at the household and national levels.

206. It soon became apparent that the early `micro' oriented production focus was insufficient to ensure progress in solving the food security problem and that `macro' and `meso' type issues were important in enabling production increases to occur, and in ensuring benefits accrue to the producers. Over time, the implementation of the SPFS has become `less rigid' and `more flexible' as exemplified in the evolutionary table provided by the FAO Regional Office for Latin America and the Caribbean (see Table 5). Thus, practical realities in implementation have compelled SPFS programmes to incorporate elements in the early part of Phase I that were previously conceptualised as being dealt with in Phase II or at the earliest during the expansion part of Phase I. Such initiatives have been found to be particularly appropriate when they can be addressed at the local level without requiring changes in policy.

207. In addition to the design flaws, which became apparent in implementing SPFS activities, another problem of a more conceptual nature became apparent to the Evaluation Team during the visits to the case study countries. This relates to the likely trade offs between fulfilling the goals indicated in the guidelines for the SPFS for addressing food security at both the national and household levels. In general, the stipulation of initiating SPFS activities in higher potential areas is likely to be better in addressing the issue of improving national food security. Poverty, and hence individual household food insecurity, is likely to exist in such areas but by the same token it is likely to be less acute than in less promising agricultural areas.⁶ As a result in the case study countries, in general the sites selected for SPFS activities have been of relatively high productivity, compared with the more marginal areas where the degree of malnourishment in rural areas is higher but the potential for increases in agricultural productivity is lower. Thus, although in the opinion of the Evaluation Team the areas selected for SPFS activities are likely to be the best as far as potentially improving national food security, in terms of improving individual household food security, the impact of SPFS would likely have been higher in more marginal areas. This suggests trade-offs between the stated laudable goals of improving both household and national food security.

208. The rigid type design that characterised the SPFS at its inception had another downside in the sense that, because it was not flexible, it was not amenable to being adjusted to the specific priorities and strengths of countries and consequently sometimes probably inhibited the development of collegiate relationships and a sense of ownership on the part of some SPFS countries. It was also introduced as a stand-alone programme not linked to other ongoing or planned activities of other agencies including NGOs.

209. Another issue which became apparent during the visits to the case study countries was that the time initially planned for the pilot part of Phase I of the SPFS, namely two or three years, was far too short, and the number of selected sites were too small, to have any major impact, on production and food security strategies. Participatory approaches, if they are to be effective, take time, and this combined with annual variations in climatic and socio-economic conditions and the time required to develop sustainable input distribution and product marketing outlets, obviously all imply the anticipated time period was too short. In fact, in the visited countries, where explicit SPFS activities were still being implemented (i.e. excluding Zambia), the major focus, except in Senegal, was still on the sites and communities where such activities were initiated up to five or six years earlier. Success of the SPFS type of approach is very dependent on the strength of the institutional structures, including extension, input distribution, marketing and credit systems. Where there are deficiencies in this, it is very unlikely that a two to three year period will be sufficient to demonstrate impact. Evidence of implementing the expansion part of Phase I (i.e. extending SPFS activities to all agro-ecological zones in a country) was only found in Senegal, although plans do exist on paper for other countries. Also, there is no country that has entered Phase II of the SPFS.

210. Given the above issues, what can be done to improve the design and implement SPFS activities in the future that will improve their potential efficacy, impact and acceptance to both national programmes and potential donors? To address this, the Evaluation Team first assessed what the current strengths are of the SPFS, which help provide a useful foundation on which future initiatives can be built. Based on this, the Evaluation Team then agreed on what should be the major focus of the SPFS. This then led to a consideration as to what might be the optimal strategy for the SPFS to adopt.

