Since the late 1980s, support to agriculture has moved from top-down agricultural extension towards more participatory approaches which better suit smallholders. Farmer field schools are one such approach used to teach specialist knowledge, promote skills and empower farmers, usually with the objectives of reducing pesticide use and increasing agricultural yields. They are typically implemented by facilitators who use participatory, discovery-based learning methods; for example, farmer field schools use practice plots so that farmers can compare the results of different farming methods. Farmer field schools have been widely used across Asia, Africa and Latin America, but the effectiveness of this approach is still debated.
The review aimed to investigate the effects of farmer field schools on outcomes for participants and on outcomes for their non-participating neighbours. The authors also aimed to identify barriers and enablers of the effectiveness and sustainability of farmer field schools, and to the diffusion of methods to non-participants.
The authors included studies of farmer field school interventions, defined as interventions that: (i) provide intensive, facilitated group training (usually weekly meetings for a season), (ii) use both practice plots (demonstrating new farming methods) and control plots (farmed using standard methods), and (iii) provide information on holistic techniques that reduce input use (such as integrated pest management). For the quantitative component, the authors included experimental and quasi-experimental study designs (with data at either the farm or household level) assessing the effect of farmer field school interventions on intermediate outcomes (such as improved knowledge or adoption of new practices) or final outcomes (such increased yields or profits) for crop farmers in low- or middle-income countries. The authors also included analysis of outcomes for crop farmers who lived in the same community as farmer field school participants and who may have been exposed to the approaches (whether via exposure or via formal dissemination methods). For the qualitative component, the authors included studies that reported on farmer field schools; were relevant to addressing barriers to, and enablers of, effective farmer field schools; were based on primary data that was analysed using qualitative methods or descriptive statistics; and provided information on the research question, data collection procedures, sampling and recruitment.
The authors included journal articles and grey literature, with searches updated in October 2012. The search strategy covered electronic databases (including ProQuest, Web of Knowledge and AgEcon), hand-searching of relevant journals (including economic, development and agricultural journals) and searching of relevant websites (including aid agencies, international development organisations, UN agencies and research organisations). The authors extracted data and appraised studies systematically. They used inverse-variance weighted statistical meta-analysis to address the question on effects and thematic synthesis to address the question on barriers and facilitators.
Headline Findings: a summary statement
In small-scale or pilot programmes, Farmer field schools (FFS) have improved farmers' knowledge and adoption of new practices, leading to increased agricultural production and increased income in some contexts. They appear less effective when delivered at scale, and there was no evidence that neighbouring farmers benefit from diffusion of knowledge from FFS participants.
The authors included 92 impact evaluation studies and 20 qualitative studies, although only 15 of the impact evaluations were considered of sufficient quality to make recommendations for policy. The quantitative studies were conducted in Sub-Saharan Africa (31 studies), South Asia (25 studies), East Asia (24 studies), Latin America (11 studies) and Central Asia (one study). The qualitative studies were conducted in Africa (11 studies), Asia (7 studies) and Latin America (two studies). Most of the studies evaluated farmer field schools that promoted integrated pest management; some studies evaluated other approaches to reduce intensive inputs, such as integrated crop management and integrated soil management.
Implications for policy and practice
The authors found that in small-scale or pilot programmes, farmer field schools lead to improved knowledge and changed agricultural practices. These changes have significant benefits for farmers, with an estimated 13 per cent increase in yields (RR=1.13, 95% CI=1.04, 1.22), 20 per cent increase in profits (RR=1.19, 95% CI=1.11, 1.27) and reduced environmental degradation due to reduced pesticide use (RR=0.61, 95% CI=0.48, 0.78). The increase in profits was higher for FFS projects which also included complementary interventions involving input or marketing support (RR=2.51, 95% CI=1.51, 4.16).'
Non-participating neighbouring farmers do not benefit from diffusion of knowledge from farmers who participate in FFS; therefore, they may not be cost-effective compared to more traditional extension programmes. The few studies that evaluated FFS programmes delivered at larger scales, such as national-level programmes, showed that such programmes do not appear to be effective in the longer term.'
The authors conclude that farmer field schools could be used selectively to solve particular problems in particular contexts, but are not useful to solve large-scale problems.'
The authors drew on qualitative studies to explain these findings:
- Facilitation: Problems in recruiting and training appropriate FFS facilitators may have been a barrier to scaling up programmes. Farmer field schools require skilled facilitators and attempts to target facilitators based on education or literacy levels may be less effective than targeting based on ability to communicate, and appropriate training which enables facilitators to use a bottom-up approach. This is most obviously a barrier in scaled-up programmes where FFS facilitators are recruited from extension staff who previously used more top-down agricultural extension methods.'
- Diffusion: At a local scale, farmer field schools rely on hands-on experience and demonstration of the benefits of new methods, so diffusion of methods to non-participant neighbours is unlikely. The qualitative studies also showed that strong social networks and concrete, simple methods could enable diffusion, while low levels of social cohesion and socioeconomic differences between participants and non-participants were barriers to diffusion.'
- Targeting: The majority of FFS projects targeted better-off farmers, which appears to have been successful. Half of the projects used pro-poor targeting, which did not always succeed in reaching the target groups because targeting mechanisms favoured elites or the characteristics of more disadvantaged target groups made it difficult for them to participate. Programmes appear to have had mixed success in reaching women.
Implications for further research
The authors found a lack of high quality quantitative studies, pointing to a need for more high-quality, theory-based impact evaluations that report and analyse a hypothesised causal chain. In particular, none of the studies used cluster-randomised assignment, which is a feasible approach for this type of intervention. To evaluate the sustainability of outcomes there is a need for studies with more than two years of follow-up data.
Studies are also needed to measure a broader range of policy-relevant outcomes; for example, farmer field school programmes often claim to have positive effects on outcomes such as health and empowerment, but few studies have collected data on these outcomes. Finally, few qualitative studies reported on the views and experiences of farmer field school facilitators.
This is a high quality systematic review. The authors conducted a comprehensive search, screened and extracted data from relevant studies using standardised methods and used appropriate criteria for the critical appraisal of the included studies. They used a program theory of change framework to provide context to the two review questions and provided a combined quantitative and qualitative analysis to enhance the narrative causal chain synthesis. They included a portfolio analysis to provide context to the findings of the review. A minor limitation is the lack of double coding of high-risk studies, but these studies were not considered in the assessment of policy implications of the studies, and the lack of explicit criteria stating non-English studies were included.