Newest in the story of developmental economics

Chennai: Neelima Khetan, 48, and Angela Ambroz, 26, come from opposite sides of the globe and had never met each other until that evening in Patna. Inside a dimly lit banquet hall at a four-star hotel in the Bihar capital, they are both part of a chattering group of civil servants, social workers, executives from donor organizations and academicians from prestigious UK and US varsities.

Also Read Food incentives spur child immunization

Affiliated with the Massachusetts Institute of Technology, or MIT, the lab carries out randomized studies—largely used in clinical trials to test new medicines—to identify and test policies that could improve education, public health and governance. J-PAL’s evaluation is carried out on two groups—a treatment group where a new policy is implemented and a control group that remains untouched by the intervention.

J-PAL is using the method to ladder down the ivory tower of developmental economics and offer groundbreaking solutions to issues such as increasing teacher attendance in schools and boosting immunization levels among rural children.

“Initially, when we were out there, people thought of us as a wacky organization with wacky ideas and, often, we got treated exactly as such," Banerjee told Mint at a Patna dinner, a day ahead of J-PAL’s first conference in India to highlight its work and partnerships.

Yet, J-PAL’s work has perked the interest of economists who place the focus of development projects on evaluating the process by which they are carried out. An example is conducting checks to see if free textbooks offered by the government are reaching the students or not, rather than confirm if the textbooks are key to higher test scores.

Khetan’s Seva Mandir, which focuses on education, health and women’s empowerment in rural Rajasthan, was the first non-governmental organization (NGO) that Banerjee, a Harvard-educated economist and famed theorist, started working with in 1996.

Banerjee’s first task: evaluating the impact of an additional teacher in Seva Mandir’s single-teacher schools. Contrary to expectations, the study revealed no significant improvement in student performance on increasing the teacher count. So the NGO rolled back its programme.

This data-intensive experiment, conducted more than a decade ago, formed the basis of further efforts by Banerjee and Duflo, his star student from MIT.

Today, J-PAL has completed or is running 235 projects in 38 countries, managed by 44 professors. J-PAL is named after the father of Mohammed Abdul Latif Jameel, an MIT alumnus and president of Abdul Latif Jameel Co. Ltd—the sole distributor of Toyota Motor Corp. vehicles in Saudi Arabia—who offered a substantial grant to the group.

Evaluating failure

“Our results about Seva Mandir’s programme were negative, and still they had a mindset to take evidence seriously. And this was at a time when we had no reputation," said Banerjee, who recently won the Infosys Prize for contribution to scientific research.

Duflo, a French national, made it to The Economist magazine’s top eight young economists list. She likes to define her team as a loose network of academicians, the government and NGOs glued together by technique. “Our team is united by a methodology of organized evaluation," the 37-year-old said.

“For instance, how do you get poor people who do not have the luxury of automatism to immunize their children? Incentives may be a way to do it. We train graduate students round the year to do this kind of work and create a culture," she said.

Being recruited as a research assistant for the group was a dream come true for Ambroz, an American student who focused on developmental economics while pursuing her master’s at the UK’s University of Oxford. As one of 12 research assistants in India, Ambroz aided data collection efforts for a J-PAL study to evaluate the impact of microcredit for Andhra Pradesh-based microfinancier Spandana.

“It (J-PAL’s work) is a new trend in development and you recognize that you are a part of something big," said Ambroz. “All of the directors are passionate about it and want to think about issues to be resolved. So, you get inspired and when you get into the field itself, you get even more inspired."

Policy outreach

Glennerster’s experience with the UK treasury won her the top nomination for dealing with policymakers. “My background is that of a policymaker and Abhijit (Banerjee) is the one who links it back to theory after challenging the assumptions in the making, " said Glennerster, 44, who joined J-PAL in 2004.

The economist, who was formerly with the International Monetary Fund, emphasizes the need to make evaluations scalable. “We are not interested in community mobilization that is costing millions of dollars in a village, because then you cannot scale," said Glennerster.

With nearly 51 evaluations running in 13 Indian states, the subcontinent is the group’s biggest hub.

To jumpstart policy efforts, particularly in India, Glennerster hired Iqbal Dhaliwal, a Princeton University graduate and a former bureaucrat who has worked with the Indian government in several capacities. He is acutely aware of unwritten protocols that show preferences for face-to-face meetings versus a phone call. At the Patna dinner event, for instance, Dhaliwal doesn’t miss out on introducing a member of the hotel staff to his table, announcing him as key to the evening’s success. And during the conference the following morning, he breaks his speech to welcome a chief secretary of the Bihar government who was taking his seat.

Dhaliwal agrees there are people in governments around the world who are so checked out and cynical that he wouldn’t even approach them. Then there are truly brilliant policymakers who don’t need any help. “But there is this middle 60% who really want to do something good but just don’t know how to do it," said Dhaliwal, who organized a meeting of the J-PAL team with Bihar chief minister Nitish Kumar earlier that evening. “My job is to identify that middle 60% and target them."

J-PAL’s Udaipur study, also done through Seva Mandir, revealed a jump in full immunization rates when nurse attendance was improved using camera monitoring and parents were offered 1kg lentils after every vaccination shot on their child.

“Donors have become more particular about measuring and tracking impact," said Khetan, Seva Mandir’s chief executive. “Having this absolutely 100% squeaky clean reliable data is necessary. But whether that evidence will be converted into action or not, is an attitude and politics that you will have to deal with."

anupama.c@livemint.com