7.2 Strengths of the SPFS

211. The SPFS, as it currently exists, has a number of positive characteristics or strengths, not always shared by other donor and FAO-supported programmes, that deserve recognition and can be usefully built on in designing and implementing future SPFS-related initiatives. The major ones are as follows:

• It helps nurture national consciousness about food security and in principle encourages national ownership and responsibility for SPFS-related initiatives;
• It focuses attention on agriculture, food and nutrition, which have often tended to be eclipsed in discussions concerning poverty, ignoring the fact that agriculture is the backbone of most poor countries, and certainly of the rural sectors, and that countries, not only individual households, often do not have adequate means to purchase food;
• It recognises, at least in theory, that the most efficient approach to dealing with food security issues is using a participatory approach at all levels and establishing linkages from the national to the field level and that empowerment of people (i.e. particularly farming households), to seek information and options, to make and take responsibility for their decisions and to influence outcomes, both individually and collectively, is facilitated/nurtured via the use of farmers' organizations, groups and Farmers Field Schools;
• It recognises and gives priority to the fact that in many low productivity systems, water availability is the primary natural constraint to agricultural development, but in doing so it also recognises that solving the problem of food security requires a multifaceted and integrated strategy not only involving water management but also intensification and diversification. It has also in recent years increasingly recognised the importance of water use efficiency and water harvesting related initiatives in situations where water is not readily available, for example in semi-arid/arid areas, particularly where the potential for irrigation is limited or non-existent;
• The diversification activity to supplement household incomes during food insecure parts of the year is particularly important in helping women, who often have the major responsibility for raising the children and feeding all household members; and
• SSC initiatives have strong political and financial support among countries in the South, and have received some funding from donors in other parts of the world.

7.3 Alternative Future Approach for the SPFS

7.3.1 FAO Should Prioritise Countries for SPFS Related Initiatives

212. The SPFS is currently being implemented in 62 countries. However, a major concern of the Evaluation Team is whether, given the limited resources (i.e. financial and human) available to FAO, it has the capacity to deal adequately with all the countries currently eligible for SPFS. Currently, the criterion for eligibility to participate in/benefit from the SPFS is generally based on being a Low-Income Food Deficit Country (LIFDC), although a few of the countries currently included in the SPFS do not fit that criterion.

213. The advantages of using this criterion are that:

• FAO has the mandate and the responsibility to serve all its members (i.e. rich and poor), with an accepted emphasis on the needs of the poor;
• The LIFDC criteria are clear cut, verifiable and are an internationally agreed definition also used for eligibility for non-emergency food aid; and
• It covers most of the food-insecure countries.

214. On the other hand, the problems of using this criterion are:

• There are currently more than 80 LIFDCs that are FAO members, which relates to the issue raised above about very limited resources available to FAO. In addition, even non-LIFDCs can opt to participate in the SPFS;
• The LIFDC category does not reflect the extent of malnutrition, although such figures are published annually by FAO in The State of Food Insecurity in the World (SOFI).⁷ Also, the LIFDC category includes relatively prosperous net agricultural exporters such as Egypt and excludes relatively poor countries, such as Vietnam and Uganda that have a comparative advantage in the export of cereals, coffee, etc.; and
• No account is taken of the political will for rural development which is reflected in the HIPC initiative or of the extent to which needs are already being addressed by other programmes.

215. After consideration of the above factors, the Evaluation Team has come to the conclusion that some form of prioritisation of LIFDC countries is necessary in order to prevent the limited resources of FAO being spread too thinly and to improve the prospects for impact of the SPFS. The issue is what criteria should be used in the prioritisation process. The Evaluation Team feels the criteria would differ according to whether the country is planning an SPFS project or whether the country is desirous of continuing an SPFS activity started earlier.

216. For a country wanting to initiate an SPFS activity. The Evaluation Team suggests that four criteria should be considered once a country (i.e. usually an LIFDC) has expressed interest in an SPFS activity:

1. Incidence of hunger and malnutrition in the country;
2. Potential institutional infrastructure in place to support an SPFS initiative (e.g. are extension and support services adequate in the public and/or private sector, can governmental policies allow an essentially bottom-up farmer driven participatory approach to addressing food security);
3. Availability of unexploited developmental opportunities (e.g. availability of appropriate technology options, diversification possibilities, accessibility to relevant marketing opportunities), especially for marginal areas; and
4. Potential for complementing, or integrating with, planned or ongoing national or donor initiatives.

