On October 14 the Swedish Academy of Sciences awarded the “Nobel prize” in economics to Abhijit Banerjee, Esther Duflo and Michael Kremer for “their experimental approach to alleviating global poverty.”
The prize in economics is not one of the original prizes endowed by Swedish armaments manufacturer Alfred Nobel; it was established in 1969 as a tag along and is endowed by the Bank of Sweden.
Banerjee and Duflo teach at the Massachusetts Institute of Technology while Kremer is at Harvard University. The trio have been at the forefront of pushing the use of randomised control trials in the subdiscipline known as development economics.
Partly as the result of their efforts, an ecosystem has developed in which the vampire squids with tentacles of influence across the globe are the Abdul Latif Jameel Poverty Action Lab (J-Pal), the International Initiative for Impact Evaluation (3ie), and the World Bank’s development impact evaluation group.
The main idea behind their work is that randomised control trials allow us to know what works and doesn’t work in development because of their “experimental” approach.
Such trials are most well-known for their use in medicine and involve the random assignment of interventions into “treatment” and “control” groups. Just like in medicine, so the argument goes, randomised control trials allow us to know which development pill to swallow because of the rigour associated with the experimental approach.
Banerjee and Duflo popularised their work in a 2011 book, Poor Economics: A Radical Rethinking of the Way to Fight Global Poverty.
Even though other Nobel prize awards often attract public controversy (peace and literature come to mind), the economics prize has largely flown under the radar, with announcements often met with the same shrugging of the shoulders as, for example, the chemistry prize. This year has, however, been different (so was 1976, when the winner was Milton Friedman, that high priest of neoliberalism).
A broad section of commentary, particularly from the Global South, has puzzled over the committee’s decision not only to reward an approach that many consider as suffering from serious ethical and methodological problems, but also extol its virtues and supposed benefits for poor people.
Many of the trio’s randomised control trials have been performed on black and brown people in poor parts of the world. Serious ethical and moral questions have been raised, particularly about the types of experiments that the randomistas, as they are colloquially known, have been allowed to perform.
In one study in western Kenya, which is one-half of the epicentre of this kind of experimentation, randomistas deliberately gave some villages more money and others less to check if people from the villages receiving less would become envious of those receiving more.
The study’s authors, without any sense of shame, titled their paper “Is Your Gain My Pain?” In another study in India, the other half of the epicentre, researchers installed intrusive cameras in classrooms to police teacher attendance (this study was actually favourably mentioned by the Swedish Academy).
There are some superficial rationalisations for this sort of thing, but studies of this kind — and there are many — would never have seen the light of day had the experimental subjects been rich Westerners.
There are also concerns around the extractive nature of the randomised control trials enterprise. To execute these interventions, randomistas rely on massive teams of local assistants (academics, students, community workers, and the like) who often make nontrivial contributions to the projects.
Similarly, those to be studied (the poor villagers) lend their incalculable emotional labour to these projects. It is often unclear whether they have been adequately consulted or if the randomistas have simply struck deals with local officials. The villagers are the ones that have to deal with all the community-level disruptions that the randomistas introduce and then leave behind once they’ve gone back to their cushy lives in the United States and Europe.
And although there is an increasing amount of posturing to compensate for this exploitation, with some researchers gushing about how they and their “native assistants” are bosom buddies, the payoffs of the projects (lucrative career advancement, fame, speaking gigs, and so on) accrue only to the randomistas. The most extreme case is obviously this week’s award.
Beyond the ethics of the Nobel winners, their disciples, and the institutions they have created in their image are two serious methodological problems that fundamentally undermine their findings.
The first is that the vast majority of studies conducted using these methods (our rough guess is more than 90%) have no formal basis for generalisation. In other words, there is no basis to believe that the findings of these studies can be applied beyond the narrow confines of the population on which the experiments are conducted. This is simply fatal for policy purposes.
The prize-giving committee addresses this only in passing by saying that “the laureates have also been at the forefront of research on the issue of [whether experimental results apply in other contexts].”
This is misleading at best and false at worst. There are some advocates of randomised trials who have conducted important research on the problem, but the majority of key contributions are not by advocates of randomised trials and the three awardees have been marginal contributors.
The more important point is simply that if the problem of whether experimental results are relevant outside the experiment has not been resolved, how can it be claimed that the trio’s work is “reducing world poverty”?
The second contradiction is more widely understood: despite the gushing headlines in the Western press, there is simply no evidence that policy based on randomised trials is better than alternatives. Countries that are now developed did not need foreign researchers running experiments on local poor people to grow their economies. There is ample historical evidence that growth, development and dramatic reductions in poverty can be achieved without randomised trials.
Randomistas claim that their methods are the holy grail of development yet they have not presented any serious arguments to show why theirs is the appropriate response. Instead, they take the case that such methods are crucial for policy largely for granted because they think they are doing “science.” But although they are certainly imitating what researchers in various scientific disciplines do, the claim that the results are as reliable and useful for economic and social questions is unsupported.
It is instead a matter of blind faith — as with the conviction many such individuals appear to have of a calling to save the poor, usually black and brown, masses of the world.
We do not have a view on whether these individuals ought to have been awarded the prize — prizes are usually somewhat dubious in their arbitrariness and historical contingency. But the claims made about the usefulness and credibility of the methods employed are concerning, both because they are unfounded and because they inform a missionary complex that we believe is more of a threat to the progress of developing countries than it is an aid.
This is an edited version of an article that originally appeared on Africa is a Country