Bachmann, Lorenz B. R.: Review of the Agricultural Knowledge System in Fiji - Opportunities and Limitations of Participatory Methods and Platforms to promote Innovation Development -


Chapter 6. Improvement of the AKIS: some possible solutions

The description of the main AKIS actors and their interrelation permitted the identification of a number of problems in particular in the working methodologies used, linkages and information flow. This chapter will take a look at different attempts undertaken, to improve the AKIS and solve some of the problems identified.

The first section provides an overview of different donor projects and their strategies to overcome some of the linkage problems. The second section confers training experiences with participatory methods, and investigates the potential for institutionalising participatory methods within the AKIS. In the third section, four case studies of donor projects are used to illustrate and evaluate different old and new approaches in promoting agricultural innovation. The last section explores the possibilities of platforms as means to reorganise the AKIS and promote innovation development and diffusion in Fiji.

6.1 Addressing linkage problems

6.1.1 Measures of various donor projects

MAFF management and experts of donor projects were concerned about the magnitude and seriousness of problems encountered in the linkage workshop (chapter 5.6.1). Further aggravating was that priority problems were not clearly expressed. However, a few areas were assistance should start up were still identified at the end of the workshop. It was concluded that donor projects should facilitate activities that promote collaboration between research and extension. Projects should assist in raising awareness of each actor‘s role, and also assist to create an atmosphere that favours the adoption of new ideas. Furthermore, projects should assist in developing human resources, provide funds for activities such as appropriate publications assist in the establishment of a technology transfer unit and conduct training in linkage issues.

In the months following the workshop, strategies were refined and the first activities were begun. In particular the experts that attended the workshop fostered dialogue with their donors and MAFF. These negotiations were bilateral and focused on how each donor could best place his assistance. Co-ordination between donors was not formalised, but it nevertheless worked. The smallness of the location assisted this process.

This informal ’coalition of projects‘ had put together a bundle of many different single and also joint activities to address the linkage problems identified. The diversity of activities undertaken are best illustrated by a short description of the projects and their focal areas of intervention:

SCEP was one of the front runners for improving linkages. Originally conceived to assist in crop related fertilisation issues, the project soon expanded its services into research and extension management training, focused on improving linkages by starting-up many joint activities, promoted PRA techniques, assisted technology transfer in selected cases and provided training on all these topics.

IBSRAM was a research focused project that investigated erosion control on sloping land. The project promoted new research methodologies and attempted to involve farmers


and extension in the search for innovative solutions. The project organised several PRAs and assisted in training activities.

GTZ was active in biological control of pests and diseases. Research and extension co-operation in this area was promoted on a regional scale together with SPC. In Fiji one focus of the project centred on vegetable production in the Sigatoka Valley. Research and extension interaction was supported. Research was trained in identifying biological control measures while extension was assisted in extension approaches (e.g. mass media campaigns)<96>.

The Pineapple*<97> project facilitated research, extension, production and marketing of pineapples. The key feature of the project was the vertical integration of various actors, from farmers that were organised in a co-operative, to MAFF and to traders and exporters in Fiji and overseas.

The PRAP programme, in its first phase focused on research, shifted emphasis in the second phase towards the consolidation of research findings and extension of findings. This implied a stronger consideration of a system‘s perspective and linkage issues by all PRAP programme components. As a regional programme, PRAP was not only active in Fiji, but also in 7 other countries. This of course, limited the time experts could spent in each country. The specific focal areas of each project were:

The Farming Systems (P1)*<98> project supported participatory technology development for agroforestry systems. Technology development involved the triangle farmers, research and extension.

The Vegetable Seed (P3)*<99> project facilitated the availability of vegetable seed and planting materials and strengthened in particular MAFF linkages to the private sector.

The Information Services (P9) project strengthened the role of agricultural information centres. Improving access to information and assisting staff in writing appropriate publications (e.g. farmer leaflets) were main activities in promoting better information flow within and between the various countries.

The Agricultural Rural Development (P11) project was designed to shift the programme balance more towards extension and to assist in better linking the various actors. Participatory methodologies and management training were the main activities of the project. The PRA training course conducted as part of the researchers own field research and presented in the following sub-chapter was financed and supported in the framework of this project.

The remaining components (Coconut Hybridisation P2, Sweet Potato P4, Taro Beetle Control P5, Biometric Service P8, Tissue Culture Service P7) represented either services (P7, P8) or rather classical research linked on-station projects (P2, P4, P 5). Their impact


on linkages was small. Their main contribution may be seen in reducing the problem of ’limited technology output‘ by providing innovations and improved genetic materials (coconut, sweet potato) and better multiplication methods via tissue culture.

This long list of projects illustrates that considerable resources were mobilised to address the problems identified. Each different project attempted to address some of the existing linkage problems. This was important in order to avoid duplication of efforts. This was also necessary in respect to the rather small Fijian Ministry. Co-operating with such a large number of projects reached the limits of The Ministry‘s absorption capacity<100>.

Due to the involvement of so many players, discussion on the best approach to follow, were frequent. However, the recommendation to create a formal linkage unit to facilitate technology transfer, was quickly abandoned. Past experiences with such a unit established through the rice development programme in the late 1980s were mixed (cf. Eyzaguirre 1996, 69). Similar experiences were made by ISNAR that investigated the reorganisation of research and extension systems for several countries across the world (Eponou 1993, 17).

These considerations were the main reasons for the adoption of an informal approach. Instead of building another formal unit, it was considered the best option to create project specific multi divisional task groups. It was hoped that in this way a large number of informal linkages could be created that could last beyond the life of the projects. In other words it was hoped that informal linkages would become a strong asset for the future work of the Ministry and that these might have more sustainability than formal (but ineffective) structures. It was further thought that this was the quickest way<101> to promote co-operation between the various divisions, clarify roles between them and reduce existing tensions. It was also seen as an advantage that in this way a larger number of MAFF staff would be trained, than if only a few staff in a new unit would be trained to resume this task.

6.1.2 Impact of the measures to reduce linkage problems

During the workshop on linkage problems in late 1994 a large number of problems were identified. It was evident that not all of these problems would be solved in a short period of time.

To facilitate a good comparison of the new situation at the end of 1997 with the problems identified during the first workshop, the researcher extracted a selection of problems and included these for assessment in the follow-up questionnaire. Researchers, extensionists and experts were asked to rate the progress made with three options: (1) no difference, (2) moderate - some improvement and (3) good improvement. The results for linkage problems are depicted in Table 30 and the results for technology development and dissemination in Table 31.


Table 30: Linkage problems: Improvement 1997 versus 1994

Problems as seen 1994




No common goals




Insufficient meeting opportunities




No transfer mechanisms between RD and ED




Difficulty to raise problems during meetings




No clearly defined roles between RD - ED




Lack of knowledge of what the others do




Staff uncommitted and lazy




Lack of agreement on evaluation criteria




Lack of guidelines, responsibilities




Lack of economic advice




No formalised structure to assist linkages




Low managerial qualifications




RD and ED information breakdown to TC&S




Overall average




Scale: 1= no different; 2= some improvement; 3= good improvement.

Table 31: Improvement of technology development and dissemination

Problems as seen 1994




Insufficient technology output




Too strong an emphasis on commercial crops only




Too little farmer participation in technology generation




Too little involvement in technology consolidation from extension and farmers




Procedure for multiplication of genetic materials unclear




Insufficient feedback of technology performance in the field




Unclear process of approval and release for new technology




Poor monitoring of technology adoption by farmers




Poor evaluation of the overall process of technology generation and dissemination




Overall average




Scale: 1= no difference; 2= moderate, some improvement; 3= good improvement.


All three groups acknowledged a moderate improvement of the linkage problems (1.8-2.1). The ratings of the RD and experts showed some parallels. The best improvements (2.4-2.7) were stated regarding the problems of ’no common goals‘, ’insufficient meeting opportunities‘ and ’difficulties in addressing problems during meetings‘. RD staff saw also some progress (2.3) in the transfer mechanisms between RD and ED. The least improvements (1.3-1.9) were acknowledged regarding the problems of ’low managerial qualifications‘, structure to assist linkages and the ’RD and the ED information breakdown to TC&S‘. Furthermore, experts stated rather little improvement (1.5) in the role definition between the RD and the ED commenting on the lack of guidelines and responsibilities. ED staff with an overall average of only 1.8 rated progress lower than the other two groups. All items were classified similarly with less variation (1.6-2.0). This may indicate that improvements at the ED were less pronounced.

Improvements in technology development and dissemination showed similar trends. The RD rated the overall improvement as moderate (2.0), followed by the ED with a slightly lower rating of 1.7. The experts, were more critical of the overall improvement giving it only a 1.6 rating. Research indicated the best progress (2.4) in relation to the problems ’too strong emphasis on commercial crops‘ and ’insufficient technology output‘. The other groups did not confirm this view. The ED still rated technology output only moderate (1.8), and also the experts considered both points still as more serious problems (1.5). While the RD and experts identified some improvement on the procedures for multiplication of genetic materials, ED staff rated this problem as unchanged (1.2).

The perception on the approval and release for new technologies was controversial, as well as the monitoring of adoption and the overall evaluation of the process of technology generation and dissemination. RD staff rated all, except the last item, with moderate improvement. The experts, however, rated three areas as practically unchanged (1.3).

The differences between the RD and the ED are also noteworthy. Overall, researchers rated the progress at their division higher than the extenionists for their division. Two reasons may explain these differences. First, researchers are more exposed to donor projects and thus have more access to training. Second, as highlighted in chapter 5.3.3, the general level of qualifications is higher at the RD. For theses reasons researchers seem to be more successful in acquiring new knowledge and skills.

Reviewing all findings together indicates that some improvements in the AKIS have taken place. The first workshop on linkages revealed that there was very little co-operation between the various divisions at MAFF and the knowledge of the role and function of each other had been very limited. Meanwhile, interaction and awareness has grown. In this respect it is now certainly more valid to speak of an AKIS, as the actors are more aware of each other‘s roles and perceive themselves more as being part of a system with a joint goal. This is an encouraging result. The magnitude of change, however, still remains modest. In some areas almost no progress was stated, whereas in other areas more achievements were reached.

Evaluating the improvements made so far remains a difficult task. Considering the many deficits identified in the beginning, the progress reached may be appraised as acceptable. Developing and disseminating innovations can be a very time consuming task, and in this respect, three years is not a very long time. Breeding a new variety of crops may easily require a timeframe of ten years. Thus, several aspects will probably still require much


more time before substantial improvements can be achieved. In this respect in particular, the stages of technology consolidation, dissemination, monitoring and evaluation should be noted. All these areas were rated as still being underdeveloped. But this is not surprising, as the number of solutions and promising innovations is still low, there is less need for activity in the areas of consolidating or monitoring adoption.

In this context a wider understanding of innovation is important for MAFF. At present, research and extension at MAFF are too limited on the rather classical agricultural domains of crop and livestock production. The socio economic areas (e.g. marketing innovations, new types of co-operation, joint ventures with the private sector, better links to food processing) are far too underdeveloped. In particular in these areas innovations may be developed and promoted in a much shorter time. A reorientation in this direction was part of MAFF policies (chapter 5.2.1), but so far it is not expressed adequately in staff allocations and tasks. To improve the output of innovations, a shift in this direction is required.

Regarding the question of linkages between divisions, a look back to the working hypothesis on the intervention strategy is necessary.

H 6

Linkage problems within small organisations may be overcome by supporting participatory methods to promote informal modes of communication and co-operation. These informal modes may bypass existing institutional or hierarchical barriers.

It was thought that the increased use of participatory methods would stimulate such informal linkages, and that for this reason no specific new ’linkage-unit‘ would be necessary. In retrospect, this hypothesis appears only partly valid. Linkages have improved, but only slightly. The potential impact of participatory methods to improve informal linkages and bypass institutional barriers was overestimated. This may be interpreted in the following way. Participatory methods remained, to a considerable extent, limited to training activities of projects. These activities were short and probably not sufficient enough to establish good and lasting informal linkages. Furthermore, informal linkages cannot fully compensate a poorly functioning hierarchical system. Both systems are interrelated. Informal linkages can improve information flow, but they have less impact on resource allocations. However, in small organisations sound resource allocations, in particular operational funds and staff are crucial for achieving a high organisational output. Thus, it may be concluded that informal linkages can have a positive effect on the overall output<102> of the organisation, but the internal efficiency of the organisation largely determines the final level of output that may be reached.

Based on these experiences, it seems necessary to establish a linkage unit to promote the co-operation between the various divisions at MAFF. The main functions of such a unit could comprise:

The main function of the liaison unit would be to further improve the linkages between all divisions. The unit could provide a more formal framework to facilitate co-operation and exchange of resources between the divisions. In such a more formal setting, the unit could also contribute to the improvement of the internal information flow in the Ministry.

With respect to the AKIS functions introduced in chapter 2.2.1, the unit would take up the central function of an information centre for monitoring, storage and retrieval of all information relevant to the knowledge system. This is of particular importance, as this kind of ’grey‘ information is often not published and difficult to access. The information would become available to all actors, instead of remaining ’hidden‘ in various actors‘ filing systems. In this respect the unit could complement the existing agricultural libraries in their role as ’formal‘ information centres.

The unit could be seen as a further step to improve the overall output of the Ministry. It remains to be seen if such a step would be sufficient. Eponou (1993, 17) points out that the worldwide success of such linkage units was mixed. After all, innovation generations are an open-ended processes and can only be planned within certain limits.