217. For a country wanting support for continuing an SPFS activity initiated earlier. This would be based on the same criteria as those for initiating an SPFS activity given in the preceding paragraph, plus three others, specifically whether:

1. An explicit national commitment has been made to addressing food security issues;
2. Government has assumed `ownership and leadership' for the SPFS; and
3. Satisfactory progress has been achieved in terms of adoption of the principles of the SPFS approach, the results obtained, and there is potential for national and/or donor support.

218. Because of the position of FAO, as a global organization, needing to be sensitive to, and equitable in its treatment of FAO member countries, particularly the poorer ones, the Evaluation Team suggests that this prioritisation exercise should not be viewed as a means of excluding certain countries. Rather, it is suggested that low priority as far as FAO supported SPFS type initiatives are concerned should be translated into high priority for other FAO initiatives (e.g. help in strengthening and/or upgrading of the extension service), particularly for those that could help rectify the deficiencies identified as reasons for the country receiving low priority for SPFS. Also, in some situations, prior to proceeding further to continue SPFS-related initiatives (i.e. in the SPFS terminology, the expansion of Phase I) it may be appropriate for FAO to assist food security development in other ways including the policy setting, institutional capacity building for development in marginal areas, and assistance in such areas as early warning, disaster preparedness and organization of targeted food safety net programmes.

7.3.2 SPFS Should Give Greater Priority to Household Food Security

219. The Evaluation Team fully supports the emphasis placed on food security by the SPFS but as indicated earlier, it is concerned about the possible trade-offs between household and national food security. To resolve the issue, the Evaluation Team suggests consideration is given to the following:

• For countries that are not food self-sufficient it is proposed that SPFS should give equal priority to areas within countries that have good potential for increasing production as well as more marginal areas. The Evaluation Team appreciates that the expansion part of Phase I is supposed to address all agro-ecological areas in the country but it is likely that the expansion strategy will need to be implemented in steps and hence in the process the more promising areas would receive major priority. Although generally food security is interpreted by the international community in broader terms (i.e. ability to produce and/or purchase food), the Evaluation Team is sympathetic with the notion that poor, highly indebted countries often lack the necessary foreign exchange to purchase the necessary food quantities to supplement domestic food production. Consequently, national food self-sufficiency is a legitimate concern as it contributes to insulating the countries from the vagaries of the market place; and
• For countries that are self-sufficient or nearly self-sufficient in food production (e.g. Bangladesh, Cambodia and China), the Evaluation Team proposes that the SPFS focuses on agricultural areas where the greatest degree of household food insecurity exists - that is if the conditions outlined in the preceding section can be met. This means that attention will be focussed on the poorer agricultural areas of the country.

7.3.3 Factors to Consider in Designing Specific SPFS Initiatives

220. Based on the Evaluation Team's observations earlier in the report, it recommends that three basic principles should underlie the initiation or extension of SPFS initiatives in specific countries. These are:

1. The design should build on the strengths or characteristics of the SPFS as it currently exists, and outlined earlier;
2. The shift in design from a rigid or package driven approach to one that is more flexible and `people driven' based on meeting needs, grasping opportunities and alleviating constraints, should be further reinforced. Specifically, this means moving away from a focus on production to also include economic, financial and social dimensions, input/product marketing and credit related initiatives. These would be approached through giving farming households the analytical tools and means to become empowered and to be able as far as possible to influence and control their own destinies; and
3. In the design exercise, the priorities and comparative advantages of national governments and donors should be recognised and, as a result, ways should be sought to develop SPFS-related initiatives congruent with, and in partnership with, national governments and donors rather than trying to focus on marketing a fixed approach and modus operandi. This would allow SPFS designs to benefit from the experiences of development partners who would be fully associated in the programmes rather than seeking their support ex-post.

