Impact evaluations in international development co- operation

Abstract

Experimental designs are currently in the methodological frontier of impact assessment and development economics. The methodology has received some criticisms: some of them are applied to any type of evaluation while others are specific to their randomized design. The paper points out and discusses four of them: the need for theory to design creatively each evaluation; the problem of heterogeneity; external validity; and the translation of evidence into public policy. In light of the Spanish experience with the Spanish Impact Evaluation Fund, the main recommendation that emerges from the analysis is that policy makers should be aware of the potential and limitations of the wide range of impact assessments, to support their implementation when appropriate and to adjust the inquiry to the necessary accuracy, time, budget constraints and to the utility of the information they receive in order to take a political decision.