Abductive Reasoning to Explain Integrated Development: Lessons from the Multi-Method Evaluation of the Millennium Villages Project

The use of mixed methods in impact evaluation often focuses on the triangulation of findings or using qualitative data to explain statistical assessments of impact (net effect). In this article, we argue for the benefits of combining and contesting ex ante etic-informed theory with ex post emic-informed grounded theory. Using the case of the Millennium Villages Project (MVP) evaluation in northern Ghana, we highlight examples of where emergent theory led to a deeper understanding that valued how local people view change. In the context of an independent evaluation that set out to assess the MVP model – and with little scope for the evaluation to influence or adapt implementation – the theories of change created abductively helped fill a theoretical void. We argue that this is particularly applicable for integrated programmes where ex ante theory may struggle to capture the complex interaction of different sector-based activities, overlook unintended effects, and undervalue local perceptions of change.