6.2 Training participatory methods

The analysis of working approaches at the RD and the ED and the results of the linkage workshop revealed a clear lack of participatory methodologies in the Ministry. This was considered as a main bottleneck for the successful work of the Ministry. As already discussed in chapter 5.5.4 , positive effects of participatory methods were expected in particular through the better integration of farmers‘ problems into the priority setting of research and extension. It was equally expected that participatory methods would lead to improvements for linkages and information flow (cf. chapter 6.1.2). Thus training of MAFF staff in participatory methods was seen as an urgent prerequisite to improve the quality and output of the work of the Ministry.

This chapter describes the researcher‘s training course on participatory methods for Ministry staff. Special attention is given to the difficulties encountered in the learning and training process. The training course consisted of three cycles that build on each other. Each cycle is discussed in a separate section. The success of the training measures and the question to which extent participatory methods could be institutionalised within the Ministry is discussed in the last section.


6.2.1 First training cycle


The course started with introductions and a discussion on the expectations of all trainees. With four trainees (1 RD, 2 ED, 1 EP&S) and the researcher as 5th team member, the group had a good working size and made it possible to address individual questions very well.

The qualifications of all team members showed much resemblance. As their main agricultural education, all team members had attended a 3-year course at FCA in Fiji. Then they joined the Ministry for their further professional career. The two extensionists, 40 and 45 years old, and the RD staff member 43 years old, all had worked for the Ministry already for more than 20 years. The youngest team member (25 years old) from EP&S had graduated from FCA in 1991 and then worked on a cane farm. Two years later he joined the Ministry. Both extensionists had spent a half-year overseas in Japan in a training programme on rice production, the younger of both also training on mechanisation in Thailand. The RD team member attended a training course in Australia, while the youngest team member had no overseas experience so far. The two extensionists had worked in different locations each, but always on crop production related programmes. The RD team member started his career on animal production and later specialised in information management in the research library. The economic planner was still in a training stage at the Ministry and his main function was agricultural liaison officer with IRETA. From these initial presentations, a sound experience of the team members on Fiji agriculture could be expected.

The didactic concept of the course was focused on learning by doing. In this respect the researcher tried to keep teaching elements to a minimum. Short technical inputs on the subject or task were followed by group works to solve the tasks. The preparation and implementation of the RRA and PRAs followed a structure of 16 steps proposed by Nagel et al. (1989, 7):


Step 1:

Defining the core problem


Step 2:

Defining output, purpose and users


Step 3:

Defining research topics and research questions


Step 4:

Making the research plan


Step 5:

Defining and grouping variables


Step 6:

Defining data evaluation categories


Step 7:

Defining (proxy-) indicators


Step 8:

Defining the survey units


Step 9:

Defining data collection methods


Step 10:

Defining the sample


Step 11:

Pretest ing the survey design


Step 12:

Training for survey execution


Step 13:

Planning and organising the survey


Step 14:

Implementing and controlling the survey


Step 15:

Analysing and presenting data


Step 16:

Utilising data


These steps were used to further structure the description of activities and results of the training course.


Using the tool of a problem tree to identify the core problem was accompanied by a few initial difficulties. Condensing thoughts on a single card and structuring these in causes and effects was felt as a difficult task.

The definition of output, purpose and users with the decisions on what to research, why and for who was grasped quickly and created no major problems.

The definition of research topics and research questions was perceived as a rather academic approach. The team members were used to work only on a few problems at a given time. These were dealt with rather intuitively. Structuring a rather complex problem as the one they selected for the RRA ’insufficient knowledge of farming systems‘ made structuring an important task.

Creating the following plan of operation, however, was very easy. All members stressed that they appreciated using cards and the matrix approach to decide on activities and the respective responsibilities and time setting.

Due to shortage of time, only 9 days were available for the preparation of the field phase. The group omitted the step of defining and grouping of variables and went directly into the formulation of the interview guidelines. Prior discussions with other experts revealed the point that many MAFF staff have considerable difficulties with preparing and implementing even simple formal surveys. For this reason it was decided to keep the first RRA rather formal. This meant that a good number of closed questions were formulated. These were supplemented by open questions and sections with keywords only to enable discussions with farmers and explore the topic. Therefore, the number of tools used was rather low: formal interviews, semi- structured interviews, observation and farm/field walks. The discussion in the team also showed that most members had a preference for closed questions, as they felt somewhat uneasy with open questions and engaging in explorative discussions.

At this stage the team decided to present the questionnaire and survey design to other interested user groups in the Ministry. The presentation provided the first chance for team members to demonstrate their presentation skills. Using media such as overhead projector, flip chart and pin boards as well as speaking freely to a larger audience quickly turned out as areas that required further attention for training of all team members.


The team used one day to pretest the approach in the field and a second day to analyse the field test. Everybody enjoyed working in the field and the day analysis was used to improve the questionnaire and cut its size.

For the implementation of the RRA the team spent 3 days in the field. The researcher was surprised by the enormous diversity of farms that made it rather difficult to identify trends or patterns. The team members, however, considered the farms visited as ’normal‘. While the formal sections of the questionnaire turned out easy, the usage of the interview guidelines caused problems. Often the team members proceeded in the same style as with


the more formal questions and did not engage in detailed discussions. Quite a few of the key words remained unfilled as the team members were unable to summarise the discussion findings or simply no discussions could be invoked in the first place. Two main explanations for this deficit were identified: First, there was a lack of experience in dealing with open questions. If a farmer found it difficult to reply to a certain key word, they continued with the next topic too quickly instead of rephrasing the question or trying to find another starting point to the question. Second, the researcher rates the very similar knowledge profile of farmers and the team members as a main reason. As the phenomena observed on the farms were ’too natural‘ or ’too well known‘ for them, they had much less incentive than the researcher to question the phenomena observed. Overall, the team felt that a wealth of information was collected.

Immediately after the field phase the team decided how to analyse the information and write-up the findings. After this step the team split in two groups. For a duration of one month the researcher worked in Germany, while the Fijian team continued in Koronivia. Both parties were working on the same steps. The objective of this section was to gain some additional insights into the abilities of the team members to work individually and unsupervised, their problem solving capacities, and analysing skills in combining different questions for drawing conclusions. For this reason the guidelines for analysis of the questionnaire included some open space for interpretation.


The local team extracted the data as far as possible into hand written tables. This approach was selected, as only one team member was familiar with spreadsheet databases. Each team member was responsible for one area. The researcher compiled all data with spreadsheets.

After the four weeks of separate work, joined work resumed with a comparison of findings. Several strength and weaknesses could be identified:

Strengths: The motivation of all team members was high. Each team member prepared a report. The findings were extracted into tables and the applied basic statistics (totals, averages, percentages) were well calculated. The team also attended together with other MAFF staff a 3-day workshop on PRA that was moderated by Jules Pretty<103>.

Weaknesses: The reports were very short (2-5 pages) and grammar and expression mistakes were frequent<104>. More important, a lack of basic scientific principles was observed. A proper description of the basic facts was lacking. Probably due to difficulties in summarising and expression, the reports did not describe the actual findings but rather each individual‘s general knowledge on the topic. The team members had problems to decide which facts were important and should be mentioned and which facts could be omitted. The open space for analysis remained unused. To give an example of this open space, the plot sizes for all crops grown and fallow areas were collected. At this stage the reseacher expected that several subtotals and combinations or relations of variables would


be calculated and interpreted e.g.: total cropping size, total farm size, relation of fallow to cropping area. However, not a single of such indicators was calculated. This highlights the weakest area. With the exception of the team member from the Research Division, who managed to state a few conclusions, the other members completely omitted an interpretation of findings. They did not feel confident to draw conclusions.

For the further write-up of the findings the team developed a new technique. All members of the team sat around a single PC and discussed the findings. One team member typed the proposed formulations, while the others watched suggested phrases and added comments or corrections. Using these techniques the team managed to write-up a full report within two weeks. It needs to be noted that these group sessions were particularly tiring and the reseacher had to resume a very active role (intensive moderation of discussions, regular summarising, suggesting formulations, etc.) to promote the documentation of findings.

The findings were presented to MAFF management and interested experts on a half-day workshop at headquarters in Suva. Good visualisation of the findings with large posters and pin boards allowed a vivid discussion. This type of presentation was a novelty in MAFF and was well received. With some training on presentation skills, the team members already had made strong progress in this area.


The first RRA cycle closed with an open discussion and a short feedback questionnaire. The team members made the following suggestions for the next round:

These suggestions appeared as all very important and it was decided to consider them for the next round. A few more methodological aspects and points of observation of the reseacher need to be presented here.

The speed of learning and participation in discussions varied from team member to team member, but could be rated overall as quite reasonable. All team members made visible progress during the training period.

A matter more concerning was the low level of analytical skills. Although these also improved during the joint analysis phase, the researcher felt that they were still not adequate. Thus, it became obvious that the working hypothesis H 5 made earlier ’participatory methods are rather easy to learn tools‘ could probably not be maintained without further differentiation. The point will be taken up again after the next cycle.

For the specific case of RRAs or PRAs, interdisciplinarity and a clear specialisation of the team members is an important ingredient to gain a deeper understanding of the matters studied. Although from different divisions (research, extension, economic planning), the team members appeared little specialised on specific subjects. The technical knowledge


on agriculture was very similar for all of them<105>. This, however, is a situation that cannot be changed in short term and PRA training had only a minor influence. Consequently, it was recommended to MAFF that staff members should be given any possible support for training and additional studies to gain more specialisation and ’profile‘.

The survey preparation steps used in the first cycle appeared as suitable and helpful in structuring the tasks. Therefore, it was decided to use them again for the next cycle.

6.2.2 Second training cycle


After a few days break the new cycle started. This time twice as much time was used for the preparation and also the days in the field were doubled from 3 to 6. Because the course followed largely the same sequence as in the first round, the description here will not include all steps, but focus on important changes and the new aspects that emerged during the cycle.

As a general observation it must be said that the local team performed all steps much better than in the first round. As the topics were already more familiar, specific deficits could be given more attention and the team members could raise the areas where they still felt most weaknesses.

Somewhat problematic was again the tool of the problem tree. In particular the step to agree on a core problem was felt as most difficult.

Defining and grouping variables was done in this cycle done for the first time. This exercise helped a great deal to discuss the scope of the PRA. To reduce the volume of the work and to enable more depths, the number of research topics was cut by half. All team members found it very useful to apply this process of defining and screening variables.

The next step, defining data evaluation categories was considered rather academic. Anticipating what categories might be useful to analyse the future information was perceived as very difficult. Only the team member from research managed to formulate a number of potential hypotheses. The problem can be seen as a problem of abstract thinking<106>. The team members had problems to express their expectations on how phenomena might be related in the field. Finally, it was agreed to revisit the question in the analysis stage.


Strong emphasis was placed on the step of defining data collection methods. A good number of tools described frequently in literature were used: annual and seasonal calendars, historic development, matrices and ranking, mapping, farm walks, focussed observation, problem ranking. And a few tools were developed by the researcher together with the whole team: enterprise profitability assessment with gross margin group discussions, matrix on fallow length and rapid crop yield estimates<107>. The graphical tool, described in the methodology chapter of this study, to depict the relation between research questions and tools helped notably to explain the concept of triangulation and to guide the later analysis. In the training stage all members quickly understood the various tools.


Working with this mixed set of tools made the fieldwork very interesting and all parties were very motivated, relaxed and attentive. Farmers quickly understood the various tools and participated lively. Working with three different farmer groups separately (women, advanced and average farmers), created additional excitement and enriched the information obtained.

During the application, not all methods proved equally easy to use. Relating to the experience of the team members, these methods may be grouped in three categories of difficulty:

Very easy:

seasonal calendars, time charts, mapping, farm walks, transsects, observation


ranking exercises, matrices, rapid plot size and yield estimates


topical group discussions, gross margin exercises

Two factors played an important role for the difficulties encountered: Facilitation/ moderation skills and routine in dealing with figures. Regarding the tools in the easy category, activities started to move on their own and farmers participated without a need for much facilitation. The tools in the medium and demanding categories, however, required more active facilitation/moderation. With this skill, the team members were still rather inexperienced. Uncertainty in leading group discussions and teaching rather than facilitation could be observed at times.

A lack in routine in making calculations or dealing with figures was the reason why ranking exercises were perceived as more difficult<108>. These problems could also be felt with the plot size and yield estimate tools, where calculations had to be made on the spot in order to discuss the results with the farmers. Particularly with respect to cross-margin calculations this problem could be felt. This tool requires a good basic knowledge of farm economics and production figures<109>. Without such a basic framework in mind, it is rather


difficult to rate the accuracy of contributions made in gross margin discussions and the final results are likely to be inaccurate. This is exactly what happened during a few exercises in the field. As the team members lacked such a basic framework on production or profitability figures for Fiji, some of the results obtained were out of proportion or inconsistent<110>.


The analysis was grouped in two sections again. After the period of separate analysis the following observations on the results of the team could be made.

The quality of reports had improved considerably. All team members had prepared a typed report. The reports were more detailed and longer (9-18 pages each). The readability was considerably better. Some figures were presented as tables and one team member even prepared a series of charts.

Compared to the first cycle, where the descriptions in the report did not match well with the field data, the basic findings of the fieldwork were well described here. The presentation of findings followed the sequence of tools used.

Weaknesses were still found in the step of aggregating the finding from the level of individual tools to the level of research questions. The description covered one tool after the other, but the links and cross-checks between the tools that are necessary to validate the information and to form a final answer for each research question was lacking in most cases. A few conclusions were made, but these again, were mainly on the level of tools. Conclusions for research topics or recommendations for the overall survey were not made. Another weakness was that unclear or contradictory information from different groups was not described as such. The information was either omitted completely or only that part of information, which was perceived as more common was presented. Given the fact that reality often is contradictory, this strong filtering risks cutting important information.