221. In the design exercise itself, there are six specific areas that the Evaluation Team believes should receive greater attention since they impact not only on the way strategies for household food security are developed but also on the potential degree, sustainability and multiplier impact of SPFS-related activities. These are discussed below.

222. Explicit consideration of seasonality. For poor households heavily dependent on agriculture as a means of livelihood the degree of food security varies seasonally. Technological recommendations have to take into account the `normality' of good and bad years. For the rural poor household food security is influenced by the agricultural production cycle, the amount of food stored, and the cash flow. This clustering of factors is most pronounced in seasonal rainfed marginal areas, but even in irrigated areas, it is a crucial issue for farming households with very limited resources. In both rainfed and irrigated areas, diversification activities not so dependent on water availability are important in cushioning households from the negative impacts of seasonality, and increasing their resilience to shock and negative trends through diversification of their production systems and income sources. The Evaluation Team believes it is extremely important to use the seasonal nature of food security, as a rational systematic starting point for designing strategies to improve household food security during all periods of the year. The components, which might appropriately be termed "Counter-Seasonal Strategies" (CSS), consist of:

• Increasing food production at different times of the year (CSS1);
• Increasing the capacity of local/community/household grain storage, with due attention given to post-harvest losses, and grain banks in order to counteract seasonal price differentials (CSS2);
• Increasing household income-purchasing power through income-generating (i.e. diversification) activities, particularly at recurring/predictable times of food shortages (CSS3a), and through the improvement of the farmers' terms of access to inputs, of the processing and marketing of agricultural produce (CSS3b); and
• Introducing/supporting self-help savings/credit groups in order to reduce the impact of resource/expenditure time lags, to allow small-scale investments, and to ensure sustainability (CSS4).

223. In the countries visited by the Evaluation Team, SPFS has mainly dealt with CSS1, not very much with CSS2, to a certain extent with CSS3a (but not in the most food insecure areas), only sporadically with CSS3b (i.e. a few food processing activities), and in a few cases with micro-credit support (CSS4). If future SPFS-related initiatives are to have a clearly visible identity and boundaries sharply defined around the issues of seasonal hunger and counter-seasonal strategies, this will require continuation of efforts to increase food production at different times of the year, to improve implementation of strategies relating to food storage, post-harvest losses, grain banks, marketing, processing and credit, and to enter into partnerships with other agencies which have relevant experience/expertise, for example in income-generating activities (i.e. including off-farm) and rural financial institutions. Mutually satisfying partnerships with other agencies require that FAO does not claim ownership and most of the limelight, and will enable FAO to focus on areas in which it has more experience. Also, much greater effort should be made to work with Farmers' Organizations (i.e. such as in Senegal), farmers' groups and communities.

224. More explicit consideration of environmental issues and ensuring congruency between production and ecological sustainability. Although the concept of the SPFS emphasises the importance of increasing production without undermining ecological sustainability, the Evaluation Team believes, at least in the case study countries, it has not always received much explicit attention. The explicit emphasis has tended to be very much on yield increasing technologies. Although in a couple of countries (e.g. Ecuador and Cambodia), there was some attention to IPM and organic production methods, there was less attention to environmental issues than the Evaluation Team would have anticipated. Therefore, the Evaluation Team recommends that in the future more explicit attention is given to designing strategies that will ensure congruency between production and ecological sustainability than has been the case to date. Attention to this issue will become even more critically important as SPFS-related activities are extended into more marginal areas.

225. More explicit attention to gender equality. The Evaluation Team believes mainstreaming gender equality in a location sensitive manner in the SPFS country programmes needs to receive more explicit attention. The recently developed FAO Gender and Development Plan of Action provides guidelines for the SPFS to mainstream gender equality, partnership with country government gender policies.

