The Affordable Care Act, commonly called Obamacare, increased the number of Americans with an insurance coverage. However, there is a perceived growing disillusionment with this policy, which was meant to be Obama’s government flagship. But what caused so many Americans’ lukewarm reaction to a policy that made healthcare more accessible? An answer could be given by Abhijit Banerjee, Esther Duflo, Michael Kremer, the three economists who won the 2019 Economics Nobel prizes for tackling poverty. According to them the reason why so many good policies fail lies in their bad implementation. In fact, the online insurance exchanges where people could choose policies presented many technical problems and, consequently, many people had their existing insurance policies cancelled as these did not comply with the new regulations, despite the fact that the president had promised that those with existing healthcare plans could keep them.

As explained by the three development economists, this can happen because policy makers are too often concerned with “what to do” in order to tackle a problem rather than “how to do” it. Duflo, a French economist, at 46, is the youngest Economics Nobel laureate and the second woman to ever win this prize. She explained that when good policy doesn’t work, it is often because of bad plumbing, rather than bad economics. According to her, an economist should be like a plumber, it should not only design the policy machinery but also install it in the real world and pay attention to how it works in order to make the necessary corrections to its initial design if necessary. That is to say that while traditional ways have looked at the bigger picture, this new approach looks at the details.

The trio of economists were awarded the prize for “their experimental approach to alleviating global poverty”. But what does their approach consist of, and, why it is considered pioneering? They are commonly called by the industry experts “randomistas” since they use randomized controlled trials (RCT) to test the effectiveness of various policy interventions to alleviate poverty. A randomized controlled trial is an experiment that is designed to isolate the influence that a certain intervention or variable has on an outcome or event. It aims at reducing certain sources of bias when testing the effectiveness of a measure by randomly  allocating subjects to two or more groups, treating them differently, and then comparing them with respect to a measured response.

Abhijit Banerjee, Esther Duflo, Michael Kremer have borrowed this research tool from fields such as biomedical sciences where the effectiveness of various drugs was gauged using this technique. Mr. Kremer was the first to apply RCT to the field of development economics beginning in the 1990s in order to study the impact that free meals and books had on learning in Kenyan schools. Mr. Banerjee and Ms. Duflo, who were university peers at the MIT and are now partners in life too, later conducted similar experiments in India and further elaborated on RCTs and made them popular through their book Poor Economics, published in 2011.

These economists are concerned with questions such as: is it better to give to people anti-malaria nests for free or make them pay? What is the best way to make children go to school and be sure that they will profit from it? Is it better to encourage immunization by sending mobile clinics or award the parents with free rice bags? Huge top-down campaigns do not interest randomistas actions since their full attention is on the bottom-up approach. Randomization is the tool these economists use to reply to these questions.

But why is the randomization approach so important? We could think, for example, to evaluate the effects of a health insurance by simply comparing the health of insured and uninsured people, unfortunately, we would be wrong. Such simple comparison are often cited  as evidence of causal effects. More often than not, however, such comparisons are misleading. Comparisons of people with and without health insurance are not apples to apples; such contrasts are apples to oranges, or worse. Among other differences, those with health insurance are likely to be better educated, have higher income, and to be working than the uninsured (Joshua Angrist and Jörn-Steffen Pischke 2014). All these characteristics are probably correlated with health status, hence simply comparing the two groups will give misleading results. When evaluating the causal effect of a treatment/policy it is fundamental to compare “apple with apple”.

For these reasons, ideally, to study the effects of health insurance in a randomized trial, we’d start with a large enough sample of people who are currently uninsured. We would then randomly divide these people in two groups: one that would be provided with insurance and another one not. Later, the health of the insured and uninsured groups can be compared. The random assignment makes this comparison doable by narrowing down the differences to the variable we want to study: the impact of having a health insurance. Randomizing allow us to compare “apple with apple”, avoiding misleading results.

The randomistas assume that small changes guided by experiments, covering a wide range of topics, rather than broad positive thinking, will serve to steadily improve the plight of the poor. What they do is to unpack the problem of poverty and divide it in smaller tasks. As recognized by the Nobel jury, they are helping to address global poverty by asking themselves “smaller, more manageable questions,” rather than being guided by big ideas. Therefore, the focus is on the micro rather than on the macro-level. This is also because the assumption is that part of the problem is the behavior of the poor and this why there is so much overlapping with behavioral economics.

Nonetheless, this school of thought, despite being nowadays the North Star of development economics, is far from being void of critics. In fact, most of the critics believe that the idea of helping poor people to overcome low incomes and precarious lives by correcting individual misbehaviors and cognitive biases is too simplistic and paternalistic. They believe that by focusing merely on the micro-level, randomistas choose to miss the bigger picture and forget that poverty is often a product of larger structural injustices of the economic and political system since rich people make also mistakes but the consequences of their mistakes are far smaller. Therefore, in their opinion, by refusing to meddle in structural change, the randomistas approach is partial. While many applaud the shift of focus away from politics and policy, many still think it is important to keep them in the solving equation. In the case of a cut to school systems, for example, they would direct their attention to absenteeism of teachers, the effects of school meals and the number of teachers in the classroom on learning and not to austerity. Abhijit Banerjee, Esther Duflo, Michael Kremer and their followers believe that policy should be de-politicized and that RCT can help policy be evidence based. However, other economists believe that policy should still be also guided by debate over values rather than just results.

While the scope of this article is not that of choosing a side, it is undeniable that this Nobel prize plays a big role in reviving the debate over poverty reduction and by awarding a woman and a man of color it marks a step closer to a more inclusive development economy field. What is sure is that global poverty and development are keeping some great minds up at night.