2019 Nobel Prize

This year’s Nobel Prize in economics, announced on Monday, was a ray of sunshine amidst the prevailing media gloom.

The award was made for the work of the new Laureates on the alleviation of global poverty. This is one reason to be cheerful about it.

In addition, they have made important developments in how economists go about solving problems. They are a key part of the drive to move economics on from an obsession with pure theory and to make it much more empirically based.

At first sight, the award is very conventional. All of this week’s joint recipients are based in top American universities, two at MIT and one at Harvard. Of the previous 81 winners, no fewer than 61 did most of their work at Ivy League institutions.

They also retain the basic view of economists. Poor people are essentially rational agents, trying to take decisions which are in their own interest. They may have much more difficulty in accessing relevant information than others. They face many more constraints on their ability to make the best decision. But they are just as rational as everyone else.

The similarity with tradition ends there. Esther Duflo, for example, became only the second woman to win the prize, along with her close collaborators Abhijit Banerjee and Michael Kremer, so something different there.

Their main innovation is to introduce the use of randomised controlled trials (RCTs) into economics. One hundred years ago, the British statistician Ronald Fisher was revolutionising the principles of statistical analysis. It is the maths he developed which enabled the testing of new medicines to become much more scientific.

The basic idea is to have a group of people who take the new drug and a group who do not. The key thing is to assign them into the groups purely at random. This way, any difference in the outcomes of the group which was treated and the group which was not can reasonably be thought to be due to the impact of the drug.

Duflo and her colleagues, along with many others they have inspired, have successfully addressed a wide range of real-life policy problems in the developing world using the same approach.

Examples include discovering how best to get farmers to use more effective fertilisers, how to increase the uptake of safe water filters, how to improve patient safety in hospitals, how to spread advice most effectively, about tuberculosis using community-based counsellors and how to improve safety in public service vehicles.

The technique of RCT has even been applied in developed world settings. For example, experiments have been carried out with job applications, sending them out with names which strongly imply the ethnic background of the applicant and seeing if the response differs across groups. (It does).

The approach of RCT is not without critics in economics, even now. An important issue, for example, is that experiments are typically on a small scale, and they may be issues when they are scaled up.

But Duflo and colleagues, unlike some economics Laureates, have definitely helped to make the world a better place.

Paul Ormerod