226. More explicit attention to linkages. Earlier in the report the Evaluation Team indicated the desirability of the SPFS using a more systematic, explicit and planned approach to establishing linkages, not only with donor agencies but also with other developmental actors/agencies (e.g. NGOs). Moves towards doing this need to be initiated even at the design stage since they can play a potentially important role in determining the specific character of the strategies that are planned and are feasible, and in increasing the scope and potential impact of SPFS-related activities that are implemented. Examples of two specific and potentially very useful linkages are with developmentally oriented NGOs and with research institutions. Development-oriented NGOs could sometimes provide support in implementing savings/credit schemes, grain banks, community marketing and processing facilities, and supporting diversification in terms of types of activities not mainstreamed into SPFS diversification initiatives (e.g. crafts and off-farm employment). FAO, because of its mandate, has to concentrate on agriculturally related activities, but potentially important contributors to household security or the sustainable livelihoods of poor households are non-farm sources of incomes that could be facilitated through the development of creative linkages with other agencies that are focussing on such areas. This would be coherent with the right to food, as discussed in the documentation for the World Food Summit Five Years Later.

227. More explicit attention to macro and meso-level institutional and policy issues. At the macro level this relates particularly to public distribution of food, pricing policies, subsidies, and WTO issues. Meso-level type issues include credit, finance, input distribution, market identification, and development, community negotiation, agreement and action in terms of watershed development and land tenure related type issues. All such issues will often need to be dealt with in partnership with other agencies.

228. Acceptance of a longer time period for achieving impact. A more realistic time period (e.g. up to five years) is required to adequately develop and incorporate the above considerations.

7.3.4 Implementing the Proposed Design of SPFS Initiatives

229. Given earlier comments about a more flexible design with reference to SPFS initiatives, the Evaluation Team believes there would be merit, before the design exercise per se, in implementing one or two activities, namely:

1. If the SPFS is already in the country, having a detailed independent evaluation undertaken, which would not only evaluate impact but would include suggestions/proposals for the future; and
2. At the start of the SPFS in a country, mounting a spearhead (i.e. exploratory) mission which would make no firm commitments about the future, but would informally explore the merits of initiating SPFS type initiatives based on considerations discussed earlier. If the prognosis arising is promising, then the next step would be to mount a formal design mission, which should be, to the extent possible, led and controlled by the host country, consisting of nationals, FAO representation, and potential donor representation.

230. An exit strategy as far as FAO is concerned in terms of handing over responsibility at the end of the implementation period has to be thought through in advance and gradually implemented. It is recommended that a participatory log-frame approach is used in the design exercise and that during the implementation phase it becomes a participatory dynamic log-frame, which is periodically revisited to facilitate monitoring and evaluation with respect to the objectives, indicators, means of verification/measurement, and hypotheses/assumptions/risks associated with the project. The later column is of particular importance since it relates to the policy environment and offers a bridge with the constraints analysis.

231. Finally, four complementary strategies need to be developed. These are to:

1. Increase the effort devoted to food security mapping (FIVIMS) in order to facilitate the identification of food insecure areas;
2. Introduce systematic, simple and efficient monitoring systems to improve management at different levels and independent evaluation at the project level, to glean and share/disseminate with national and partner agencies experiences with, and lessons from, implementation of the SPFS, that can help in improving later initiatives and enhance FAO credibility;
3. Assist countries in organising training and capacity building programmes in planning and project formulation; and
4. After carefully assessing the true needs of each country in terms of the level of expertise needed (i.e. low, medium or high), and matching those needs with available technical and human resources of other countries in the South, introduce SSC programmes that use small numbers of cooperants with adequate language skills, to give hands-on training to, and mentor local experts and technicians.

__________________________

5 As indicated earlier (see Section 5.3.3) such economic analysis is usually conspicuously lacking in the case study countries.

6 While the numbers of relatively poor may be fairly high in high potential areas, the type of strategy that the SPFS focuses on, namely improving agricultural productivity, may at best only benefit such households indirectly (i.e. through creating seasonal employment opportunities since many are likely to be landless). Some types of diversification activities may provide some direct benefits but, the most useful strategy for helping such households is likely to be outside the remit of specific SPFS activities, e.g. creation of off-farm employment.

7 However such data do not cover all countries, are not as reliable as LIFDC data, and are not as widely accepted as LIFDC data.