Nobel laureate's experimental approach aims to tackle root cause of poverty
In 1985 a new college graduate named Michael Kremer traveled to Kenya hoping to learn more about a topic he had studied at Harvard: economic development in low-income countries. He didn’t anticipate that an official in the village where he was staying would invite him to teach at a new school—but Kremer stayed in the country for a year to do just that. In fact, he cofounded an organization, WorldTeach, to send more teachers to Kenya and ultimately to several other developing countries.
Given the opportunity to make a difference, Kremer jumped in, backing up a moral commitment to improving human life with hands-on work and entrepreneurial spirit.
When he returned to Kenya in the early 1990s, Kremer was just starting his academic career as an economist at MIT. His plan was to meet up with old friends and spend time just “being a tourist,” he says. But the opportunity to help arose again. One friend, a Kenyan who was working for a small Dutch nongovernmental organization operating in Africa, was trying to identify schools in which to roll out a program intended to improve learning. Kremer offered a casual suggestion: select schools in a systematic way, then compare outcomes in those schools with outcomes in the same number of similar schools where the program was not operating.
After returning home Kremer heard from his friend that the NGO was interested in trying out the approach. It was the start of a relationship that helped determine the trajectory of Kremer’s research in economics: he got involved not only with the Kenyan school study but with other experiments being done by the organization, today called ICS Africa, helping it develop evidence-backed programs.
Kremer’s desire to help was evolving into a scientific methodology, one that would complement the most established approaches in his discipline.
“Randomized controlled trials had been used in medicine for a very long time,” he says, noting that some economists and other social scientists had also used them. And yet the technique had not broken through widely in economics, which largely either focused on theoretical models or relied on existing data. Early in Kremer’s career, the subfield of development economics—which studies economic issues in low- and middle-income countries—was no exception.
We know a lot less than we think we know.
- Prof. Michael Kremer
The results that came back from the Kenyan schools were unexpected. Studying a rural part of the country that had one textbook for every 17 students, Kremer and fellow researchers hypothesized that increasing the number of textbooks would increase average test scores. Their 2009 paper concluded that it did not. Only the the students who scored highest on pretests improved, since the intervention failed to address a deeper problem: a curriculum heavily geared toward the strongest- performing students. Historically, Kenyan schools have been judged on their ability to produce a handful of excellent students, not how well they meet the needs of the majority. Once it was clear that more textbooks alone wasn’t the solution, other interventions—such as remedial education for students who have fallen behind and flexibility for different schools to cover materials at their own pace—could come into view. Since then, other researchers have worked with NGOs, particularly Pratham in India, to develop such interventions, which now reach hundreds of millions of students each year.
Kremer and his colleagues had hit upon a virtuous circle: an experimental result can uncover hidden factors, suggesting new directions for intervention and evaluation. This process can be repeated for as long as it proves fruitful.
Kremer was born in New York City to parents who were both children of Jewish immigrants from Europe. He came of age in the orbit of Kansas State University, where his father, Eugene Kremer, taught architecture, and his mother, Sara Lillian Kremer, was a professor of English. He credits his mother, who authored two books on literary representations of the Holocaust, with teaching him the necessity of addressing preventable suffering and injustice in the world—the key motivation behind his work, from founding WorldTeach to conducting research in Kenya and beyond.
Kremer and his wife and collaborator Rachel Glennerster—a British development economist who joined the UChicago faculty as an associate professor in 2021—take this intention to heart in their research as well as in their personal philanthropy: they’re part of the Giving What We Can Pledge, a commitment to give 10 percent or more of their lifetime income to high-impact charities.
To prevent suffering, from an economist’s perspective, means asking what interventions will do the most good in the most efficient way. Kremer holds up another ICS Africa program as an example of how a “catalytic investment” by an NGO can scale up. This program addressed the problem of intestinal worms, parasites that infect hundreds of millions of people globally—over a quarter of the world’s population—especially in warm areas with poor sanitation. Sometimes the worms lead to debilitating digestive, nutritional, and developmental problems.
In the early 2000s, Kremer and a colleague set out to study a group of 75 schools in Kenya. They found that a school-based program providing deworming pills to all students reduced absenteeism by more than 25 percent—a staggering result improving the health and well-being of the children directly affected, and likely of the entire local population and its economy over time. In their 2004 paper, the authors argue that the positive spillover effects easily justify providing free universal treatment to school-age children in high-risk areas.
Officials took note. After the study, with the support of NGOs, the program was scaled nationwide by the Kenyan government. From there it was picked up by multiple Indian states—and then by the Indian national government. Today the world’s second most populous country holds National Deworming Day twice a year, administered through public schools.
Bringing actionable research findings to the governments of developing countries is a major goal of Kremer’s research. “We’re not mainly writing for foreign aid donors or philanthropists,” he says, although he’s happy if they read his work. “The big win is if the Indian government decides to scale up deworming, reaching hundreds of millions of people.”