These points highlighted a more general underlying problem. None of the team members had a real scientific education. The FCA diploma is more equivalent to a technical school degree and still below a scientific degree such as a Bachelor or Masters degree obtained at a university. Abilities such as working independently, planning and conceptual thinking or problem solving capacities are not promoted sufficiently during such a technical college course. The lack of this scientific base was certainly one of the important reasons why the team members had difficulties with the tasks of combining various sources of information and drawing conclusions.

The further joint analysis focused on reducing some of the above weaknesses. Emphasis was placed on drawing conclusions on all research topics and formulating


recommendations. The more difficult tools were discussed in detail, calculations were revised where necessary and some training on economics was given.

For the final write-up the team used the team write-up approach again. Three weeks were required to complete the report. These group sessions were again long and tiring. This stimulated reflections on how report writing could be simplified. This will be discussed in the evaluation section.

The findings were presented to MAFF management and interested experts on a half-day workshop at the main extension offices in Nausori. The audience was large with a good number of extension staff present. Again the team members managed to further refine their presentation skills and a very vivid discussion followed the presentation. The findings were valued as very useful and this recognition from senior management staff and other experts created a big boost for the morale of the team.


Analogue to the first cycle, the second cycle closed with a short evaluation questionnaire and an open discussion. Comparing the two cycles, the team members commented that they considered the new cycle as both “more interesting“ and “more specific“. It was also seen as “a lot easier as it was a follow-up.“

The understanding of the various steps applied can be rated as reasonable to good. Comments ranged from “yes, and can realise their importance“ to “basically yes, but need more practise“. Some steps required more consideration, while others took less. All steps were considered as useful as they “support each other“.

The overall time allocation was considered “just okay“ or “good for novice practitioners“.

The volume of research topics covered was considered as rather high again. As reason it was explained that “we had more sub-research topics and variables, which was still again a lot of work “.

Individually, the confidence to work with the approach was not yet high. “More experience“ and external support was required as one team member mentioned “I still need the assistance of a facilitator“. As a team they already felt stronger: “team work helped in gaining experience and confidence“. And on the ability to carry out a PRA survey as a team, one team member expressed confidence “certainly yes, we will give it a go“.

Suggestions for improvements of the last cycle included a number of issues: deeper consideration of systems research aspects, more field work and again inclusion of village settings, training on report writing, computer classes and exposure to other countries.

These comments of the team members do not require further elaboration. It was clear, however, that not all suggestions could be taken up. The main consequences drawn for the last cycle are summarised here:

6.2.3 Third training cycle


For the definition the core problem a new method was used. Similar to the assessment of farmers‘ problems in the last PRA, the team made a brainstorming session together with two resource persons and collected a list of problems. These problems were then discussed and ranked according to importance. Some of the problems were later reformulated into research questions. This procedure proved much faster than the work with a problem tree and the team members preferred using this approach. For providing the direction of the survey, the method appeared equally useful.

While the first two cycles examined the farming systems as a whole, the new PRA focused on just one aspect: farmers‘ seed requirements and preferences. This narrow focus permitted a very detailed look at the matter and it facilitated concentration for everybody, as fewer aspects had to be considered. Another consequence was that the formulation of research topics and research questions was easier.

The team worked out the research plan and then drafted the organisation of the survey (steps 4 & 13). The organisation of the survey was then finalised together with the field staff in Sigatoka. A decision on survey units (step 8: various farmer groups) and samples (step 10: no. of villages / settlements) were also taken at this early stage. Visualising the results on cards facilitated this step considerably and some more time could be saved<111>.

The definition of variables was combined with the definition of tools. Working on those two steps in parallel made the process more creative. The team members could contribute from both ends: on the one side thinking about potential questions and on the other side reflecting about the tools and what information these tools could provide in answering


questions. This made it easier for the team members to contribute ideas. A reason may be that this approach required less conceptual thinking. One reason certainly was the increased familiarity with tools and the planning process as such. As all steps were well known and their purpose was clear, it become easier to change the sequence or try out modifications. The steps of defining data evaluation categories and proxy indicators were dropped as these proved difficult and little rewarding in the last round and it was expected that they would add little in this new explorative phase.


The fieldwork of the last PRA ran very smoothly and the atmosphere was the best of all cycles. The different farmer groups participated very actively and the days were very lively. The work was divided in four sub-teams and in each sub- team a mix of 2-3 extension and research staff participated. On the first day the team members facilitated the tools and explained them to the participating staff. On the following days, staff took over and facilitated the exercises while the team members play a more observing role.

It was again interesting to observe that the tools more or less worked by themselves. The new staff copied the skills very quickly from the core team members and managed to facilitate the various tools correctly. The difficulties encountered with the tools were similar to the ones observed in the last round:

Very easy:

observation, seasonal calendar, tastes ranking.


problem ranking, matrices variety ranking, semi -structured interviewing.


gross-margin exercises.

To explain the difficulties, the two factors determined in the last round (facilitation/ moderation skills and routine/experience in dealing with figures) also played the major role this time. Regarding the tools in the easy category, activities started to move on their own and farmers participated without a need for much facilitation. The tools in the medium and high categories, however, again required more active facilitation/ moderation. Here farmers had to think harder to reflect about their situation and this slowed down discussions. In this situation, moderation skills were required to keep the discussions going: e.g. providing examples, offering knowledge inputs, summarising findings or questioning facts. The team members had again improved somewhat in this area, however, weaknesses were still observed. This was also the case for the local staff that participated. However, the fact that now 2-3 staff members assisted in each group work led to spontaneous co-facilitation/moderation. They helped each other out and that greatly improved the quality of the moderation, made the sessions richer and livelier.

The second factor, lack in routine and experience in dealing with figures and making calculations, also played a significant role. It could be felt to some extent in the ranking exercises, however, co-moderation also helped here. But in particular with respect to cross-margin calculations the old problems could still be felt. This highlights an important area were further qualification of MAFF staff will be required.

At the end of each field day, all participants gathered in the conference room of the research station to discuss the findings of the day. This was very important to enable the exchange of experiences between the various groups. On the last evening a short


evaluation questionnaire was distributed. The outcome is discussed below under the evaluation heading.


In this cycle no separate analysis steps were used. The findings were compiled by the whole team within only three days and presented in another half-day workshop.

To achieve this speedy analysis the team split in sub-groups, each working on the results of a single tool. The individual findings of each tool were then compiled into large tables to allow for a comparison on all levels (single locations, farmer groups and overall)<112>. Other tools not in the matrix format (e.g. key problems, their causes and solutions) were attached in groups on the wall of the office to allow a visual overview and to enable reading them quickly in sequence. This compilation process was rather easy for everybody involved, as it had been practised also in the prior cycle.

Then, the whole team discussed the data for each tool. These discussions proved at times quite difficult. While the findings were often rather conclusive on the level of a single location, findings between locations or farmer groups often varied in a number of cases. This made the identification of trends and conclusions more difficult. One reason for the difficulty may be associated to the many ’anonymous‘ figures in the table. Reality suddenly was compressed to a bulk of more or less conclusive figures. At this stage again facilitation of the discussion became very important. Each of those ’inconclusive‘ figures had to be traced back to its location and context. This made the figures even more real. Looking at the specific location, the tool and the group, thinking about the overall context, helped to decide if something was a special case, or only a minor deviation from the overall trend (e.g. if women ranked lack of water number one problem and men only number five it had to be decided if this was a serious or minor difference of opinion). In this respect a good number of cases were rather straight forward and conclusive, but quite a few cases remained where this was not as easy. In these cases, the group had to make a decision. This certainly also contained a degree of subjectivity. However, this was the only way to condense this enormous amount of qualitative information to a manageable amount of results. This step of decision making was an area where the team members faced some difficulties.

Group decisions and findings were documented immediately. To reduce the volume of the report, the description of the basic data was reduced to a minimum. Only important aspects were described in a few sentences. This was then followed by the conclusions and recommendations. Using this approach the report had only about ¼ of the length of the previous reports. This size seemed more appropriate; both in terms of time requirement for report writing as in making it easier for the readers.

The final presentation of results to MAFF management and interested experts was again a well-visited event. In the meanwhile the team members had reached a very good level of presentation skills and managed to speak very openly. This also added to a good reception of the findings and several management staff expressed that they were surprised about the


quality of information that could be generated with PRA tools. Thus, the workshop helped to convince quite a few rather sceptical staff of the utility of participatory methods.


The evaluation of this cycle was done after the completion of the fieldwork and included a short questionnaire that was distributed to all staff (16 persons including core team) who had participated in the fieldwork.

The new participants learned how to apply the tools very quickly. One good reason may be that the majority of them (9/13) had already participated in an earlier PRA training course run by SCEP. A small group (3) even mentioned that they had already started to integrate some tools into their routine work.

A question regarding the difficulty of the PRA tools used, revealed the same grouping as for the last PRA.

All participants enjoyed working with different farmer groups. While group work was familiar for them, subdividing farmers in advanced and average groups as well as working with women separately was a novelty to them. A few mentioned explicitly that this subdivision contributed very much to gaining a better understanding of problems.

To gain an idea of to which extent conceptual aspects of PRA were retained, the participants were asked to define the term ’triangulation‘. Half of the participants (8/16) were unable to say anything. The other half at least partly understood the concept. Thus, weaknesses in conceptual thinking, as found in the previous cycles, also manifested themselves in this larger group.

The participants showed good confidence in applying PRA tools, and a small group were already able to give good examples of how they could use PRA tools in their daily work. All participants were interested in more training on participatory methods.

Overall, the training results reached were somewhat below the initial expectations of the researcher. However, they indicated that the direction was right and considerable training successes were visible. The good atmosphere that the PRA tools created definitely was encouraging. All participants were convinced of having learned something useful that they could apply in the future.

A few more aspects discussed during the final presentation of the report shall be mentioned here:


6.2.4 Review of the training process

Assessing the success of training courses is no easy matter. It should be noted that any judgement on the training performance is only subjective. To reduce this element of subjectivity, different perceptions are used here. At first, the own perceptions of the team members are discussed. Then, this is compared with the perception of the researcher. In addition, the views are matched with the experience made by other experts<113>. A factor that makes these comparisons more delicate, are the different cultural backgrounds<114>. In this respect, it is important to stress that there is no right or wrong perception. Seen by itself each perception certainly is justified, and viewed in comparison, they provide a more complete picture of reality, or more prudently, of what reality might look like.

The training success is viewed according to its short-term and medium term achievements. Short-term achievements refer to the knowledge and skills learned in the course. The medium-term achievements examine the progress made after two years.

The above aspects are considered in the following evaluation. The first sub-chapter examines the training results reached immediately after the training. The second sub-chapter investigates the situation two years later. Evaluation after the completion of the training course

Alike the evaluation in the previous cycles, a feedback questionnaire was used to collect the views of the team members on the overall success of the three training cycles. In Table 32, the team members rated their abilities before and after the course on a scale of zero to five. Zero corresponds to no knowledge or skills where as five corresponds to an excellent level.

Overall, a marked improvement in all skills could be observed. As highlighted in the previous evaluation sections, the team members were improving very well from cycle to cycle. For most of the criteria listed, the team members rated their abilities prior to the training course as poor to fair (1-2). Experience with matrices was rated as the lowest item with only 0.5, while the presentation skills were already considered as medium.

After the training, most abilities were ranked two to three points higher in the good to excellent categories (4-5). The criterion with the best rating was ’use of RRA/PRA tools in the field‘ while the lowest rating was attributed to ’literature search‘<115>. Overall the differences were rather small and the range between the lowest and highest value only just exceeded one point. Excluding the minimum and maximum values, this suggests that all abilities are on a similar level with little deviation in the category ’good‘.


Table 32: Improvement of abilities during the training course




Familiarity with visualisation methods



Organise a survey (plan of operation, etc.)



Carry out a literature search (key word definition etc.)



Modify RRA/PRA tool to suit a problem situation best



Use of RRA/PRA tools in the field



Use of matrixes



Apply ranking procedures



Analyse information collected in field survey



Compile summary tables of information collected



Interpret survey information



Drawing and formulating conclusions



Report writing



Present information at meetings



Knowledge on farming systems in Fiji



Confidence to train other staff in above techniques



Overall average



Ranking scale: 0=no knowledge, 1=poor, 2= fair, 3=medium, 4=good, 5=excellent.

These ratings, that represent the average of all team members, are somewhat misleading as they provide the impression of very limited differentiation. The actual existing differences came out clearer with two further questions on what worked best and what was most difficult.

Under what worked best the team member from the RD mentioned “visualisation, data analysis and report writing“. The two extensionists wrote “the RRA/PRA tools in the fieldwork with farmers“ and the “problem analysis done in the field“. The team member from EP&S preferred the “working with matrices“ and the “presentation of information“.

As most difficult the RD staff member specified the “organisation of the field activities“. The two extensionists mentioned “gross margin analysis with a lot of figure work“ and “to put those facts and findings in writing in a proper manner“. Similar the EP&S staff member wrote “analysing the information collected in the field“.

These statements showed individual differences better and indicated an influence of the work place. The two extensionists had difficulties in analysis and writing, while the RD staff member found the organisation of field activities more difficult. This indicates that the team members had more difficulties in those areas where they still had the least exposure and practise.

Comparing the team members‘ assessment of the training course with his own observations, the reseacher largely agrees to the above self-assessment. In respect of the ranking results presented in Table 32, a judgement appears more difficult. Here, a certain overoptimistic assessment of the abilities appeared to be the case, considering that these were rated as good to excellent. The researcher would rather rate these a full category


lower in between medium to good. This view was also supported by the other experts with whom the researcher discussed the training process in Fiji<116>.

