Work on randomised control trials (RCTs) in economics and more generally
Since the award of the 2019 Nobel Memorial Prize in Economics there has been a resurgence of interest in the use of randomised trials in development economics – including problems and limitations of this approach. This page serves to collate my work on this subject, which formally started in 2010 when I began my PhD in economics on the external validity of results from randomised trials.
I have co-authored three relatively accessible pieces that were written after the announcement of the Nobel: “The Poverty of Poor Economics” in Africa is a Country with Grieve Chelwa; and two pieces in The Conversation with Grieve Chelwa and Nimi Hoffmann, one of which provides some background to the use of RCTs in development economics and the other outlines some key criticisms.
In August 2020 I published a mid-length article entitled “Randomised Trials as a Dead-end for African Development” in a CODESRIA Special Bulletin: Randomised Control Trials and Development Research in Africa.
Criticisms focused more sharply on methodological limitations can be found in a short, relatively accessible contribution to a World Development special issue on the 2019 Nobel published online in early 2020.
I have recently completed a book chapter that contributes to the philosophy and methodology of economics literature and should be published in 2020 or 2021: “Randomised trials in economics”, in H. Kincaid and D. Ross eds. The Modern Guide to Philosophy of Economics, Edward Elgar [pre-print available]
All of this work follows from my PhD dissertation on the problem of external validity (generalisation) for randomised experiments in economics (and more broadly).
The first chapter was published as a SALDRU working paper (which has been widely-read but less-widely cited…) and an updated version presented at the World Bank’s Annual Bank Conference in Development Economics (ABCDE). A compact version of the core econometric critique was subsequently published in the World Bank Economic Review.
In my work I try as much as possible to connect different academic disciplines, which also means paying attention to the corresponding literatures. The working papers above provide quite detailed references to the pre-2016 literature and I subsequently added more here. (The overlapping areas are large and growing rapidly: my current desktop bibliography folder contains 535 references and is still not comprehensive…).
In addition to that work mostly addressing economics, I conducted parallel work in philosophy. Unfortunately it turns out that publishing and citation dynamics in that discipline are not much better than in economics, but my 2012/13 philosophy of economics paper presented at the Evidence and Causality in the Sciences conference series is available on request and I have various related papers under submission – one of which has been accepted for presentation at the Philosophy of Science Association biannual conference (originally 2020 now rescheduled for 2021). The book chapter referred to above provides a fairly comprehensive explanation and overview of the philosophical and methodological issues with RCTs and my views thereon.