Helping to establish the experimental method in development economics earned Kremer a share of the 2019 Sveriges Riksbank Prize in Economic Sciences in Memory of Alfred Nobel. In awarding the honor to Kremer, Abhijit Banerjee, and Esther Duflo, the committee heralded a burgeoning approach to reducing global poverty that “involves dividing this issue into smaller, more manageable questions.”
Standing on stage at Stockholm University, Kremer opened his acceptance speech by describing the mentorship of a Kansas State University theoretical physicist, Larry Weaver. He recalled that he was surprised one day when Weaver mentioned being happy that a colleague had discovered an error in his work. He was happy, Kremer explained, “because he cared about advancing science.”
Kremer views experimental economics as complementing the more abstract, model-based work that has long characterized the mainstream of the discipline. As he put it in his Nobel lecture, Kremer began his academic career working “mostly alone or with one or two coauthors, using models and data which others had collected.” It was serendipity, he said, that his time in Kenya gave him the chance to work in the field, to collaborate with people in other fields, and to experiment with randomized controlled trials—a method that could “cleanly isolate the causal impact of a program or policy from confounding factors.”
Through doing that work, one experiment at a time, he came to see how the experimental method could be elevated to “a fundamentally different type of economics research,” one distinguished by the richness of its empirical context and the specificity of the practical questions it attempts to solve.
This way of working also requires a high degree of collaboration. “It’s obviously fantastic to get the prize,” Kremer says, but he is quick to spread the credit to a broad group of partners that includes fellow development economists; the personnel collecting survey data; allied researchers from disciplines such as education and psychology; funders like the British Department for International Development (now the Foreign Commonwealth and Development Office), the World Bank, and the US Agency for International Development (USAID); and the farmers, teachers, and other people he has met and learned from. “You wind up spending a lot of time in the field,” Kremer says. “That exposes the economist to a richer set of ideas.”
In his Nobel lecture Kremer shared his greatest hope for the work he does: “innovations in social institutions designed to accelerate scientific and technological change and orient it toward human needs.”
In 2020 Kremer brought his engaged experimental approach to the University of Chicago when he joined the Kenneth C. Griffin Department of Economics and the Harris School of Public Policy as a University Professor. It’s technically not the first time Kremer has taught on the quads—he was a visiting professor in the spring of 1993—but his responsibilities are much greater today. Notably, he is serving as faculty director of the new Development Economics Center, part of the Becker Friedman Institute for Economics (BFI).
Announced in January, the Development Economics Center continues the interdisciplinary work of BFI’s previous Development Economics Research Initiative: developing collaborative approaches to the challenges faced by lower-income countries, engaging with policy makers, and awarding grants to UChicago PhD students and junior faculty working in the field. The Development Economics Center also includes the Weiss Fund for Research in Development Economics, to which Kremer and his co-recipients donated their Nobel Prize money, and the Development Innovation Lab, where Kremer has been serving as founding director since his arrival in 2020.
“The University of Chicago is extremely fortunate to have Michael Kremer leading our collaborative development economics efforts,” says BFI director Michael Greenstone, LAB’87, the Milton Friedman Distinguished Service Professor in Economics and the College. He calls Kremer “a true pioneer in conducting experiments with randomized controlled trials to explore the root causes of poverty in developing countries.”
That work, agrees economics chair and Alvin H. Baum Professor in Economics and the College Robert Shimer, “has been really transformative for the type of research people are doing in development economics, the kinds of questions they’re asking, and outcomes that are improving people’s lives around the world.” Even so, this tells only part of the story. Kremer “is a broad economist whose work is both theoretical and empirical,” Shimer adds, and the fruitfulness of his experiments relies on the strength of the theory informing them.
“To do good empirical research, you also need to understand economic theory to have a sense of where to look for solutions,” Shimer says, pointing out that Kremer’s theoretical acumen has been applied widely. For instance, building on a body of past work on the economics of vaccine development, Kremer has turned much of his attention since arriving on campus to the pressing global public health issue that emerged around that time: COVID-19.
Early in the COVID-19 pandemic, Kremer was asked by several governments and international agencies for advice on vaccine financing. He assembled the Accelerating Health Technologies group, which includes economists as well as biostatisticians with expertise in epidemiology and pharmacology. Together they conducted analysis and made recommendations to policy makers. For example, they advised the US government in the run-up to Operation Warp Speed, the public-private partnership initiated in 2020 in which different vaccines were developed simultaneously and produced before their efficacy was known. Speed was critical, and it would have been “penny-wise and pound-foolish,” Kremer says, to delay building large factories to develop these unproven vaccines, even if the facilities cost millions of dollars.
The United States and United Kingdom went for this approach, but Kremer says he would have preferred to see it happen on an even larger scale—with vaccine production capacity being built in advance for as many as 15 or 20 different vaccines. The fact that multiple vaccines worked for COVID-19 is “really lucky,” he believes, given that there are far longer-known viruses like HIV that we still have not figured out how to inoculate ourselves against.
