Asim Khwaja and Michael Kremer before the event, 2019.

At a CID Event, Nobel Economics Laureate Michael Kremer Describes the Value of a Synergistic Relationship Between Research and Practice.

By Rosemary Berberian

 

International development needs scholars from a variety of backgrounds and the input of practitioners, Nobel Laureate Michael Kremer told the audience at a special event organized by Harvard’s Center for International Development.

The event took place on Tuesday, November 5th in Harvard Commons, the Smith Center’s public auditorium. Throughout the evening, Kremer’s insights, the questions posed, and the sheer diversity of the expertise represented in the room came together to send a clear message. Kremer’s work testifies to the synergistic relationship between research and practice: efforts to affect real change in the world are most fruitful when the academic exercises of theory and experimentation are undertaken alongside and directly informed by policymakers, practitioners, and the research beneficiaries themselves.

Here are some excerpts from Kremer’s talk, lightly edited for length and clarity. Listen to the full event on CID’s podcast “Michael Kremer in Conversation with Harvard Students.”  

Michael Kremer speaking at Harvard Commons Smith Center, 2019.
On an inclusive approach to poverty reduction:

International development work requires, and is naturally suited to, teams of people with very different subject matter backgrounds … from economists to public health specialists, education specialists, and psychologists. Depending on the project, it also involves very deep interaction with government workers, NGO workers, and ultimately the clients of the research—students in a school or farmers—who are motivated by certain issues and will offer valuable insights. What’s perhaps unusual about this field is that we feel we’re all working toward a common goal.

On his origins as a development economist:

I had been interested in international development and cared about this issue—I know that Esther and Abhijeet, my co-laureates feel the same way—in part because of how I was raised as a child by my parents who believed that we have certain obligations to help address injustices in the world.

I was a social studies concentrator as an undergraduate and benefited immensely from that program. People who work in international development very quickly realize the interdisciplinary nature of the problems. Social studies for me was very helpful because it offered me exposure to a lot of different fields and different ways of thinking about issues, as well as to the great thinkers that one reads in social studies.

I took more economics classes over time, and I took a class with [a professor] at the precursor to CID—the Harvard Institute for International Development—who worked in Kenya. Through him, I applied for funds—as I assume many of you will—to do my senior thesis research over the summer in India and in Sri Lanka. I received the funds, went to Sri Lanka, went to the library and the government offices to look up the records on the program I was studying, and diligently wrote them all down. Then at the end of that summer, I got the chance to visit a village and I realized what I’d been missing out on the rest of the summer. Having done that, I realized that if I wanted work in this field, it was important for me to spend more time in developing countries.

On his first use of randomized controlled trials:

After graduating, I got a job at MIT and went back to Kenya to visit some friends of mine while on vacation with my wife, Rachel Glennerster, who is very active in this field and was the head of the Poverty Action Lab at MIT. We were visiting a Kenyan friend of mine, who had been a headmaster of a school when I was teaching, but who had gone on to work for a very small NGO. They were starting to work in a new area of Kenya where they hadn’t worked before and wanted to work with schools, but they were considering a bunch of different ways of working. As we talked about it, they agreed that they wanted to explore how to operate most effectively.

What they decided to do was to try different programs in different areas. Since they were just starting and needed to phase in gradually, they decided to time the phasing in such a way that they had comparable groups of schools that were phased in earlier and phased in later. This meant that of part way through the implementation, you could compare the schools that had already received a particular program to others that hadn’t. Of course, it might seem like basic logic to do something like that, but this was not common at the time. It’s a very simple approach and obviously used in medicine and many other fields, and it has proven to have a huge effect on the field.

When I got involved, I was interested in this approach primarily from the standpoint of stronger and more reliable evidence. Trying to statistically control for various factors and to isolate the impact of a program is very hard, and a lot of work in the field suggested that we weren’t very successful at that. And [randomized controlled trials] seemed like a good way to get at that. What’s turned out to be the case over time is that in addition to any benefits from better causal inference, [the interactions on the ground] were another very important aspect of this [approach]. Economists tended to do models with paper and pencil or to analyze data sets, which were both very valuable approaches. But there’s also something to be gained by richer interactions, such as talking to farmers, teachers, students, government officials, and people working at NGOs. These interactions bring in all sorts of new kinds of insights.

Michael Kremer speaking with students following the event, 2019.
On future work:

Economists have been pretty shy about working on issues of culture. Our tendency is to assume that people are people everywhere—you can just look at the fundamental economic forces of supply and demand or game theory, and that people are going to work the same everywhere. There’s exciting work in economic history and more broadly to try to understand the role of culture. My Harvard colleagues Nathan Nunn and Alberto Alesina do a lot of that work. That obviously involves bringing in insights from other disciplines. I think that’s one of many exciting areas.

I’m obviously very excited about the movement toward randomized trials, but the economic theory and macroeconomics of development are also very important. I hope there will be more work in those areas as well.

Listen to the full event on CID’s podcast