Development impacts of value chain interventions: how to collect credible evidence and draw valid conclusions in impact evaluations?

Abstract

In development policy and practice, support to or interventions in value chains are considered to be instrumental for achieving outcomes such as poverty alleviation. This paper reviews methodological discussions on how to show the effects and workings of value chain support in a context of donors demanding rigorous impact evaluations. The paper starts with a discussion of evaluation methods strongly anchored in ex-post statistical analysis of effect measurements, and argues in favour of a theory-based evaluation protocol, equipped to handle threats to valid conclusions. Value chains are open, multi-layered systems and development outcomes are multi-dimensional and contingent on contextual particularities. Moreover, development interventions in value chains are often time, place and commodity specific and unlikely to be repeated in a similar way, which complicates generalisation and constrains evaluative conclusions. The example of a small-grant fund promoting collective marketing by smallholder organisations illustrates these methodology challenges and shows the value of using a mix of methods for addressing the problems of outcome measuring, impact attribution and generalisations from highly diverse contexts.