Evidence-based development needs a diversity of tools, with a bottom-up process of “embedded” dialogue
Along with the rest of Economics, Development Economics has been changed for the better by the ‘‘credibility revolution”
Abstract
Along with the rest of Economics, Development Economics has been changed for the better by the ‘‘credibility revolution”. Twenty years ago, the standard method of establishing causality was with the use of a (usually dubious) instrumental variable. We can no longer get away with that. The seminal work of economists such as Angrist, Card and Krueger raised the standards for what constituted a causal claim with the use of quasi-experimental methods including natural experiments and regression discontinuity designs. Banerji, Duflo and Kremer took this a step further by conducting a series of experiments with random assignment to examine the impact of development interventions. They coupled this with something akin to leading a social movement – a relentless process of education, persuasion and mobilization to make policy interventions more evidence-based. This combination of scientific innovation and advocacy was wildly successful and led to a fundamental transformation in how development is assessed, and in many ways, how it is done. Following Duflo (2017), some development economists now see themselves as ‘‘plumbers” with a good practical understanding of how to design interventions and solve problems. There is no question that we have become better plumbers over the last twenty years, but are we good enough plumbers? Do we developmentistas (of which randomistas are a subset), know as much about the functioning of development dynamics as much as (good) plumbers know about the hydraulics of buildings? The paper discusses Evidence-based development approach.
This is an output of the World Bank’s Strategic Research Program
Citation
Rao, Vijayendra “Evidence-based development needs a diversity of tools, with a bottom-up process of “embedded” dialogue,” World Development, Volume 127, March 2020. https://doi.org/10.1016/j.worlddev.2019.104823.