Go to content

8 Conclusions and Recommendations

8.1 Conclusions

The relevance of applying a FFS approach has remained high within the Bangladeshi context over the period of evaluation in support to the country’s efforts to become a Lower Middle-Income country and graduate from a Least Developed Country (LDC) to a developing country. However, more support seems needed at the policy level.

The FFS approach fits well within the Vision 2021 and the 7th FYP (which is aligned to the SDGs), where focus is on increasing the pace of poverty reduction and reducing income inequality. Nevertheless, recent trends in Bangladesh show that the pace of poverty reduction is decreasing, and income inequality is increasing. Therefore, more support and efforts may be needed to bring the positive experiences from use of the FFS approach into national policy development and implementation processes. Although the evaluation found less uptake of technologies in IFMC compared to AFSP and earlier phases, the relevance of these technologies was still found to be high. The reduced uptake is mostly a result of the change in training approach and quality, compared to previous phases (see Conclusion 6).

The FFS approach, as implemented in both IFMC and AFSP, has delivered several positive short-term results along the lines of the hoped-for changes in AGEP.

Since 2013, almost one million poor people (representing close to half a million households) in rural Bangladesh have benefitted directly from new knowledge and techniques related to agricultural production and nutrition introduced through FFS in AGEP. The impact from FFS on household income (around BDT 10,000 annually, equivalent to an average monthly household income), food security, diversification of agricultural production, women’s empowerment and nutrition is highly significant. FFS households have reduced their vulnerability and increased their intake of most food items significantly more than control village households. Although it has not been possible to specifically assess job creation, it should be noted that other studies have found that a 10% rise in farm incomes generates a 6% rise in non-farm incomes,[110] thus indicating that increase in income resulting from learning through FFS is linked to broader effects.

The scaling-up of FFS in IFMC, and especially the combination of many training modules into one package, has led to reduced quality of the FFS training.

The approach of exploratory learning has been diluted in most FFS training in IFMC and is the main reason for the decline in quality. The training has still been of sufficient quality to lead to positive results in production and income, albeit not as good as in earlier phases.[111] Many of the FFs conducting the training had limited experience as facilitators and, due to the many different modules in the FFS, the FFs also had to deal with topics outside their own experience and “comfort zone”.[112] On top of that, the FFs have been supervised and guided by DAE staff, who are actually “crop” people, with limited knowledge on fish, livestock, poultry, nutrition and women empowerment topics. All this has contributed to lowering the quality of the FFS. In AFSP the results related to increased production and access to extension appeared more impressive and could be augmented further by increased use of systematized farmer-to-farmer extension. This could also serve to reduce training costs.

Favouritism and clientelism in the process of selection of FFS villages and participants have to some extent led to exclusion of the power-poor from participating.

While the upscaling of FFS has made it possible to support a large group of farmers who would not have been reached through DAE’s mainstreaming approach, the support has, at the same time, been less pro-poor oriented compared to previous programme phases. Although the group of landless, marginal and small farmers has been well represented in the FFS groups, in many cases it has not been the poorest segments of these groups that have been reached, but the relatively wealthier. As an explanatory factor, the evaluation found indications of favouritism and clientelism in the process of selection of FFS villages and participants in IFMC. This is supported by recent studies, showing that in rural areas in Bangladesh local elites are now increasingly diversifying their power base beyond landownership and money lending into multiple and often flexible party-political affiliations. In this way, political attachments have become an important determinant in maintaining power relationships, often resulting in exclusion of the power-poor from wider benefits.

AGEP has contributed to a significant enhancement of women’s empowerment within FFS households, however intra-household issues need to be addressed more explicitly to challenge existing barriers for women’s mobility and decision-making power in relation to farm management.

The income of women from FFS households increased significantly compared to women in control groups, mainly due to enhanced production of poultry and vegetables from homestead gardening. Income from eggs and vegetables are largely controlled by the women. Assets such as chickens and cattle are still sold by men and, therefore, women have less control over profit. This also applies for Bengali women in CHT, but ethnic minority women have acquired more control over income from their own activities. Women who have participated in FFS have become significantly more involved in the process around decision-making but the final decision, still lies with the men. In addition, the workload on women has increased with the number of activities, although lack of mobility still limits women’s full participation and benefit from FFS and larger income sources continue to be controlled by men. These issues form part of rural livelihood and farming systems, where farming is seen as a family business and gender inequalities addressed in a co-operative manner with women and men. Some improvement in mobility of ethnic minority women in CHT has however occurred. Women in all the regions are gaining bargaining power in communities and in their households. The involvement in FFS and FOs have contributed to this development.

The decision to implement the FO model and market-oriented activities through a government institution (DAE) has lowered the quality.

