Towards a generic, comprehensive and participatory approach for assessing the impact of agricultural research in developing countries

International public agricultural research for development is increasingly requested to contribute to solving societal challenges related to food security, ecological transitions, climatic change and inequalities in development, among others. At the same time, in one strand of the scientific communities modalities and criteria to assess research are shifting towards demonstrating and explaining the causal link between research outputs and development impacts (Gaunand et al;, 2015). This paper describes a novel approach for impact assessment of agricultural research conducted in a developing country context adapted from the “impact pathways” and ASIRPA approaches (e.g. Douthwaite and Gummert 2010; Joly et al., 2015). A key methodological choice was to give an active role in the assessment to the multiple stakeholders involved in innovation and/or impacted by it. This was considered essential to identify impacts and indicators that evaluation teams might not have thought of by themselves, and to understand the complexity of innovation processes eventually leading to impact, particularly in a developing country context for which accountability towards end-users of research is weak and availability of or access to reliable quantitative data is a challenge. The resulting participatory methodology, called “Impress” (IMPact of RESearch in the South) focuses on establishing and explaining the relationships between the outputs produced by research, the outcomes that involve and affect the actors directly or indirectly interacting with research and ultimately the primary and secondary impacts for development. After developing it iteratively over several years, CIRAD is currently testing it by assessing 13 case studies throughout Africa, Asia and Latin America. Cases cover a wide diversity of innovation domains (plant breeding, post-harvest processing, pest and disease management, value chains, etc.), research approaches (from transfer of technology to action-research) and innovation trajectories, some of which were assessed ex-post and others in itinere). The perimeter of each case study includes suites of consecutive or closely related research or R&D projects which took place around a common theme in a given area over time. Evidence and indicators of impact are being identified, measured and validated through multi-stakeholders workshops, focus groups, semi-structured interviews and surveys with key concerned stakeholder groups. Initial results show a wide diversity of mostly positive impact types, including increases in production and incomes, improvement of the natural resources base, increased access to remunerative markets and at times changes in regulatory frameworks and policies, among others. Some impacts and indicators thereof were identified by the case stakeholders themselves, vindicating the choice of a participatory approach. Impacts seem linked among others to the development of and access to new knowledge and the strengthening of partnerships among stakeholders. Developing the capacities of stakeholders and of smallholder farmers in particular, also seems to have had a significant effect, which may have contributed to scaling out of some innovations. After cross-analysing this round of case studies and addressing the methodological lessons and challenges, CIRAD will release by end of 2016 an improved, user-friendlier version of IMPRESS which may be used within and outside CIRAD. It also hopes to encourage its researchers and Triomphe et al. Towards a generic, comprehensive and participatory approach... 3Southern partners to adopt an “impact culture” so that they can adjust their research planning and practices in an ex-ante fashion, and in doing so, improve the probability their future research work will indeed contribute to more and better impact. (Résumé d'auteur