The team members‘ abilities may be summarised in terms of strengths and weaknesses as such:


Organisational skills, field use of tools, practical implementation, presentation and visualisation skills.


Conceptual thinking, analytical skills, interpretation and report writing.

Quinney (1994, 87) made similar differentiated experiences with PRA trainers in Sudan<117>.

To judge the implications of these findings, it is important to recall the intention of the PRA training course. Designed as a ’train the trainer‘ course, the idea was that the team members would play an active role in training other staff at MAFF to achieve a rapid spread of the methodologies. It was further assumed that the increased application of PRA tools by researchers and extensionists would lead to better understanding of farmers problems‘ and consideration of those problems in the process of innovation development and diffusion<118>.

The final validation of the success of the training at this stage seemed premature. The researcher discussed the findings with the PRAP P11 Teamleader and two other PRAP experts. In these discussions the following preliminary conclusions and an approach for further action emerged:

Shortly after the course, the researcher returned to Germany. Further process was then followed mainly via reports and email. The next chapter highlights these details. Longer term training impact two years later

In 1996 and 1997 several donors continued to support work with participatory methods. According to the experts that participated in the follow-up questionnaire, 6 short RRAs/PRAs were implemented at different locations. Four other workshops, which also made use of some participatory methods, focused on strengthening planning and management capacities at research, extension and top management level. These activities provided some further training opportunities for MAFF staff.

To assess the progress of the team members in their abilities in using PRA tools, a list of basic PRA principles by Schönhuth and Kievelitz (1994, 7) was used in the follow-up questionnaire. Once again, the team members were asked to rate their own abilities on a scale of one to five. These results were compared with experiences that other experts made in training MAFF staff. The findings are compiled in Table 33.

These findings are very similar to the first assessment directly after the training. This is most evident in the equal overall average rating (4.1) of all questions. The team members


rated their abilities overall as high in the category ’good understanding‘. The fact that the categories are not identical makes the direct comparison a little more difficult<120>.

Table 33: Rating of abilities of MAFF staff in PRA principles

Key principles

PRA team (n=4)

Experts (n=5)

Appropriate instruments






Adapting instruments






Visual sharing



Avoid biases



Learning in the community



Optimal ignorance



Overall average



Scale: 1= concept/ method not understood; 2= incomplete understanding, still major difficulties; 3= reasonable understanding, assistance required; 4= good understanding, little assistance required; 5= full understanding, can work independently.

Very interesting to note, however, is a shift in some of the abilities. The team members now rated the planning and concept related criteria ’triangulation, appropriate tools and adapting tools‘ as the areas with best understanding. The more field related skills such as ’learning in the community, visual sharing and avoid biases‘ fell a little in behind. This may indicate a progress in learning. The more difficult parts of conceptual thinking and planning took more time to be understood and absorbed. The field skills, which were learned quickly in the beginning, rather consolidated than improved further. This may also be explained by the fact that field application of PRA tools remained rather limited, as will be shown later<121>.

The expert rating results confirm the researchers own observation made during the course very well. The field abilities such as ’learning in the community, visual sharing and avoiding biases‘ were rated as stronger areas, whereas the abilities related to developing concepts and planning ’appropriate tools and adapting tools, triangulation and optimal ignorance‘, were rated as weaker abilities. Overall, the experts rated all abilities in the medium category ’reasonable understanding, still assistance required‘. This is a full point lower than the self-rating of the team members and confirms the earlier observations made by the reseacher. Regarding the general abilities of MAFF staff, the experts equally confirmed the observations made earlier. ’Speed of learning‘ and ’self confidence‘ were rated as good to excellent, while ’report writing skills‘ were rated as poor and dealing with figures and calculations was seen as satisfactory (see Annex 3).

The above rating results of the team members and the experts do not match completely. The team members rate their abilities higher than the experts. This is certainly a normal behaviour. As differences in perception always exist, a final clear-cut answer cannot be given here. However, the following lessons can be drawn:


The working hypothesis H 5 formulated at the beginning of the research phase ’participatory methods are rather easy tools that can be learned quickly‘, has to be considered as a too simplified assumption. While it proved correct, that in particular PRA field tools were learned quickly, it was also shown that the approach required conceptual and planning skills which caused considerable difficulties at the beginning of the training. After 2 years of further exposure to the approach, these abilities also had improved. This showed that training participatory methods has to be considered as a long-term issue. Several factors such as the intensity of the training, opportunities to practice the new skills, availability of backup training support and favourable institutional support determine the speed of learning. Overall, a timeframe of at least two to five years appears as necessary to ensure that the training measures are successful. Consequently, integrating participatory methods in an organisation cannot be achieved in a short time, but represents a challenge that requires long term attention.

A good indicator for the success of training is also the rate of application of the training contents in practice. The use of participatory methods at MAFF will be examined in the following chapter.

6.2.5 Institutionalisation of participatory methods

The question of institutionalisation of participatory methods is of crucial importance in the evaluation of the various support measures undertaken. Therefore, these aspects were given ample room in the follow-up questionnaire. The main indicator for the success of newly introduced methods is their rate of application in daily work. For this reason MAFF staff were asked to state to which extent they make use of PRA tools in their work. The results are presented in Table 34.

Table 34: Use of participatory methods

Frequency of PRA use in %

RD (n=8)

ED (n=18)

in %

in %

Not at all



Very little






Only one Officer in each group stated that they did not use PRA tools at all. The majority in each group (50 % RD and 67 % ED) stated that they used the methods very little. 38 % at research and 28 % at extension indicated that they used participatory methods regularly.

The fact that about 1/3 of the staff already used PRA tools regularly indicates quite a success. This shows that the process of institutionalisation is on its way. The fact that about 2/3‘s of the staff seldom use the tools, however, also shows that the process is still vulnerable and that it will still require some time before the process will be consolidated.

A more sceptical view on the level of practice of participatory methods at MAFF was raised by the experts: “very little (usage)“, “only a small percentage“ and “only approached by a small number of people who have been actively involved in the past and ongoing activities, hardly by policy or decision makers“. More optimism was indicated by the following reply: “very little, but struggles on!

This trend was also visible in the answers to several other questions. Table 35 presents the use of participatory methods at selected application fields by research and extension. Overall, the rating showed for both divisions that the new methods were applied only ’at times‘. The category ’participatory follow up activities‘ was ranked the most applied option (2.6) by the RD. This indicates that some degree of on-farm research and testing is now ongoing. The highest rating (2.8) at the ED was attributed to ’identification of farmers‘ preferences and attitudes‘. The low rating of 1.4 for ’technology validation‘ by the ED and the higher rating by the RD (2.3) gave an indication that co-operation between the divisions in this field of work is still weak. It may also be an indirect indicator that the numbers of research programmes that have reached this stage is still low.

Table 35: Use of participatory methods in selected application fields




For determining priorities with farmers



General data collection, statistical purposes



Identify farmers‘ preferences and attitudes



Participatory follow-up activity



Technology validation, testing



Overall average



Scale: 1= rarely; 2= at times; 3= moderately often; 4= often; 5=very often.

The question of institutionalisation of participatory methods was examined in three ways. First, the question to which extent participatory methods were practically institutionalised within MAFF, was posed to the PRA team members and the experts. Using a ranking scale with four categories (not at all, partly, medium and fully) the PRA team members and experts were asked to rate the degree to which participatory methods were accepted and institutionalised in the Ministry. The team members opted equally for medium and partly, while the experts all ranked the process as only partly achieved. Thus, these ratings corresponded well with the findings of the above question of frequency of PRA use.

Secondly, a theoretical concept of Pretty and Chambers (1993, 13) was utilised to review the degree of institutionalisation. The concept puts forward that three conditions must be met to achieve a sustainable institutionalisation of participatory methods: active participatory field activities, interactive learning environment and institutional support. The team members, experts and management were asked to indicate to which extent these conditions were fulfilled in Fiji. The results are presented in Table 36.

Table 36: Framework conditions for participatory methods


PRA team (n=4)

Experts (n=5)

Management (n=2)

Active Participatory field activities




Interactive learning environment




Institutional support




Overall average




Scale: 1= poor; 2=marginal; 3=partly; moderate; 4=good; 5=fully met.

The results showed a controversy. While the team members considered the conditions as, overall, only marginally met, management<122> considered the conditions as moderate to good. The experts‘ rating was closer to the rating of the team members and ranged in between. While all groups shared a similar view on the PRA field activities (2.6 -3), the main differences were seen in the interactive learning environment and institutional support. Both were considered as moderate to good by management, while the other groups rated them as marginal or even poor.

It remains very difficult, to judge which perception comes closer to reality, but the views of the PRA team and the experts appear more substantiated when considered in the context with the third group of questions on institutionalisation. The three groups were requested to list the concrete consequences to which the introduction of participatory methods had led so far.

The team members replied to all options with a clear “no“ or “nothing that I know of“ and added “there is no effect as MAFF is still adopting old methods and policies. There is still no (PLA) unit to date“.

Both management members left the question blank. This could be interpreted as an indirect admission that there were no consequences<123>.

The experts made more detailed comments in respect of each question that are listed below:

“There have been policy changes which have re-emphasised on commodities." “No, see CDF<124>." “The main consequences was the introduction of some simple tools, which remain under-utilised".

“No". “The priorities remain the same, but these are better understood and with stronger consensus".

“No“. “The priorities remain the same, but these are better understood and with stronger consensus“. “Yes there are efforts to implement further training for all staff; to perform rural development work accordingly“.

“Few recommendations". “No glue, must ask extension staff". “Waibau and Nadi<125> yes".

“Zero". “Definitely no, the contrary". “MAFF did not follow recommendations, e.g. on seed quality".

These comments may be grouped into three categories. The first group stands for no impact (e.g. no, zero). A certain disappointment could be felt in their comments. The second group could be labelled unsure (e.g. questions left blank, “no glue“). The third group saw some impact in specific areas. This impact was seen rather on a field level than at a policy level.

These different perceptions indicate that the process of institutionalisation participatory methods is certainly not homogenous throughout the Ministry. This may explain to some extent why some people saw progress while others did not.

However, a major reason for the rather limited institutionalisation of participatory methods must be seen in the weak support or in other words passive role of management. On the one hand the approaches were not hindered or blocked, but on the other hand these methods did not obtain sufficient active support or encouragement. It is not sufficient to provide an opportunity for experimentation with participatory methods. The space for experimentation is only used by a few staff. To ensure full-scale use of the methods, it would be necessary for the methods to be promoted more actively and integrated into the existing work plans and job descriptions of staff.

Overall, it may be concluded that participatory methods now play a small, growing role within MAFF, but clear official policies are still lacking. One consequence of the lack of policies is, it could be said, that participatory methods are still in an experimental stage. The experiment is ongoing, but although several results are available, no decisions for the future of the experiment have been made. Unless promoted more actively, it remains uncertain if participatory methods will fully become regular working tools within MAFF. Considering the fact that the initiative for participatory methods was largely donor driven, there is a considerable risk that the efforts will not be sustainable, once donor assistance has been be terminated.

6.3 Comparison of working approaches: selected study cases

This chapter looks at the Fiji AKIS with the help of a few practical project examples. The cases illustrate examples of projects that were implemented by MAFF in co-operation with different donors. A selection of four cases are described according to the following structure: brief history, project goals, research and extension approaches, farmers‘ participation, configuration of actors involved, and finally a comparison of results achieved and main weaknesses. The lessons from these examples together with an attempt to summarise AKIS-relevant features are outlined in the last section.


6.3.1 Sigatoka Valley Rural Development Programme (SVRDP)

The Sigatoka Valley Rural Development Programme (SVRDP) was a large scale integrated project funded as a loan by the Asian Development Bank (ADB). The project began in 1987 but was terminated before the completion date of 1990. The project was designed in the mid 1980s when national policies focused on import substitution and increased production. This, together with export of surpluses, was expected to produce an impact on macro policy level (e.g. food self-sufficiency, balance of payments). The development goals of the project were:

To achieve the above goals the project made large investments within the rural area of the Sigatoka Valley. Focal activities were irrigation and drainage development (bore holes, pipes, sprinklers) for vegetable production and improvement of the transport infrastructure with rural access roads. The activities were implemented by a consultancy bureau. In addition MAFF was involved with the provision of a number of agricultural support services:

With this entirely top-down and ’green-revolution‘ type of design, the project ran into numerous problems during its implementation. Prior to the start, the hydrological studies for the borehole drilling were inadequately carried out. Many sites could not be used, and consequently, the area to be irrigated was much smaller than originally planned. Road construction was delayed, leading to excessive cost at the expense of marketing measures.

The involvement of farmers during the design and implementation phase was minimal. The area is populated by two main ethnic groups: native Fijians and Indians. Both groups have very different cultures and farming systems. While Indian farmers operate more as individuals with stronger market-orientation, Fijian farmers are more community- and subsistence-oriented. These fundamental differences were not considered. The extension strategy consisted of a single package that favoured a minority of resource rich farmers. These progressive farmers were used as demonstration model farmers. However, these model farmers failed to generate the anticipated ’spin-off‘ of technological benefits to neighbouring farmers. Extension was busy with delivering inputs, rather than providing sound advice and explaining to farmers what were the function of all components of the package and why all combined inputs were necessary. Hence farmers adopted bits and pieces of the package that they felt were of relevance and affordable. Consequently, yields under farming conditions were low and the ambitious project production targets could never be met (Taukei 1993, 122).

Farmers‘ experiences in producing vegetables in the area, in particular their knowledge of marketing problems were completely ignored. Project storage facilities contributed little to reduce seasonal market gluts. Fresh vegetable and fruit exports proved very difficult to


promote. Quarantine regulations of importing countries and basic produce quality standards for export could not be met.

