Does Research Reduce Poverty? Assessing the welfare impact of policy-orientated research in agriculture

ALINe publication: Sumner, A., Masset, E. and Mulmi, R. (2010) IDS Practice Paper, under review. Available online


In the current context of the global financial crisis and its aftermath, development resources are likely to be getting scarcer. Resources for development research are too. The set of circumstances generating the resource scarcity is also putting pressure on development gains. More than ever before, every dollar spent on development will have to count towards sustainable poverty reduction as will every dollar spent on development research. However, understanding the impacts of development research on policy change and on poverty is weak at best, with agriculture being no different.

The area of research impact is not a new area of enquiry but an emergent one. Our paper seeks to build on the work of others, notably, IFPRI, CGIAR, IDRC, ODI RAPID, GDN, NR International and ECDPM. In our paper we survey the literature and identify different ways of assessing the impact of ‘policy-oriented’ research. We then take the available literature on agriculture as a specific focus to survey.

Our paper surveys the different types of ‘policy-oriented’ research; the literature on the ‘theories of change’ for policy research in international development; methodologies for analysing the impact of policy-oriented research; the relevant agriculture literature and outlines the types indicators that can be used for impact assessment of research with examples.

The key findings are:

  • There is no standard practice for the evaluation of research projects and every evaluation strategy should be designed on a case-by-case basis.
  • Provided we are willing to accept some assumptions, it is possible to test research project impacts along some dimensions of social welfare (agricultural output, income or poverty) by finding the appropriate indicators (and methodology). The overall goal – welfare impacts of research – is highly desirable but not always feasible (especially so due to time-lags).
  • When a welfare assessment of research projects is not feasible, it is recommended that evaluators test intermediate project outcomes. The articulation of the theory of change of the project allows testing critical links in the causal chain running from research to welfare.

%d bloggers like this: