Use of Randomization in the Evaluation of Development Effectiveness.pdf

Duflo and Kremer 1 1 Use of Randomization in the Evaluation of Development Effectiveness Esther Duflo2 Michael Kremer3 Paper prepared for the World Bank Operations Evaluation Department (OED) Conference on Evaluation and Development Effectiveness in Washington, D.C. 15-16 July, 2003 Abstract In this paper, we argue that current evaluation practices often encounter problems that prevent them from effectively determining program impact, and that there is considerable scope for greater use of randomized evaluation methods in addressing these problems. We review recent randomized evaluations of educational programs in developing countries, including programs to increase school participation, provide educational inputs, and reform education, and then extract lessons from these examples. Finally, we discuss the role the international agencies can play in promoting and financing rigorous impact evaluations, including randomized evaluations. 1 We thank Ted Miguel for extremely detailed and useful comments. We are particularly grateful to Heidi Williams for outstanding research assistance. We are also very grateful to Abhijit Banerjee, Francois Bourguignon, Anne Case, Angus Deaton, Rachel Glennerster, Emily Oster, and Paul Schultz. We gratefully acknowledge financial support from the Alfred P. Sloan Foundation. 2 Department of Economics, MIT, BREAD, and NBER, eduflo@mit.edu 3 Department of Economics, Harvard University, BREAD, The Brookings Institution, Center for Global Development, and NBER, mkremer@fas.harvard.edu