Not achieving targets, and pressure from the ADB, increased respect towards the timetables of implementation. This further re-enforced top-down decision making and geared implementation towards quicker results. The issue of sustainability was forgotten. The large-scale design of the project with the high number of key actors (ADB as donor, MAFF as implementing agency, experts for management and implementation, separate Government Ministries involved for ground water studies and road constructions, and finally provincial administrations and beneficiaries) created major co-ordination and communication problems.

In the design stage, research was not consulted sufficiently. The technology was transferred from the ’international shelf‘ without being tested properly by researchers for its adaptability to local conditions. Therefore, many of the recommendations did not work in the new environment. Crop diseases caused considerable losses. The timeframe of the project did not allow for long-term research strategies to identify solutions to these field problems. With the strong focus on increasing production, research only followed strategies to increase yields of crops. Low input strategies and emphasis on quality rather than quantity were not considered. The configuration of this case study corresponds very well with Engel‘s description of a policy driven configuration (chapter 2.1).

The organisational culture was strongly biased in favour of the common civil service and Government administration protocols, and left little space for the rural communities to express their needs. Existing community structures were not integrated. Instead, new water- user groups were created. As this was done in a rather authoritarian way, and membership criteria favoured richer farmers, tensions grew within and between the communities, as not all villages benefited equally from the irrigation infrastructure measures.

The many problems encountered during implementation seriously undermined the financial viability of the project, and it had to be terminated in 1990 due to funding constraints. Since then, MAFF has been trying to make use of the existing infrastructure, and develop a new approach in order to reach the original goal of developing the area into the ’salad bowl of Fiji‘.

The SVRDP is a typical case for the work and approach of MAFF up to the early 1990s. An even larger rice development programme, financed by ADB and Japanese aid, followed the same blue print approach with an equally poor outcome.

These types of projects and the associated thinking and mentality are still common within MAFF. This could be felt in the discussions with field staff during the fieldwork for this study. In this context it is evident that considerable time is required until new approaches become institutionalised and readily applied.

The following case studies show different aspects of such new approaches and discuss the success of this ’fresh wind‘.


6.3.2 Vegetable seed and planting material (PRAP P3)

The project, vegetable seed and planting material, was the third project (P3) in a series of 11 that form the Pacific Regional Agricultural Programme (PRAP). The Program was initiated in 1989 and went into a second phase from 1994 to 1998. The first phase focused on agricultural research and technology development. Phase II projects were expected to finalise research activities, field test them and initiate extension activities in the countries.

Phase I P3 was entitled ’development of vegetables, grains and pulses‘. It was expected to recommend varieties, collaborate with and distribute screening tasks among national collaborators, and address regional problems of seed production and supply. Implementation showed that this formulation encompassed an almost unlimited number of crops to deal with<126>. and thus the project title was rephrased to ’seed and planting material‘ for the second phase. It was agreed that the project would work with 22 crops, grouped into 3 categories<127>. With varying emphasis on the 3 categories of crops, the project was expected to produce the following results:

The overall approach of the project could be characterised as supply-oriented. It aimed at improving the supply of seeds available at NARES <128> and in particular in the shelves of private seed vendors. This should give farmers a better choice (more varieties with different characteristics) and quality (e.g. high germination, timely availability etc.) of seed. Thus, the main target groups of the project were research and private seed companies and vendors. Besides the supply of seed and planting materials, the project assisted the classical research domains of on-station trial design and analysis.

Co-operation with extension remained limited. The main activities were demonstration plots of promising varieties. Furthermore the project co-operated with extension for the organisation of occasional joint field trips. As farmers were not considered the primary


target group, project findings were not analysed towards formulating appropriate extension messages. Results were documented in internal project papers or research papers, but these were not translated into extension materials for field use. However, the project invested some efforts into improving farmer - researcher - extension linkages. PRA activities helped to understand farmers‘ priority problems related to vegetable farming<129>. This granted very useful criteria for screening of new varieties. It showed in particular that farmers consider a large number of criteria, among others crop yield, often only of minor importance. This was an important learning process for researchers and motivated extension staff. These exercises helped to contribute towards a modification of existing working routines. Outside these PRAs, the project had no immediate direct impact on farm level. It remains to be seen if farmers will make use of a wider supply of vegetable seed in the future.

The project was very successful in linking MAFF with private partners and professional seed associations in Asia. It managed to get local vendors in Fiji in touch with a large number of suppliers overseas. Thus, one of the strengths of the configuration of actors in the project was the integration of the private sector.

One particular weak point emerged, namely a lack of seed legislation and strict quarantine restrictions. While the project managed to identify procedures for faster exchange of genetic materials, proposed seed policy changes got stuck in Government bureaucracy. This is a clear indication that the project lacked support on policy level.

Another weak aspect was the lack of a systematic monitoring system. While research variety trial routines were followed up quite regularly, no monitoring on a farmers‘ level (e.g. adoption rates) was initiated. Some monitoring of the seed supply situation on the market was undertaken, but no data on project effects and market developments were collected (cf. Lemonius 1997, 5). This makes an assessment of the overall impact of the project very difficult.

A retrospective view of the main project achievements and their impact one year after the official hand-over of the project is given by the project expert Dr. Steven Preston<130>:

Fiji MAFF is playing a more active/positive role in vegetable variety testing, and knows how useful this is to other countries in the region. In this, the Fiji MAFF is interacting well with the Asia and Pacific Seed Association (APSA), which helps arrange seed samples for trial. SPC/MAFF/APSA are developing this into a regional function. Fiji Quarantine has a better understanding of seed imports and I think this has opened up a lot of seed imports (especially pasture seed). The leading seed importer in Fiji has a much wider range of overseas seed sources in stock. Seed importers in other countries have better information on alternative seed sources, and some have introduced new varieties from new sources. (These have usually not been widely tested - the lack of systematic testing remains a problem). Some leading seed importers are active in a regional seed group, and the prospect of National Seed Councils, in which the seed importers will play their


part, is much stronger. There is some discussion about seed policies and a draft seed policy document is being reviewed. Slowly, the benefits will be seen by farmers and consumers. But the range of locally produced off-season vegetables available in Sigatoka market (for example) is virtually the same as it was before the project. However, a wider range of varieties, including new varieties are available (in Suva at least) and it is now only a matter of time before the farmers find better varieties amongst these.

Little more can be added to this statement. The main benefits can certainly be seen in drawing MAFF‘s attention to the private sector and building links with major players in overseas seed market. This also strengthens Fiji‘s role in becoming a hub for seed and planting materials to the surrounding small island nations. However, farmer and extension relations remain underdeveloped.

6.3.3 Agroforestry and farming systems (PRAP P1)

The agroforestry and farming systems project (P1) is also a component of the PRAP. The project was staffed with two experts, the team leader based in Western Samoa and an assistant based in Vanuatu. The team leader co-ordinated the activities in Fiji and visited the country several times a year. Thus, the project is an example of more limited external input in terms of advise, and consequently, more direct responsibility lying in the hands of the MAFF counterpart staff of the project.

In the first phase of PRAP (1990-94) the project started working with alley cropping systems, mainly with on-station research trials. During implementation more traditional agroforestry systems and grid planting patterns were studied. In the second phase of the programme (1994-98), emphasis was shifted towards better regional research networking, increasing farmer participation and extension of results. In the last phase (1999) capacity building of local collaborators with participatory methods and extension were further strengthened.

The project purpose, as defined in the latest phase, was to increase the adoption of sustainable smallholder crop production technologies<131>. This purpose should be attained through the following results:


When the project started it had a strong focus on research departments with classical station provenance trials and vigorous statistical analysis. Gradually, analysis of farming systems and the diverse nature of farming problems was given more attention. This led to the design of on-farm trials to evaluate potential solutions. Parallel to this, emphasis on conventional statistical analysis was reduced, as it often proved not possible or inappropriate in this context. New methodologies including PRA tools such as problem identification, causal analysis, prioritisation and ranking were introduced. This promoted farmers‘ participation encouraging them to test and evaluate possible solutions to their priority problems. Links with station research were maintained as necessary. Extension services were increasingly involved in the process and benefited from the PRA and on-farm technology testing. This inspired extension to use on-farm trials as a new method to promote what they considered as already proven technologies. The project also assisted extension in promoting on-farm demonstrations/field days and the preparation of extension materials. In Western Samoa the project started supporting and promoting farmer to farmer training and extension.

Farmer participation in the project increased over time considerably. Within the project mandate area<133> (e.g. agroforestry, soil fertility), farmers had possibilities to determine priorities, influence trial designs<134> and evaluate results. In this respect the project was among the first to promote a far reaching participation of farmers in the region.

The main project intervention was in the Waibau area, close to the capital Suva. The project used a set of criteria for the selection of farmers for co-operation. These were: semi-commercial farms, leaseholders, ALTA tenants that had soil fertility or perceived land sustainability problems. No studies were undertaken to estimate the number of farms with similar conditions on a national scale. Thus, remains unclear the number of farmers who could potentially benefit from the innovations developed.

Noteworthy is the technology development process. Trying to find solutions to farmers‘ problems, available international solutions, traditional knowledge and new ideas were tested. Among the innovations finally developed, farmers‘ traditional knowledge played a major role. One example (Erythrina trees) is a local legume tree that grows wild in Fiji and Samoa. In Tonga farmer yam production innovations were further developed.

Regarding actor configurations, the project assisted in building farmer, research and extension linkages. However, due to rather weak policy support within MAFF, these linkages were still largely limited to the staff directly involved in the project. A well visible institutionalisation and ’mainstreaming‘ of the methodologies developed has not yet been achieved. Outside the farmer, researcher and extension triangle, no other actors


were involved. In particular links to policy, market and private sector actors remained weak.

The main strengths of the project are the methodologies for participatory technology development and the identified innovations. The still rather limited field level impact of the project may be seen as a weakness. Only twenty farmers in Fiji have adopted innovations so far<135>. Another weakness, mainly of the MAFF system, is the lack of a monitoring system for adoption rates of farmers.

6.3.4 Pineapple export production

The project began in the late 1980s as a micro-project<136> funded by the Fiji Government and the EU. The development goal was to create income opportunities in remote rural areas of Vanua Levu<137>. A co-operative with smallholder farmers was created and a centre for collection, handling and marketing of pineapples was set up. Several start-up difficulties had to be overcome until the first exports were shipped in 1994. After this successful start was made and all capital works were completed, EU funding ran out in 1996.

The main project goal was to promote tropical fruit exports and start-up pineapple production and marketing. This should have been achieved through the following results:

The project analysed current experiences of farmers in growing pineapples and promoting a package of innovative measures: closer spacing with higher planting densities, use of mill mud as organic fertiliser, fertilisation according to plant leaf diagnosis, year round production by steering plant flowering periods with plant hormones. This package enabled high yields even on poor soils. The fertilisation according to plant leaf diagnosis allowed excellent fruit quality with a perfect balance of sugars and acids<139>.

The fruit packaging centre was built on the compound of the existing Seaqaqa research station. The small station also hosts extension staff, enabling good co-operation between research and extension. Research was involved in the project from the beginning. Its role was the testing and adoption of technology that was taken off the ’international shelf‘. At


first there were tensions between the project expert who promoted the new cropping techniques, and researchers that did not consider them feasible. However, both on-station and on-farm tests soon demonstrated their advantages and feasibility. Research revised the old recommendations and developed fast multiplication techniques to produce planting materials on a large scale for farmers. A nursery was set-up specifically for this purpose. The existing local pineapple variety proved as very disease resistant and thus the crop did not require any pesticide applications.

Extension was organised as a classical T&V service. Extension staff were trained by the project. Extension officers then passed on their knowledge to selected farmers in various ways: formal training at the centre, extension leaflets, group discussions, on-station and on-farm demonstrations and farmer to farmer visits. Besides training, extension was responsible for input delivery and pick up of fruits. Thus, extension could provide ongoing advice to farmers according to the needs of the cropping calendar and the harvesting cycle. But training was not limited to agricultural production, it also included all issues of transport, processing and handling. In particular, the co-operative staff and manager had to be trained in management and marketing skills. This included practical trips to New Zealand to familiarise themselves with the views and requirements of importers and customers at supermarkets.

Overall, this was a very comprehensive approach with a combination of different training methods. This was certainly one reason for its good success. An important difference to the ordinary extension service was that the staff were really well trained in all issues relating to the commodity<140>. Extension workers benefited from having a message to offer and discuss<141>. Another crucial issue was that extension workers were to some extent accountable to farmers, as they had the chance to raise problems during the co-operative meetings that were also attended by MAFF representatives. To further strengthen this aspect, it was planned that MAFF extension staff should slowly phase out in the future (once production is reached and consolidated on full capacity). Then, the co-operative should devise and pay it‘s own extension service.

Farmers rarely participated at the technology development stage, but their influence as members of the co-operative was strong. Farmers could voice their interest during co-operative meetings, and thus influence decisions on extension activities, management and production issues, and also in the crucial area of price policy. In 1995 about 60 farmers participated actively in the project and the requests pending for membership exceeded the target of 100 farmers by far. This was due to the visibility of the success<142> of the innovation and also due to the good prices paid to farmers. However the growth speed of the project with about 20 farmers per year remained moderate, as the multiplication of


planting materials created the main bottleneck that determined the expansion of the project.

This example showed that economic incentives could have a significant impact on adoption rates<143>. Participation may be an important ingredient, too, but it is, as this case shows not necessarily the single most important factor.

