This article, written with Professor Dean Karlan of Yale University, appeared in Stanford Social Innovation Review.
Public debate about two prominent poverty-alleviation programs shows that over the past 15 years international development has become much more scientific.
The international development world is currently hosting rows about whether two poverty-alleviation programs actually work.
The Millennium Villages Project, founded by economist Jeffrey Sachs and supported by Angelina Jolie and others, aims to help nearly 500,000 out of extreme poverty. A paper published in June in The Lancet, a leading health journal, was scrutinized and roundly criticized for the logic and analysis it used to argue that observed changes were due to the Millennium Villages rather than changes already taking place in society.
The second row concerns treating children in less developed countries for intestinal worms, which are endemic in many countries. Because the worms share a child’s food, they are thought to contribute to malnutrition, reduced physical and cognitive development, and lethargy. Deworming children has been found by randomized control trials to reduced absenteeism from school, and hence is recommended by the World Health Organization and the Copenhagen Consensus Center, a think tank that publicizes the best uses of development money. But a systematic review and meta-analysis published last month by the respected Cochrane Collaboration, focusing on non-educational outcomes, found that “deworming children seems like a good idea, but the evidence for it just doesn’t stack up.”
The striking shift here is not in the details or merits of the specific programs, but in that these rows happen at all. They are precisely how science is supposed to work. For instance, Maxwell published his theory of electromagnetism, which turns out to be inconsistent with the maths of radiation from a black box, and from that tension arose the much broader quantum theory. Andrew Wiles published his “proof” of Fermat’s last theorem in 1993, somebody spotted an error, and Wiles revised and strengthened the proof as a result. In Einstein’s maths for his general theory of relativity, the Russian mathematician Alexander Friedmann found a term being divided by zero (“a complicated form of zero,” a physicist once said), which suggested—contrary to the prevailing view—that the universe is expanding, subsequently confirmed by observation and from which cosmologists have estimated the universe’s age.
International development has become much more scientific in the last 15 years: evaluating ideas through randomized control trials; publishing enough detail about a program’s methods and results that it can be replicated elsewhere; subjecting analysis to peer review; and publishing in respected journals. The organizations whose data are being contested should be proud that their data are capable of such contest. They contrast starkly with much activity in charities, philanthropy, and even social policy where performance data are often too scarce, too private, too vague, and/or otherwise too flaky to be meaningfully debated.
Science—knowledge—progresses through vigorous public debate about rigorous data. This process has shown that many things that everyone “just knew” to be true are actually false—from the “fact” that the Sun goes round the Earth, to the “fact” that severe brain injuries should be treated with steroids, a practice common until 2005 when randomized control trials showed it to be fatal. Similarly, many things which we “just know” to be true about international development are being shown by this careful, empirical, scientific approach to be false: providing more text books to Indian schools rarely actually improves learning; microcredit does not singlehandedly lift millions out of poverty; anti-malarial bednets should not be sold but rather given away for free; cooking stoves that use less wood as fuel do not always reduce respiratory diseases from reduced smoke inhalation.
The current rows are therefore a sign that international development is moving beyond “just knowing because I saw it with my own eyes” into properly understanding what works. We need more and better data to enable more quality debates on many subjects about development— debates that get settled, not by personalities or popularity or politics, but by the evidence.