fb-pixelMIT researchers are fighting extreme poverty with science - The Boston Globe Skip to main content

MIT researchers are fighting extreme poverty with science

Study co-authors Abhijit Banerjee and Esther Duflo in India in 2007. Both are directors of the Abdul Latif Jameel Poverty Action Lab at MIT.Courtesy of David Baron

One of the most pressing social, moral, and economic quandaries of our time lies in this simple and depressing fact: more than a fifth of the people on the planet live on $1.25 or less a day. What can be done to lift up the "ultra-poor"?

A team of MIT researchers has harnessed the tools of science to test one promising solution: a "big push" approach that helps poor people establish a livelihood.

The three-year study, implemented in six countries starting in 2006, gave the poorest households a variety of supports, including a small savings account, mentoring, a resource such as goats or chickens, and money for food. A year after the intervention ended, there were fewer skipped meals, more income from livestock and farming, and a durable, though small, increase in how much they consume each day.


Elements of the program may sound familiar, but the study is part of an ongoing revolution in which the public policy world draws on the rigorous methods used to test new drugs in order to understand how to help poor people. More than 10,000 households were randomly assigned to either receive the "graduation" intervention or maintain the status quo, and the analysis in the journal Science on Thursday found the intervention more than pays for itself and has lasting benefits.

The fight against global poverty has often utilized policies that depend on heart-rending anecdotes about successes or failures — something that the economists who led the work guarded against. Esther Duflo, a leader of the study and an economist at the Jameel Poverty Action Lab at the Massachusetts Institute of Technology, said that in the field, the extreme poor often seem mired in an intractable, desperate situation.

"What is remarkable when we see the data" is that people are able to go from a hopeless situation to "doing something productive," Duflo said. "The ingenuity was there, somewhere, but buried under despair."


Duflo and her coauthors are part of an all-star team at the forefront of a shift in how interventions aimed at addressing global poverty are evaluated.

"It's a big paper. I hope — I'm sure — it will produce waves," said Laurence Chandy, a fellow at the Brookings Institution who was not involved in the work. Chandy said the approach will garner wide attention because it was written by so many leaders in the field and because tackling extreme poverty has become an international priority.

But he added that it is not, by any means, a complete solution for extreme poverty. There are a number of caveats — for example, the improvements in absolute terms were small, about 10 to 15 cents improvement in how much people were able to live on each day.

The work will also add to a raging discussion in the policy world about how to best reduce poverty. This type of discrete intervention can't deter many of the major contributors to poverty, such as corruption or government instability, according to John Crowley, a chief within the social and human sciences sector at UNESCO.

Dean Karlan, a development economist at Yale University and another leader of the work, said the strength of the paper is its demonstration that the program had success in so many different countries and cultural contexts; often a study showing something that works in one country may not apply more broadly.


The intervention was individually tailored for each country, with families able to choose a livelihood asset that ranged from bees to goats to guinea pigs. In most countries, households were given health care support and either food or financial support to ensure they did not use up their asset right away.

In addition to the savings account, they also received mentoring and regular visits from a local nongovernmental organization. The up-front costs ranged from $1,455 per household in India to nearly $6,000 in Pakistan.

The programs all began at different times during the past nine years. In each one, two years after the program began, the households received their last visits from nongovernmental organizations, and a year after that researchers returned to evaluate the long-term effects. Overall, there was a 5 percent increase in the amount people had to live on each month, nearly a doubling of the amount they saved, a 37 percent increase in monthly revenues from livestock, a 9 percent increase in farming income, and a similar increase in their own evaluation of their economic status.

A cost-benefit analysis of the study found that the benefits outweighed the considerable up-front investment during the three years in Ethiopia, Ghana, India, Pakistan, and Peru. The intervention in Honduras failed to show a favorable cost benefit, in part because the asset that was given to the households tended to be chickens, and they got sick and died.

The results have been helpful in persuading other governments to give the program a try.


Tatiana Rincón of Fundación Capital, a development organization that works to end poverty, has helped set up and expand a program in Colombia.

Carolyn Y. Johnson can be reached at cjohnson@
. Follow her on Twitter @carolynyjohnson.