The project maintained extensive relations with several actors. Government and donor were updated on project progress regularly which secured policy support and reduced funding delays. Besides good research and extension linkages, farmers‘ participation was insured through a strong influence in the co-operative. Very important were the links to marketing actors. The project created good links to all actors in the marketing chain (e.g. local market, carriers, importers in New Zealand). In this respect, the project certainly represented the largest configuration of actors involved in all four case studies examined.

The main strengths of the project were that it created considerable impact on farm level using a very comprehensive training approach covering field production to processing and marketing. The project used a good monitoring system to steer implementation. Furthermore, the functional multi-actor involvement was certainly an important key to the success of the project. At the end of the field research period in 1995, the pineapple project appeared to be a very promising case. Meanwhile, the project faces some marketing problems overseas<144>.

The ’green revolution‘ type of high input technology may be seen as a weakness. This strategy comprehended more environmental risks. The marketing in this sector is riskier too, as the project competes with many conventional tropical fruit producers (e.g. big producers in Australia and Indonesia). Given the limited size of the overall intervention, an organic production strategy might be more sustainable in the long run.

6.3.5 Conclusions of these experiences

All four cases describe projects with considerable differences, not only in goals and approaches used, but also in terms of size, financial volume, staffing, and time frame of the interventions. Nevertheless, some general conclusions may be drawn from the examination of the case studies:

For the further assessment of the case studies the hypotheses formulated at the beginning shall be recalled:

H 7

Good co-operation of several actors may be achieved by small networks or platforms. Such platforms will enhance system output in terms of useful innovations.

H 8

Platform creation is enhanced through donor assistance or existence of good export opportunities.

Viewing the projects as platforms, the first hypothesis H 7 could be confirmed. All projects co-operated with different actors and this certainly could be seen as effective networking. Thus the projects created an interest for the different actors to work together. All projects produced innovations which were more or less successful in the diffusion of these innovations. Therefore, it may be said that the projects also enhanced the output of the AKIS.


The second hypothesis H 8 could also be confirmed. In all cases, donor funds were the starting point for platform creation. Thus, donor funding can have a strong impact on the output of an AKIS. As no official statistics on the output in terms of innovations of the Fiji AKIS exist, it is difficult to estimate the contribution of projects. According to the researcher‘s experience and observations, donor contribution could be estimated in the range of 50 to 75 % of all innovations developed and diffused. This illustrates the strong dependancy of the AKIS on external technical assistance still.

The question to which extent the sustainability of platforms depends on donor funding cycles, could not be answered definetly. At the end of the research phase in 1998, several projects were still being funded and therefore it is still too early to give an answer to this question. Nevertheless, the case of the pineapple project (funding ended 1996) showed that platforms may be sustainable after donors withdraw funds. The cooperative continued functioning and the other actors also co-operated in this platform. The agroforestry platform should also be sustainable as the Ministry supports the concept with a specific unit in the extension division. In this respect the sustainability of platforms seems to depend on the question to which extent the participating actors can find alternative funding sources. If that is not the case, the further promotion of the respective innovations is endangered. For donors, the key question would be to determine the right point in time when to end funding. In this respect the case studies do not yet provide indications.

A closer look shall be given to platform leadership in the configuration of actors of the four case studies. Figure 25 is an attempt to display the influence of key actors graphically. The influence of each actor was ranked according to the perception of the researcher<145> on a scale of 1 = weak, 3 = average to 5 = strong. As the assessment is based on the perception of the researcher only, this constitutes a strong element of subjectivity<146>.

The figure shows that all projects had established completely different configurations. In SVRDP farmers and research played a weak role while extension, policy and donor influence was strong. The vegetable seed project was divided in two groups with weak farmers, extension and policy actors on the one side and strong research and private sector on the other side. The agroforestry project had strong farmers, research and extension, while in the private sector, experts and policy actors were weaker. The pineapple project had overall a rather balanced profile. All actors ranged in between 3 and 4 as average to moderatly strong.

Considering the fact that the pineapple project was the overall most sucessful project, with a rather balanced influence of all actors, it could be concluded that this is a potential


lesson for future interventions: a balanced participation of actors in platforms ensures the best system performance.

In order to achieve such a ’balanced participation‘ good platform leadership is required. May be it is useful to distinguish between leadership and dominance in this context. Any set of actors working together will need some sort of leadership to achieve its goals.

Figure 25: Relative influence of selected actors in the case studies

Source: own data.

However, this leadership is only accepted by the other actors, as long as every actor in the configuration feels that his own interests are taken care of, and not dominated by other actors<147>. Perceived dominance could, consequently, be a reason for other actors to become more passive or even to withdraw from the configuration. This would then reduce the overall system output or performance. An important guideline for platform leadership could therefore be facilitation rather than strong leadership. Facilitation would further require that recognition be attributed to all platform members. It would also require taking over responsibility to ensure that all knowledge system functions are gone through by the platform.

Another aspect of ’balanced participation‘ would be the configuration as such. All relevant important actors should be represented. In other words, if important actors are


missing, the platform is not ’properly configurated‘. It could be seen as an important task of platform leadership to ensure that the platform is well configurated. This interpretation could also explain why some case studies were less successful. In the case of the SVRDP, farmers and researchers held too little influence, while policy and donors dominated. The vegetable seed project missed out extension while private sector issues dominated. The agroforestry project in the contrary, missed out the private sector.

Concluding, the graph could be used as a monitoring tool to analyse co-operation and leadership in platforms. Obviously, the number of case studies presented here is too small to fully validate the conclusions formulated. However, the findings may provide an interesting starting point for additional research.<148> The question of leadership in platforms will be investigated in more depth in chapter 6.4.2.

6.4 Outlook: platforms as means to promote innovations

The findings in this chapter are based on a separate section in the follow-up questionnaire that was distributed to research, extension and management staff at MAFF and the group of experts<149>. Prior to answering the questions, the respondents were asked to read a two-page outline of the platform model including graphical illustrations to familiarise themselves with the concept.

The findings in this chapter are, as the term ’outlook‘ in the heading already implies, an attempt to look into the future. Could platforms be a suitable means to improve the efficiency and performance of MAFF and the Fiji AKIS as a whole? It is important to note that this outlook or projection was based on what the respondents thought about the model, not on their practical experience working with the model. Nevertheless, some practical experiences were included, as several respondents have worked in projects that could, in retrospect, be perceived as such platforms. Such a retrospective assessment, which implies an ’unconscious‘ application of the model, is not the same as if a platform had been set up and managed with full intention. These limitations should be kept in mind during the assessment of these findings.

The presentation of the subject is segmented into six sections: platforms as a linkage mechanism, leadership and co-ordination in platforms, policy framework, driving forces for platforms, ways to introduce platforms and the overall suitability of the model for Fiji.

6.4.1 Platforms as linkage mechanism

’The platform model appears as a flexible solution that permits relevant people to work together, without the need for the creation of new formal units‘. In the first question, the


interviewees were asked to comment on this statement and express their opinion on the suitability of the model to solve the linkage problems within the Ministry.

The researchers all agreed on the suitability of the approach. Replies ranged from a simple<150>yes“ to “strongly feel that the approach will solve linkage problems." One researcher further specified “as a participatory approach, the platform approach suits well to the RRA/PRA methodologies introduced earlier.“

ED staff confirmed the suitability as well. Fourteen staff members welcomed the approach with a simple “yes.“ Others were more enthusiastic by saying: “really suitable“ and “better way“, “saves time and money“. One comment stresses the dissatisfaction with the current Ministry set-up and welcomed the platform model “...the platform approach is long overdue in our Ministry. It has been observed that most projects undertaken by the Ministry were undertaken without consultation and dialogue with the right personnel.“ Some concerns indicated further training would be required: “a thorough understanding is needed.

The experts rated the approach more reservedly. Four experts confirmed the potentials, while two experts considered the approach as non feasible. The advantage of linkages was stressed strongly in two comments: “this is exactly what is needed“, “fully agree, informal linkages are the key, whereby those key players in an area need to form linkages.“ Others pointed to the fact that the idea is not at all new: “platforms already exist in successful projects“, “several projects ... have adopted this approach and provide platforms to different degrees.“ A weakness that was raised was that linkage problems may be due to “human nature“ and therefore not all problems could be solved by the approach. Another critical aspect was seen in the fact that the approach “requires a champion - often a donor funded project to create the initiative, and facilitate the process.“ Fundamental scepticism was raised in one comment: “although maybe a very good idea, it is not implementable in Fiji due to cultural and social background, and lack of continuity of policy.

These first comments showed that the model was acknowledged with interest and the majority of the interviewees rated the approach as useful for improving linkages.

In the second question, the groups were asked to comment on the following statement: ’An advantage of platforms is, that depending on the actual work activities (e.g. technology generation or dissemination) relevant actors can increase or decrease their input relevant to the work needed for the platform. Not all actors need to be fully engaged at all times, which saves time and resources.‘

Again all researchers confirmed the statement as very relevant. Specific advantages were seen in “complementary roles“ and “absence of duplication“. “Being informed about things“ was seen as another positive aspect of platforms. “recognition“ was mentioned as an important steering instrument for platforms.

Extension staff equally strongly supported the statement which was best illustrated by some of the adjectives and words used: “fine“, and “excellent“. A few staff added some


potential limitations of the approach: “depends on the location“, “consultation necessary“ and “full participation required.“

The experts agreed to the increased flexibility in platforms and the potentially better resource use. But from experience, one expert raised the point that economies of time through platforms should not be overestimated: “much effort is required to keep all parties up to date with current states of affairs.“ Another problem associated with flexibility and voluntary access to platforms was raised as a question: “how to ensure people interact in areas of need and not interest.“

The comments indicated that platforms can create additional flexibility, but as the last comment stressed, the limits to a free choice of platforms in small organisations like MAFF are likely to be rather narrow.

In the first linkage workshop, a considerable part of the existing linkage problems were attributed to the stiff hierarchical structure of the Ministry. For this reason the following statement was presented for comments: ’another advantage of platforms is that the approach can create necessary flexibility in stiff hierarchies. At the same time the advantages of hierarchies (direct responsibility, administrative convenience, routine procedures) can be maintained‘.

Also in respect of this question, both researchers and extensionists strongly supported the statement. This was best illustrated in some comments: “excellent“, “breaks the barriers“, “faster“, easy for operation“, “room for interaction“. Only a few reservations were articulated: “some may use time to follow private purposes“, “superiors should regularly try to adhere to it“ and “the whole team (platform) should be made responsible for success or failure.

The experts agreed that more flexibility in the hierarchy was needed. One expert explained: “rigid hierarchies often do not work efficiently because of weak links that cannot be bypassed. In a hierarchical system, informal networking becomes difficult, yet it is the key to progress. A classic example is that the private sector usually forms no part of the public sector hierarchy (by definition), yet as in the case of seeds, these people are crucial to the whole seed supply system.“ Too much flexibility was seen as an obvious danger: “at the end of the day you still need good management and mechanism to plan, monitor and evaluate how resources are being allocated and used.

The comments showed that platforms were seen as a potentially appropriate means to complement the Ministry structure that was perceived more as a ’barrier‘ rather than an enabling structure. The main difficulty was seen in finding the right balance between increasing flexibility on the one hand and ensuring sound management of available resources on the other.

To assess the potential of platforms to create functional links to actors outside MAFF, the groups were asked to comment on the following statement: ’Platforms which aim at a particular commodity (e.g. ginger, fruits, coconuts, dairy) or more general aspects (e.g. farming systems, agroforestry) could bring together relevant staff from middle management, subject matter specialists, project experts, interested farmers and members of respective commodity boards or other market actors to achieve innovative progress‘.


Both researchers and extenionists agreed with the statement. The parallel between platforms and commodity boards was acknowledged directly by one expert “agreed, except a commodity board is probably already a platform. An effective commodity board certainly is!“ Commodity boards were seen as efficient structures, and the existence of commodity boards was rated as a good indicator for the chances to achieve innovative progress: “where (platforms) do not exist, the likelihood of success is very low.“ Two existing commodity boards in Fiji were considered as a suitable example. The functioning of these boards, however, also caused some problems: “the case in point there is now an industry group involved in ginger and another in taro but here appear to be inadequate resources from MAFF to address the major issues confronting these commodities.

The case showed that funding was considered as a crucial problem. This was also raised in another comment “platforms need money to run. Who is going to pay for subsistence platforms?“ Another comment took a similar direction and stressed the need to provide incentives for farmers and the private sector to support their participation in platforms “only very few are able to invest time and resources in such a network.“

Summing up, it may be said that the potentials of platforms presented in the three statements were largely acknowledged in the different comments. Platforms were seen as a potential means to promote innovative progress. Funding and co-ordination of resources were seen as crucial point. The latter point will be examined further in the next section.

6.4.2 Leadership and co-ordination in platforms

Platforms need leadership to pursue their goal of developing a certain innovation. All groups were asked to rate the leadership potential of different actors in Fiji as low, medium or high. The rating results with comments are shown in Table 37.

All three groups had different views on the question, ’which actor might be most suitable for platform leadership‘. Researchers saw the highest potential in the private sector (2.8), followed by top management (2.6) and themselves (2.5). Extension staff rated experts, top management and themselves all on the same level (2.6) as the actors with medium to high potential. The experts rated themselves as the only actor with high potential<151> (2.9) followed by the private sector with medium potential (2.3). The fact that all actors included themselves in the group of actors with higher potential, showed that all groups showed a good self-confidence.

Researchers gave their lowest rating (2.1) to experts, while extension attributed the same rating to farmers. The experts considered MAFF top management as the actor with the least potential (1.2). The following comments justified the ratings given. The RD and the ED ranked top management high as they have the “authority“ and “the final say“. Experts rated them low because they saw the “risk of overload“ and the fact that management have “other priorities.“


Table 37: Potential of different actors for platform leadership


Potential for leadership

Comments by the three groups










“Good knowledge“, “expertise, give direction“, “can have leading function“, “liaise, could facilitate -but outside“, “broader perspective“, “depends on individual“, “not sustainable“.