The implementation of the model developed in IFMC for FOs and their linkages to BFPs has not been working well and has been decided upon with insufficient testing and learning. However, women have gained access to markets through FOs with the establishment of collection points. Although IFMC has contributed to establishing FOs, the evaluation found that these were often not operating as intended, mainly due to political interference, elite capture and power relationships. In addition, the programme target of forming more than one thousand FOs has been too ambitious. The complex context, large geographical coverage, insufficient financial and human resources, implementing market-oriented activities through a government institution, etc., has lowered the quality. However, some good examples of better-functioning FOs and their market linkages were identified and are of use as a source for inspiration. Well-functioning FOs have proven instrumental for women’s empowerment, as they provide a space for women to sell their produce without involving husbands and sons.

FFS as implemented in IFMC has shown signs of institutional weaknesses and management challenges at various levels affecting the efficiency of the interventions.

The evaluation found clear indications of institutional weaknesses at various intervention levels and, moreover, that the dual and decentralised management structure applied for programme implementation has not worked as intended. The established M&E system has only been partly functional in support of managerial and operational activities while the set-up for backstopping and quality assurance included inherent risks for inefficiencies. However, the fact that it has not been possible to fully mitigate/avoid the risks does not imply “failure”, but rather that achieving the full potential benefits of the FFS has been hampered, for instance when the ability to detect, act and follow up on weaknesses or problems has been limited. Some risks have been the price paid for working in partnership with a large governmental institution to institutionalise processes and approaches, rather than having independent implementation units with a short-term implementation focus. The AFSP institutional set-up seems to have included less inherent risks and better captures the possibilities for check and balances than IFMC, as it has operated at a smaller scale as a project modality.

Despite high cost-effectiveness from the supported interventions, the evaluation of the FFS approach raises critical questions about the future, not least in relation to IPM, finance and institutional partnerships.

Despite a number of overall positive results from the FFS approach, including a high cost-effectiveness with a pay-back time of less than a year, it is worrying that important areas such as IPM, in relation to high value crops, has not been sufficiently covered by the supported interventions. In IFMC, the absence of a dedicated curriculum and an adequate exploratory training approach is an important missed opportunity and puts at risk a longer-term sustainability. Institutional arrangements and finance for future FFS support are still unclear and, while it is unlikely that DAE will have the capacity and capability to continue with the FFS approach on its own, alternative institutional and financial modalities for FFS have not been tested as part of IFMC. AFSP has experiences with another division of labour between line departments and implementing partners (not just UNDP but also NGOs) and both the costs and quality of the support has been higher than in IFMC. Although the evaluation found indication of some spill-over effects from FFS farmers to non-FFS farmers within FFS villages, there seems to be potential for larger impact here.

8.2 Recommendations


Future development interventions in Bangladesh, aiming at reducing vulnerability and improving food security, nutrition and livelihoods among poor rural households, should continue making use of the FFS approach, incorporating the recommendations given below to address current weaknesses and opportunities.

This should also include concerns about bearing the costs of adaptation to climate change in Bangladesh, which have major implications for the most vulnerable. In view of a recent slowdown in the pace in poverty reduction and an increased inequality in Bangladesh, a properly designed FFS approach with an explicit pro-poor focus could contribute to a reversing of these trends, as it will be able to foster a rise in income and improved food security and nutrition amongst poor farmers.


The season-long exploratory learning should be brought back as the heart of the FFS approach in Bangladesh.

Future FFS’ should become more flexible and focused (fewer modules and participants in one FFS and use of explorative learning). It is crucial that the exploratory learning aspect is brought back as the heart of the FFS approach in Bangladesh. This is needed to ensure that not only technologies are introduced but also that FFS will stimulate, encourage and empower farmers to develop problem-solving skills and have the confidence to apply them on other innovative and developing practices. This will make them more resilient for dealing with challenges caused by e.g. climate change. It should be ensured that the experience-based learning will also be used in the establishment of FOs. In each FFS the participants should have more influence on selection of modules, based on their needs and priorities. Depending on the location and the interest of the participants there could be a fish FFS in one area, a poultry/ homestead vegetables FFS in another area, and a potato or rice FFS somewhere else. Each of these should also pay attention to gender and FO formation, but focus should be on season-long experiential learning of the main topics. A more flexible household approach, where a couple (husband and wife) could split up and attend different FFS sessions, could be explored.

In addition, facilitators should have practical experience in the topic they teach (e.g. an experienced fish farmer should be used as FF for fish, etc.). Nevertheless, the FFs will need to be retrained to make the main modules participatory and experiential. The different FFS modules will have different timing (depending on the topics) and different budgets. This will be a complex task to manage but will most probably end up with higher quality FFSs. Furthermore, in view of the general trend towards the production of more high-value crops, it appears important to develop FFS curricula for these crops and to ensure that IPM is adequately addressed. Finally, the FFS modules should also be adjusted with a view to promoting larger spill-over effects, including encouraging FFS farmers to share information with others.


