On Oct. 14, Abhijit Banerjee and Esther Duflo (Mr. Banerjee and Ms. Duflo are married: She was his student), both of M.I.T., and Michael Kremer of Harvard, won the Nobel Memorial Prize in Economics for “field studies” in poor countries.
That is, the Nobelists — Esther Duflo is the youngest ever and the second woman in the 50-year history of the prize — performed numerous experiments in which they give one group a treatment, such as eyeglasses to Chinese children learning to read, and use a matched group as a control. The control group doesn’t get the glasses. Question: How much better do the glasses-children do in learning to read?
The idea is to imitate in economics what is called in medicine the “gold standard” of randomized trials. The makers of field experiment claim they mine scientific gold. If the glasses-children do a lot better in reading, then the economists can recommend the policy of giving eyeglasses to children with poor vision, like recommending a baby aspirin a day if you’ve already had a heart attack. Doctor’s orders. Give the kids eyeglasses, do not give people mosquito nets, and on and on to thousands of medical-economic prescriptions. World poverty solved. And above all we’ve been scientific about it.
What could go wrong?
Plenty. The eyeglass example is not hypothetical. Three followers of the Nobelists, the economists Glewwe, Park and Zhao, reported in 2012 on its success for 19,000 school children with bad sight. With financing from the World Bank, thousands were given glasses at $2 a pair, and did learn Chinese characters. Thousands were not given glasses, and didn’t.
Any problems here? In 2016, two other economists, Stephen T. Ziliak and Edward R. Teather-Posadas, studied the experiment. They uncovered devastating errors in the use of statistics by the field experimenters and their supporters. Many other economists, such as the Nobelists Angus Deaton and James Heckman, also have grave worries of a technical sort about field experiments. A big one is inferring from a study in a locale to a recommendation for a national policy. Educational experiments have been misleading this way, as the philosopher of science Nancy Cartwright has argued.
But Messrs. Ziliak and Teather-Posadas make also the, shall we say, blindingly obvious point that we know already that a child who cannot see will not be able to read the alphabet, not to speak of hundreds of Chinese characters. The experiment is, to use a technical word, stupid. It’s like the joke about the experiment to test whether parachutes work. Throw 10 people with parachutes out of an airplane at 5,000 feet. Then throw 10 without. Science satisfied.
Messrs. Ziliak and Teather-Posadas remark that the eyeglasses experiment, and a good deal of the work of the Nobelists, is as startlingly unethical, and stupid, as the notorious Tuskegee syphilis experiment run from 1931 to as late as 1971. African-American men were randomly assigned to not get the penicillin that the medical scientists from the U.S. Department of Health already knew cured the disease.
The economists, like the medical researchers, seem to have lost touch with their proper role. They are not ethically assigned to master our lives. The mastering assignment is what they assume when they focus on “policy,” understood as tricking or bribing or coercing people to do what’s best. It sounds fine, until you realize that it is what your mother did to you when you were 2 years old, and had properly stopped doing to you by the time you were 21. The field experimenters scorn adult liberty. And that is the other way many economists have lost touch. As noted by economist William Easterly, another critic of the experimental work, and as argued at length by your reporter in numerous books, the real way to solve world poverty is liberty. Not dubious, fiddly, bossy little policies handed down from the elite.
How do Mr. Easterly and I know? We look around as economic scientists, and see that when India in 1991 gave up Gandhian illiberalism, and when China in 1978 gave up Maoist illiberalism, and when Japan in 1868 gave up Tokugawan illiberalism, and when in 1776 the Blessed Adam Smith’s “The Wealth of Nations” crushed the case for illiberalism, the poor were enriched. By 3,000 percent.
And the Nobel committee seems to have lost touch with actual science, which does not make a fetish out of method, which closely watches its ethics, and which is guided by actual, non-stupid curiosities. When the great physicist, a Nobel winner, Richard Feynman suspected that the Challenger disaster had happened because the seals hardened and broke on an unusually cold day, he tested it by dunking a piece of the rubber seal into a glass of ice water, and breaking it in front of the other commissioners. When the historian Lord Acton found that power tends to corrupt, and absolute power corrupts absolutely, he got it from a lifetime of studying how all great men are bad. That’s science. Even economic science.
• Deirdre Nansen McCloskey is an economist and historian. Her latest book is “Why Liberalism Works: How True Liberal Values Produce a Freer, More Equal, Prosperous World for All” (Yale University Press, 2019).