Private sector





“Can move anywhere, always in contact with farmers“, “can be biased“, “independent“, “best in touch, yes -see sugar industry“, “high potential -few time“, “too little developed“.






“Could co-ordinate and facilitate from their areas“, “restricted by superiors“, “together with privates“, “pretty organised“, “role could improve their status & motivation“.






“Could co-ordinate and facilitate from their areas“, “poor resource personal“, “restricted by superiors“, “in co-operation with privates“, “depends on individuals“.

MAFF top management





“Depends on policy makers“, “have the final say“, “can have leading function“, “far from reality“, “authority“, “horror“, “risk of overload“, “busy, other priorities“.






“Provide information“, “should lead“, “opinion leaders influential“, “poor accessibility“, “not empowered“, “support needed“, “small actor“, “no associations“, “lack of confidence“, “industry farmers to lead“.






“Could co-ordinate and facilitate from their areas“, “restricted by superiors“, “technically unsound, too economical“.

Scale: 1=low; 2= medium; 3= high.

Farmers were seen as an actor that “should lead“, but that are limited by “poor accessibility“ and lack of “empowerment“ that predetermines their role often as only to “provide information.“ Both research and extension had some potential to “co-ordinate and facilitate from their areas“, but they were “restricted“ by the existing hierarchy and “superiors“. In particular, for extension it was mentioned that the “role could improve their status and motivation.“ EP&S were rated to have some potential for leadership, but others considered them as “too economical“ and “technically unsound.“ Experts reached the highest overall recognition due to their “good knowledge“ and “expertise“. Their main weakness was seen in the fact that they are “outside“ MAFF and their contributions may “not be sustainable“ due to their short term assignments and limited presence to guide a platform over a longer period of time. The private sector was equally recognised as an actor with high potential due to its “independence“, “good contacts to farmers“ and proven success for example in the sugar industry. However, the actor‘s weakness was stated as “little time“ and at that time the linkage was viewed as “under developed.“

The question of platform leadership was difficult to answer as the comments illustrated. Based on the fieldwork, the researcher largely agrees with the expert rating. Capacities at the RD and the ED are probably still too weak to take up a responsible platform


leadership. However, the potential is there, and the progresses made in the training course on participatory methods demonstrated that staff could improve considerably if exposed to the right challenges. Private sector actors already play an important role for established commodities such as ginger or taro. Private sector actors would have an interest to gain something out of the co-operation with the Ministry. At present, the Ministry with its slow bureaucratic structure certainly does not yet appear a very attractive partner. The existence of a platform structure with flexible funds might change this impression quickly. For these reasons, a start-up promotion of platforms is likely to require expert assistance. The question will be taken up again in the last sub-chapter.

Besides leadership co-ordination is required in platforms. That co-ordination was perceived as a difficult task, and can be seen in the following comment: “a major problem with informal networking, is how to maintain co-ordination and avoid anarchy!“ Most of the other comments received equally pointed out the difficulties of achieving co-ordination. As reasons for co-ordination problems were mentioned: incompatibilities in goals between actors, a lack of a sense of joint ownership, little experience for co-ordination and conflict of ideas. However, others on the contrary, rated co-ordination as “easy to follow“ and pointed to the fact that Fijian culture has a good tradition of achieving consensus. Again, others made recommendations such as “positive thoughts and good co-operation need to prevail“ or “control required to control uncooperative guys“. Others added that training for teamwork was required. Strong leadership was mentioned as an important factor to achieve co-ordination.

Besides co-ordination between individual actors of each platform, overall co-ordination is equally important. In a new question, the interviewees were asked if an overall AKIS co-ordination platform was necessary, and if the directors of the RD, ED, AH&P, EP&S plus project experts or other actors would be suitable to take over this role.

The majority of research staff (6/8) approved the above suggestion as a suitable option for an AKIS co-ordination platform. The remaining two officers preferred middle management staff or a board of private sector actors and farmers. Among extension staff, the majority (15/21) opted for the above proposal as well.

The experts discussed the topic very controversial. Comments supporting and rejecting it, were made<152>. A co-ordination platform was considered as “somewhat artificial and not sustainable.“ More pragmatic, the need for a co-ordination platform was seen to “depend on how many competing platforms (projects!) were being facilitated simultaneously.“

If the Ministry started experimenting with only a few platforms, an additional co-ordination platform would certainly not be necessary. These few platforms could then be linked closely to the Ministry. One option would be to attach these platforms to the linkage unit that was proposed in chapter 6.1.2. This might facilitate the process of resource allocation. The suggestion will be further illustrated in the last sub-chapter.


6.4.3 Policy framework for platforms

The Fijian Government introduced a new policy, the commodity development framework strategy (CDF) in 1997. This policy framework determined commodities for future programme activities. In this respect the question arose, to which extent these new guidelines would be compatible with the platform model.

Among the researchers, all agreed that platforms could be combined with a CDF strategy. Extension staff was more sceptical. Only about half of them (12/21) indicated that platforms would be compatible with the CDF strategy now favoured by MAFF. The major point of concern raised was: “the CDF is a top down approach. The AKIS is a participatory approach. To combine both would be incompatible to promote sustainable development for the people at large. The CDF only takes care of small groups of commercial farmers.“

The experts were divided into groups that considered the new policies as a major obstacle to any kind of more participatory oriented methodologies and groups that did not expect major changes.

Considering the cases of platforms reviewed so far, these were either projects or commodity boards. Therefore, a policy focus on commodities is rather unlikely to affect the platform approach negatively. Nevertheless, it would be better if the commodities selected also corresponded to farmers‘ needs and potentials. A model of how farmers needs could be introduced into the AKIS was presented in chapter 5.5.4. Such a discussion process could ensure that both the views of farmers, Government and other key actors would be co-ordinated. Such a process would certainly ensure a higher motivation of all actors involved, and consequently also increase chances of success compared to a situation where decisions are imposed by the Fijian Government.

6.4.4 Driving force for platforms

A main assumption of the platform model is that the main driving force for different actors to work together in platforms is the formulation of a common goal and the joint understanding that this goal can only be achieved by close co-operation. The interviewees were invited to comment on this assumption and if necessary, to suggest additional incentives necessary to set-up and run platforms.

Among the researchers, three officers considered a joint goal as a sufficient driving force for platforms. Three other staff mentioned several incentives that were necessary to make platforms operational: autonomy, financial gain, recognition, satisfaction and other physical benefits. One officer suggested “an institution building process similar to the approach of PRA/RRA training“. In the same direction another comment pointed to additional training requirements: “the actors should have more qualifications“.

Among the extension staff, only six officers considered a joint goal as a sufficient driving force for platforms. Five were undecided while the remaining eleven staff mentioned a number of other incentives for platforms. Government support, quality reward system, resources and training were mentioned often. Other individual comments stated the


following incentives: “flexible support service“, “feedback and impartiality“ and “daily practice“.

The experts agreed that a joint goal on its own was not sufficient to set-up and maintain platforms. Similar to RD and ED staff, a number of different incentives were mentioned. Resources and funding continuity were seen as the crucial factor: “would require a sustained support system (donor?) until positive results become apparent to those involved.“ One expert stressed that besides external funding, MAFF should make better use of its own funds. As an example he mentioned the Ginger Council that was seeking outside assistance and funding for the research into pythiium rot: “...CDF has a lot of money to address these issues but the funds are not allocated to the area with the greatest need. Other important aspects raised, focused on the question of status and recognition: “job satisfaction, remuneration, medals. Being seen as a key player ... at least being able to say ’we did that‘.“

These comments provided a strong indication that a joint goal by itself is not a sufficient incentive to keep platforms operational. A joint goal will help to provide direction and facilitate consensus, but it cannot stand-alone. Other minimum factors that must be ensured are funding and recognition of the various actors for their contributions.

6.4.5 Ways to start-up platforms

Engel and Salomon (1997) developed a workshop approach designed to support the start-up of platforms. The approach is called rapid assessment of agricultural knowledge systems (RAAKS) and focuses on co-ordinating the potentials of different actors, analysing their goals and possibilities for co-operation, set platform goals, identify linkages, determine necessary information exchange, define working steps etc. The interviewees were asked to comment on the suitability of such a workshop promotion for platforms in Fiji.

Without exceptions, both researchers and extenionists welcomed such an approach to test the platform approach. This is best illustrated by some of the comments made: “seems ideal“, “yes, big potential“ , “try every possible avenue“, “yes, do it don‘t talk.

The experts considered the approach as a good potential starting point to introduce the approach in Fiji: “I would be interested to see the practical output from such a process. We should give it a go - all dependent on the quality of facilitation and the resources available.“ For an initial test of the approach it was suggested to apply the workshop to a narrower subject area rather than an entire department of agriculture.

This strong support for RAAKS showed a big interest in testing the platform model. An advantage of RAAKS would be that it could provide the necessary training that was raised by some interviewees, at the same time as setting up platforms. This is certainly one reason why this approach for the introduction of platforms was considered useful. Alternatively, the Ministry could experiment with platforms without external assistance, take up the concept in a running project or integrate it in the planning of a new


proposal<153>. The best experiences might be gained in a parallel set-up of platforms by the Ministry and promoted through donors via projects. A scientific follow-up of such a process could result in a more systematic analysis of future experiences. This would be an interesting field for future research that could further validate the insights gained with the present study.

6.4.6 Suitability of the platform model for Fiji

In the last point of the follow-up questionnaire, the interviewees were asked to specify strengths and weaknesses of the platform model and give an overall judgement of the suitability of the model for Fiji. Table 38 presents an extract of the main strengths and weaknesses mentioned.

Table 38: Strengths and weaknesses of the platform model



“helps achieve goals“

“brings the parties together“

“enhances team work“

“people learn where they stand“

“improves linkages“

“allows more flexible work practices“

“appears to reward competence and output, not length of service/seniority“

“could legitimise / regularise interactions with the private sector“

“it might improve status, morale and performance“

“better pooling of resources“

“better work commitment, better co-ordination“

“reduces duplication“

“works if there is a strong economical interest and an already expanding market“

“time consuming, too many actors“

“needs much facilitation also from top management“

“involves a weakening of hierarchical authority (this is a good thing in my view but will be resisted)“

“in some ways incompatible with current management systems: undermines the authority of the PS“

“egoism might be problematic“

“needs initiative, innovation, leadership and experienced-trained people. None of this is abundant.“

“flexibility creates a co-ordination problem that might be viewed as ’untidy‘ “

“it‘s difficult to see how it can usefully be formalised“

“less independence of actors“

“consensus difficult, needs good leadership“

“as a generalised, conceptual structure it has limited appeal“

The comments compiled in the table speak for themselves. Better co-ordination, improved linkages, better interplay of actors, means to reach goals, reduced duplication and better use of resources were identified as strengths of the platform model. Furthermore, it was assumed that platforms might have a positive influence on status, morale and performance by rewarding competence and output rather than length of service. Some of the strengths were also seen as potential weaknesses. Co-ordination and reaching of consensus might


be difficult to achieve or time consuming. Too much flexibility was seen as a danger for co-ordination. A risk was seen in the fact that platforms might weaken the hierarchical authority and therefore might be resisted. Other weaknesses were seen in the need for strong leadership and experienced-trained staff that are not available in abundance.

The overall judgement of the researchers expressed a strong support for the concept and considered it as suitable (7/8) for Fiji. Only one research officer was less certain and stated prudently “seems efficient.“

Extension staff showed an almost equally high support. Sixteen staff rated the approach as good or very good. Only 2 officers were “not sure“ or “reserved“. Four staff made no comments.

The support for the model was also shared by the experts. Various weak points or reservations were stressed, but overall, 4 experts favoured the model, one remained undecided while one firmly rejected the model. Two quotations may best illustrate this support with proviso: “It's an approach I would support, indeed we have adopted elements of this with a degree of success. As stated several networks already exist using a variety of participatory approaches - one should build on these and enhance positive experiences“, “to be cynical, Fiji has many of the elements of the classical framework for failure. Yet compared to many PICs<154>, Fiji has many if not all the ingredients for success, and should perform better. Therefore, playing with the institutional arrangement may be the key.

Overall, the comments of the interviewed researchers, extensionists and experts supported the platform model as a potential means to address current inefficiencies at the Ministry and make the development process of agricultural innovations more effective. The support and the reservations raised, indicated that the model has a good potential to prove valid in practice. The three case studies of newer projects presented in chapter 6.3 equally pointed in the same direction.

However, the empirical base is still too small to allow for a definite judgement of the model. Using the terminology of the participatory action research applied in this study, a new learning cycle would be necessary to test the platform model in practise. The experiences gained so far and the potential strengths and weaknesses identified could be used to formulate new working hypotheses to further guide the process of introducing platforms to the Fiji AKIS. A scenario for such an introduction process of platforms will be outlined in the following paragraphs.

In order to allow good judgement of the approach, the model should be tested in a number of case studies. Two to four case studies should be sufficient for this purpose. Such a number of case studies would permit comparison of different start-up settings and actor configurations. It would appear useful to compare a start-up with the help of workshops (e.g. RAAKS) and external assistance with a start-up without specific assistance. The latter case could include projects that are already implemented under full responsibility of the Ministry. This would provide insights into the suitability and necessity of start-up facilitation, and indicate how much training for platforms would be necessary.


