DFID commissioned an evaluation of EngKaR to support decision-making about the future of research on pro-poor infrastructure within the Research Funding Framework. The evaluation, conducted by Technopolis and ODI, tackles the way the programme has been designed and run, its results and - as relevant - opportunities for improvement.

This study included interviews with DFID staff, other policy makers and relevant NGOs; questionnaires to project leaders and 'expert' reviews of selected projects; a review of programme logic and management processes; and country visits to assess impact on policy.

We conclude that EngKaR is an inherently sound and valuable programme whose 'public good' character offers high 'leverage' in poverty reduction. Its performance would be improved by:

• Clustering projects thematically and geographically to increase the programme's 'clout' in country, and to make capacity building more cumulative
• Anchoring project design in the needs of specific beneficiaries at an earlier stage in the project life cycle
• More effectively coupling the projects to strong stakeholder interests in beneficiary countries or among donors who can ensure implementation and exploitation of results. In particular, they should be better linked to DFID's own plans and activities at country level