Applying mixed methods to evaluate pro-poor interventions for a pork value chain in southern Myanmar

Abstract

Applying mixed methods to evaluate pro-poor interventions for a pork value chain in southern Myanmar Introduction Pro-poor value chain approaches are common within development initiatives to improve smallholder incomes, but few deal adequately with the complexities of these chains, limiting their ability to quantify, ex-ante, the impacts of chain interventions. The research described in this review applied participatory methods to engage stakeholders in the process of building a system dynamics (SD) model that accommodated complexity and allowed trade-off analysis across chain actors when simulating the effects of chain interventions on incomes and livelihoods. This review shows how these qualitative and quantitative research methods combined to generate information that informed decision making at relatively low cost. Data, methods and results The research used participatory spatial group model building (SGMB) and reference group (RG) sessions in which qualitative and quantitative data were collected to co-construct SD models with chain stakeholders. We used five sequential SGMB workshops, attended by 13 value chain actors (>50% female) including pig producers, brokers, butchers, and wholesalers. The RG sessions involved subject matter experts who helped refine SGMB outputs and the SD model by triangulating data and analysing preliminary results. An average of eight people attended RG sessions, with 25% female representation. The SD model developed for the pork value chain comprised seven modules, each representing a sub-sector of the pork value chain. The model extended recent agri-food models by creating additional modules covering the provision of formal credit in the system, collective marketing by producer groups (PGs) and collective value-adding by producer organisations (POs). A production module covered biological processes of pig production and alternative farming and marketing systems within 32 target villages. Simulations showed that a well-sequenced combination of technical training, collective action, disease prevention, and micro and meso-credit would deliver the best outcomes for small-scale pig producers and intermediaries. While some farmers can successfully upgrade their enterprises with the assistance of just one or two interventions, poorer farmers require more layered support. Greater inclusion and better financial outcomes can be achieved by directing project investments to broadly support a smaller number of committed smallholders within functional producer groups rather than spreading project investments across large target groups. Conclusion Testing interventions in a participatory SD modelling environment is a cost-effective way of evaluating pro-poor interventions in agricultural development projects. This approach is particularly applicable to complex agri-food value chains where interventions often create winners at the expense of losers, and result in unintended negative consequences that manifest long after implementation. It is also a useful way of generating collaboration among chain stakeholders.