Experimental approach to development economics: a review of issues and options

Abstract

ABSTRACT Randomized control trials (RCTs) are recognized as the preferred tool of analysis in modern development economics literature/research and policy evaluation. This may lead to methodologies, including case studies, tabular analysis, simple regressions, taking a back seat. This survey explores the implications of such a methodological hierarchy and the implications of preoccupation with a particular evidence/methodology for research and policy. Similar developments in macroeconomic modelling are also discussed. Major advantages and limitations of RCTs and the attempts to address them are highlighted. The article argues that preoccupation with a methodology can sometimes lead to important inquiries for research and policy getting side-lined on methodological considerations. This leads to inferences favouring a particular technique/methodology or issue. Focusing solely on methodologies that emphasize quantifying the ‘effect’ may not be appropriate to address all questions relevant to development. As policies involve multiple and conflicting social concerns, methodological pluralism may be preferable.