The case studies could also provide further insights into the question of platform leadership. In this respect, the performance of different actors in leading roles could be compared. Based on the present research findings it would be interesting to compare the three actors that were attributed with the highest leadership potential: experts, private sector and MAFF staff. The concern of limited platform sustainability with experts as platform leaders could be further investigated by continued observation of the case studies presented in this study. Meanwhile, in all three cases, technical assistance ended<155>. How far this termination of donor support effected the sustainability of platforms, changes in actor configuration occurred and how the diffusion of the respective innovations was affected, could be studied.

The new case studies could further provide insights to which extent platforms could increase flexibility and improve co-ordination between actors. Regarding the question of overall platform co-ordination and the type of linkage between platforms and the Ministry, it appears suitable to attach these platforms to the linkage unit (that would also need to be created). This revised platform model is depicted in Figure 26.

This set-up would create a formal link between the platforms and the Ministry. Platforms would no longer be ’floating‘ completely outside MAFF. This might reduce reservations and potential resistance against the model. At the same time, this should facilitate the process of resource allocation to platforms and reduce the risk of co-ordination problems.

Compared to the first platform model presented in chapter 2.2.4, the co-ordination platform is replaced by the linkage unit in the new model. To indicate the connection of the unit with the Ministry, a double line is drawn to the policy framework of the model. Platforms could then be perceived as satellites of the Ministry. Besides promoting internal linkages these satellite platforms could improve contact and co-operation with all outside actors.

Overall platform co-ordination would thus become the task of middle management staff in the linkage unit. The advantage of this set-up would be that decisions would be closer to the executive staff of the Ministry. This might reduce some of the information flow problems, with information flows becoming shorter. More decision power would be delegated to middle management, while top management would be released from an excess workload. Only major<156> decisions would have to be taken by top management. This might have positive effects on the efficiency of the hierarchical structure of the Ministry.

As the follow-up questionnaire findings indicated, a joint goal is unlikely to be a strong enough incentive to run platforms. Therefore, in parallel to the above organisational modifications, the Ministry could start to develop a new reward system to promote and recognise the initiatives taken in platforms. Using the soft systems perspective, all actors would need to develop a more sympathetic understanding for each other. This would mean that actors could learn to actively take up and learn from other actors‘ point of view. In other words each actor should learn to look at matters from different perspectives (e.g. farmer,


extension, research, economist, sociologist, politician), to obtain the best possible understanding and consequently provide the best chances for platform success.

Figure 26: New platform model

Source: own design with ideas from Engel (1995, 203), Roux (1995, 19), Röling (1995, 12).

A scientific follow-up of this scenario could assist a more systematic analysis of the progress of implementation. It could provide interesting opportunities for further education for participating MAFF staff. Assistance by external researchers would certainly also be beneficial. After an implementation period of three to four years, the experiences should be evaluated. The evaluation should then enable a sound judgement as to, if and how the platform model should be introduced on a wider scale at MAFF.



In 1994 SPC-GTZ conducted two workshops. The first on rapid rural appraisal (RRA), knowledge attitude practise (KAP) and agricultural economics. The second on extension campaign planning, message design and materials development. Both workshops were well documented (cf. SPC-GTZ 1994 a and b).


* These 3 projects are described in detail as study cases later in chapter 6.3.




Several project required counterparts and office space that actually became a scarce factor.


The official Fiji Government bureaucracy turns at a very slow pace. Working in the EU Delegation, the researcher experienced that even simple project proposals took more than 2 years to pass all administrative hurdles.


In this context output could be defined as the number of innovations developed that reach a significant level of diffusion.


This workshop stimulated many discussions on what RRA and PRA are. This also made preparations in cycle two easier, as some basic principles of PRA were already known.


All team members had to express themselves in a foreign language. This made reporting more difficult. Kievelitz and Forster (1994, 62) talk in this respect of a "filter of foreign language".


Some reasons may be seen in the educational system. FCA offers a rather general agricultural course with few opportunities for specialisation. Enrolment in a masters course (e.g.) at USP is difficult for staff, as they have to bare the costs. Besides rare opportunities to participate in overseas training courses, or co-operation with projects, few opportunities for adult education and professional training courses exist.


For a long time the discussion centred on the question of what is a hypothesis. When the researcher provided an example the trainees understood the example but still found it difficult to abstract from the example and apply the concept to another area. Even several examples did not help to really resolve the problem of abstraction completely.


The tools are described in detail in the field report on the PRA (Bachmann et al. 1997 b).


Very few calculations were made by mental arithmetic. A pocket calculator was used for every simple calculation task. This slowed down work and made the task look more tedious.


Extension workers in many developed countries rely, to a large extent, on figures for both production and profitability issues. Such a framework of basic figures is a good tool to gauche farms in relation to other farms and to derive conclusions for the best possible advice to be given. Given the fact that Fiji needs to raise its international competitiveness, productivity issues will become more important and MAFF staff will need to work on this topic.


A crucial aspect here is that experience with figures and facilitation skills are both equally important. It is a task of the facilitator to question any figures that seem unrealistic and pass them back to the group for further discussion. With such a good moderation, farmers will be able to correct figures during the discussion process. During the field work, this proved particular true in the case of the advanced farmers groups, that showed a better understanding in financial matters and often discussed longer and corrected figures more often in the process.


On the previous surveys these details were clarified mainly by telephone. Difficulties to get the right people to agree on the activities, making the arrangements with farmers etc. proved time-consuming. Finally, one team member still had to visit the stations to get all details confirmed.


In addition marketing data were cross-checked with the interview of middlemen and merchants and dealt with separately.


This is a small group of 6 experts from different projects (P1, P3, P10, P11, SCEP) with whom the researcher exchanged experiences frequently.


The different cultures involved here are on the one side the ’Fijian‘ culture, that is already a mixed society with ethnic Fijians and Indians just to name to two major groups. On the other side the reseacher and experts come from the ’European‘ culture, but at the same time also belong to different countries with their own traditions (Germany, France, England, Australia, New Zealand). In this respect cultural biases should not be seen as a problem, but rather as an enriching element that promotes joint exchange and learning.


A literature search was done only once in the long preparation for PRA 2. This certainly explains the low rating.


Whose reality counts? The difficulties of judgement have to do with the cultural factor mentioned in the introduction. The question is: What rating scale is appropriate? Qualitative categories like 'good' or 'excellent' offer a wide range for interpretation. Furthermore the cultural context has an effect on the scale to be used. Viewed only in the context of Fiji, the researcher would agree that the team members certainly qualified as good to excellent compared with their colleagues at MAFF. However, seen on a ’European‘ scale, compared to working requirements at NAERS institutions in Europe, the rating of their abilities has to be lowered.


She worked with local PRA trainers that used a standard PRA package for a longer period of time. When the trainer group was faced with the task to modify and adapt some tools to the requirements for participatory monitoring and evaluation, the group encountered considerable difficulties.


Participatory methods were equally seen to play a vital role in this process of innovation development and diffusion (e.g. participatory technology development, participatory monitoring and evaluation, diffusion of innovations in a more participatory, advisory way rather than transfer of technology through extension).


The manual has been published in the meanwhile as a multimedia toolbox: PRAP P11 (1998): Participatory Learning and Action - PLA tool kit for the South Pacific.


The new formulations were used, because they covered the essence of PRA principles more precisely. The first set of rating criteria was wider; covering more general training aspects, while the new set used here focused more on core PRA principles.


More details are provided in the following chapter 6.2.5 on the institutionalisation of participatory methods.


Due to the very low number of respondents (2) these data have to be interpreted with caution. They may represent individual opinions rather than the view of management. Nevertheless, two out of seven management members provide a partial picture at least. Furthermore, the low participation in the questionnaire speaks for itself; it points towards the low priority attached to the question of participatory methods and organisational reform by management.


It could also indicate that they simply do not know anything about the consequences. This, however, would not be a good indicator for good management either, as managers should be well informed about all relevant activities going on in their organisation.


The new Commodity Development Framework (CDF), now requires the formulation of development plans in the format of the Logical Framework (In German known as ZOPP: Ziel orientierte Projekt Planung or English: Target Oriented Project Planning).


Two locations on the main Island Viti Levu in Fiji.


Decision-making is a difficult process in the Pacific. Decisions for PRAB were discussed at the annual meeting of the Regional Advisory Board (RAB) that was consisted of the Directors of Research of all countries. As 8 countries forwarded their needs and problems, the project was soon overloaded with tasks. In Pacific culture, where conflicts and offending other parties (counties) is avoided at all price, priority setting becomes an almost impossible task. Mandates grow rather than reduce and become focused.


The main problem was that all countries had different priority crops and no stakeholder was willing to drop one‘s country interest. Readers may still consider this number as much too high. However, at the time, this formulation represented a big step ahead. The number was reduced from “unlimited‘ as not clearly specified, to only 22 crops with further restrictions for the mandate for each crop category. This ’success‘ was possible with the help of PRA ranking tools applied during the log-frame workshop for the project.


This seed and planting materials were distributed for research test trial purposes. It should help the countries to identify the best varieties for their local conditions. The quantities were too small to provide materials for extension or for sale to farmers. It was left open to NARES to organise a local multiplication of the materials or purchase them from overseas.


Some of these activities were carried out jointly by the project and the researcher. Some results at the example of tomatoes were discussed earlier in chapter 5.5.3.


Email correspondence 6.2.1999.


Email correspondence from Dr. Steven Rogers 11.2.1999.


Email correspondence from Veronika Schwanz 4.2.1999.


The typical dilemma of projects: Priorities are set on policy level before the project starts. Hardly any donors approve open pilot phases, where beneficiaries can express their priorities without major restrictions.


Here the project faced another dilemma. As farmers have difficulties in expressing their understanding of matters, there was always a big risk that researchers or extensionists moved too fast and suggest ’their view‘ on designs, which were then adapted. This way researchers still dominated the process too much and farmers own ideas were not considered as much as might be desirable (email correspondence by Dr. Steven Rogers 11.2.1999).


Estimate by Taito Nakalevu, the responsible project collaborator in Fiji (email correspondence of 3.6.1999).


Micro-project is a special funding line of the EU. The main characteristics are limited funding volume, short period (2-3 years) and high own contributions by recipient country.


The second largest island of the country.


Interview with the project expert Mr. Van Santen 5.3.1993 (cf. Van Santen 1994).


Fruits sold on the local market soon built a very high reputation for the co-operative as the taste outstripped other local pineapples by far.


Interview with project expert Mr. Van Santen 23.3.1994.


Other positive aspects were that staff could focus on one single commodity only, the ratio of extension officers to farmers (4 to 100) was very good and the project secured regular transport (cf. Van Santen 1994, 8). Normal conditions in extension (in particular in remote locations like Vanua Levu) are usually much less favourable.


Some of the farms were located close to the main highway, so that the fields were well visible. Many of the first farmers started building new houses soon after joining the project. This obvious indicator for success attracted many of their colleagues to apply for membership in the co-operative.


This is also an explanation why the proceeding example with agroforestry showed comparatively much less impact. Here, benefits are much less visible and economic returns turn up only in medium term.


Communication with Mr. Bahn Singh, MAFF officer dated 10.3.2000.


The researcher was familiar with all projects (except SVRDP) through several field visits during implementation in the role of donor representative and later as program co-ordinator. As SVRDP was an older project, the assessment was based on discussions with former staff, visits to the project area, and available reports.


It would have been interesting to compare the researchers view with the views of other actors involved. However, as the idea for this comparison emerged during the analysis phase of the study, it was too late to seek the view of other actors. Despite this limitation, the findings are presented here, because the findings and the tool may be an interesting starting point for future research.


In contemporary management literature, a large number of models to determine the best type of leadership exist. In recent years agreement has been growing towards the theory that the effects of the type of leadership depend on the requirements of the context conditions. “The perception of one best type of leadership is replaced by a more relative : it all depends“ (Steinmann, Schreyögg 1993, 563). Among these context sensitive models in particular ’moderator models‘ gain importance (cf. Steinmann, Schreyögg 1993, 564).


It could be interesting to analyse how different actors rate each others influence in such configurations. Monitoring of changes over time could further reveal to which extent the tool could be useful in identifying and solving problems of system inefficiencies.


Management was also included as a separate group. However, as only one of the two management questionnaires returned contained answers on this subject, these responses were not included in the following analysis.


To preserve more of the richness of the views presented in the follow-up questionnaire, in a first step many of the comments were listed as quotes without immediate comment. Then an attempt was made to interpret the findings as a whole.


Interesting leadership reflections were raised in this comment: “ I‘m certain I could design some strong platforms with a large local component. I would almost invariably include a foreign element, not because it is foreign but because that is the source of specialist advice, which is often essential."


Rejections went so far as to question if the Ministry could at all play a constructive role: “Ideally, one would close the Ministry and rely much more on commodity boards and other ’platforms‘. Perhaps aid to the Ministries merely perpetuates an inefficient and unproductive system?“


The proposal for a follow-up project for PRAP P3 in the seed sector integrated the platform concept. (Email correspondence with Dr. Stephen Preston 6.2.1999).


Pacific Island Countries.


The last PRAP project ended late in 1999.


Division of labour and tasks would require a comprehensive review to be negotiated in the process of organisational reform.

© Die inhaltliche Zusammenstellung und Aufmachung dieser Publikation sowie die elektronische Verarbeitung sind urheberrechtlich geschützt. Jede Verwertung, die nicht ausdrücklich vom Urheberrechtsgesetz zugelassen ist, bedarf der vorherigen Zustimmung. Das gilt insbesondere für die Vervielfältigung, die Bearbeitung und Einspeicherung und Verarbeitung in elektronische Systeme.

DiML DTD Version 2.0
Zertifizierter Dokumentenserver
der Humboldt-Universität zu Berlin
HTML - Version erstellt am:
Thu Mar 29 16:23:51 2001