Current guidelines and procedures for selection of FFS/FO participants and group composition should be reviewed and more clearly defined, emphasising inclusion and focusing on the poorest and most vulnerable farmers.

This will imply some clear choices in a future Danish country programme, including: i) more narrow definition and targets for inclusion of the poorest and most vulnerable farmers (e.g. more clear definition of vulnerability, more strict requirement to land access and ownership, specific targets for participation of female-headed and other vulnerable household groups); ii) a stronger supervision of the selection process at a time when changes in rural power structures in Bangladesh are observed and are making it more difficult for the power-poor to be included in development projects such as FFS.


Future FFS interventions should include a broader definition of women’s empowerment as well as inclusion of more specific goals and targets.

While AGEP basically has focused on women’s participation and income, other relevant parameters to include in a women’s empowerment definition would be time consumption; decision-making; mobility and control of assets; etc. These and other intra-household issues form part of rural livelihood and farming systems and would need to be addressed more explicitly to challenge existing barriers for women’s mobility and decision-making power in relation to farm management. One of the already proven methodologies in this area is the Gender Action Learning System (GALS), where farming is seen as a family business and gender inequalities addressed in a cooperative manner with women and men. Close cooperation should be ensured with other programmes working in this field in Bangladesh (such as IFPRI). Finally, the current approach for identification and employment of FFs should be revisited to ensure a more equal gender balance with a particular view to strengthening women’s participation in FF.


The approach to establishing and training of FOs should be reconsidered.

This should include an assessment of alternative partnerships to DAE, which appears not to be the right partner for this activity. The approach to training should be based on the principles and praxis of exploratory learning and could well focus on farmers’ clubs (and similar groups) already developed in previous Danida supported programmes. Topics could include identification of markets and their demands; how to operate on a market; how to assess prices and other conditions (down payment, delayed payment, timing of payments, interests on outstanding payments, quality issues); and how to agree on a deal. Exploratory group learning is highly suitable for such subjects by applying direct investigations (e.g. of markets), followed by group assessments and agreed, common actions. Any new model should be well piloted and tested before scaling up.


A more effective management information and monitoring system should be established for subsequent FFS programmes/phases

It is recommended for subsequent FFS programmes/phases that this will include: i) a Baseline Study that should be designed, conducted and preserved to make it as useful as possible for ex-post evaluation (due to the complexity of this process, an expert with required skills and expertise should be consulted); ii) a performance monitoring system for FOs should be established based on a few, easily collected indicators; and iii) the monitoring system should include specific targets and indicators for measurement of women empowerment and qualitative participation (see Recommendation 5), spill-over effects from FFS as well as the direct and indirect employment effects from FFS and FOs.


Continued support should be provided to the Bangladesh Agricultural Extension Network as a platform for national dialogue on farmer-centred approaches and multi-actor consultation.

This could ultimately lead to involvement of a broader group of key stakeholders in planning and implementation of FFS. By implication, there is a need to continue the process of establishing and strengthening the national dialogue, and to ensure that Danida continues to engage in advocacy for the various important aspects of the FFS approach. This includes support to policy development, such as development of a Gender Policy in DAE.


Alternative “FFS models” should be piloted to make the support as self-financing and best practice oriented as possible (such as establishing of FFS networks and commercialization of services and income-generating activities).

Strengthening of peer training networks should also be considered a key element for development of a more sustainability approach, including with a view to promote larger spill-over effects and encourage FFS farmers to share information with others. Here it will be important to build further on the combined experiences from IFMC and AFSP (working with national partners, NGOs, project unit vs. GoB management, the role of GoB line departments; use of different approaches for selection and use of trainers/facilitators etc.).

[110] WB 2016; Gautam et al. “Dynamics of rural growth in Bangladesh : sustaining poverty reduction”; https://www.worldbank.org/en/news/feature/2016/05/17/bangladeshs-agriculture-a-poverty-reducer-in-need-of-modernization.

[111] See the 2011 Evaluation Report.

[112] E.g. a FF giving training on fish culture would not necessarily have any practical experience from cultivation of fish, or a female FF giving training on rice, would never have grown rice herself.


This page forms part of the publication "Evaluation of Agricultural Growth & Employment Programme (AGEP), Bangladesh – October 2019" as chapter 8 of 8.
Version no. 1.0, 2020-01-16
Publication may be found at the address http://www.netpublikationer.dk/um/evaluation_agep_bangladesh_oct19/index.html