Vaccines also raise an issue that economists think about regularly: supply and demand. After effective vaccines are developed, there is still the question of getting them where they are needed. Chicago Booth’s Eric Budish, the Paul G. McDermott Professor of Economics and Entrepreneurship, and Canice Prendergast, the W. Allen Wallis Distinguished Service Professor of Economics, have contributed to the work on this problem. Building on his prior research on using market mechanisms to meet demand at food banks, Prendergast—also an affiliate of the Development Economics Center and a member of the Accelerating Health Technologies group—worked to design a system that would allow a country to send its surplus vaccine supply to other countries in need and then get paid back with doses later, when its own supply falls low.
For the past couple decades Michael has been one of the leading thinkers on incentives for creating new vaccine research.
- Prof. Christopher Blattman
Kremer’s contributions to the world of vaccine development predate the emergence of the novel coronavirus. In 2004 Kremer and Glennerster proposed something called an advance market commitment (AMC) in their book Strong Medicine: Creating Incentives for Pharmaceutical Research on Neglected Diseases (Princeton University Press).
The coauthors were aiming to address a basic problem in market economies: no matter how many people need a product, they will not constitute a market for that product if they do not have the money to pay for it. Malaria killed over 600,000 people in 2020, for instance, but markets do not necessarily respond to sickness and death.
An AMC creates a market where there was none. It asks donors to agree in advance to subsidize the initial purchase of effective vaccines by lower-income countries if those vaccines are developed. The purchasing country pays a nominal copayment; the vaccine maker agrees to a price affordable in the developing country; and the donor fund contributes a per-dose top-up price to ensure the company’s investment in research and development is repaid. In this way, private firms are incentivized to undertake research and development for treatments that might not otherwise be profitable.
Kremer’s studies use economic analyses to design interventions for human good, and AMCs show how diverse these interventions can be: a change in curriculum, a deworming pill, or an entirely new kind of market.
In 2007 a pilot AMC intended to spur a vaccine for pneumococcus—the main cause of pneumonia globally—netted a pledge of $1.5 billion from a group of donors comprising the Bill & Melinda Gates Foundation and the countries of Canada, Italy, Norway, Russia, and the United Kingdom. The AMC kicked off in 2009 under the supervision of Gavi, the Vaccine Alliance (formerly the Global Alliance for Vaccines and Immunization), with firms competing for 10-year vaccine supply contracts intended to inoculate children in more than 70 countries meeting specific criteria. By the time of a 2020 paper, Kremer and two coauthors, all of whom had helped design the pilot, were able to report the development of three vaccines by different firms, 150 million children immunized, and an estimated 700,000 lives saved.
“For the past couple decades Michael has been one of the leading thinkers on incentives for creating new vaccine research,” says Christopher Blattman, the Ramalee E. Pearson Professor of Global Conflict Studies, citing the pneumococcus vaccine as well as a promising new malaria vaccine as examples of incentives at work.
AMCs are not limited, in principle, to vaccine development. It’s a heady idea: if there is no market for addressing some need, you can try to create one.
For instance, AMCs may become a new tool in the fight against climate change. In a story that cites the pneumococcus AMC as precedent, National Public Radio has reported on the idea of using an AMC to incentivize entrepreneurs to develop carbon-removal technologies. While climate activists agree that carbon emissions must be reduced at the source, the idea of extracting massive amounts of existing carbon out of the atmosphere is gaining traction. Among those calling for a carbon-removal AMC is Glennerster, who helped make the case for the strategy in a December Politico opinion piece, floating the idea of a $1 billion commitment to provide a “critical demand signal to scientists, founders and investors, giving them the confidence needed to begin building now.”
Again and again, Kremer has conducted small studies to approach big problems. In many ways, these studies are a natural complement to existing research at UChicago. For instance, Kremer has used ideas from behavioral economics—another empirically informed subfield of economics that has grown in recent decades thanks to the work of 2017 Nobelist Richard Thaler, the Charles R. Walgreen Distinguished Service Professor of Behavioral Science and Economics, and others at UChicago. As opposed to neoclassical economics, which assumes that people act according to well-defined, rational preferences, behavioral economics draws on insights from psychology to model and describe people’s real-world behavior.
For example, in addressing the problem that hundreds of thousands of children die of diarrhea each year due to unsafe drinking water, Kremer and colleagues did not merely test the efficacy of a treatment (a simple dilute chlorine solution). They also asked how to get people to use it. What is the right “nudge,” in the language of behavioral economics, to get the desired behavioral result?
As with intestinal worms, a promising treatment was at hand, ready to be backed up by experimentation. Water treatment with chlorination is cheap, effective, and widely available. But in rural Kenya, few people treated their water. Kremer and his colleagues found that placing dispensers of water treatment solution at public water sources substantially increased usage. Public dispensation gives people the convenience of adding the treatment right as they draw their water—and the fact that it’s public means they are modeling the practice for others. Sensitivity to habits, biases, and social cues, as opposed to abstract preferences, proved useful to understanding and influencing people’s actions in the real world.
Notably, Kremer’s prior research had already indicated that use of the water treatment significantly increased when it was free instead of merely low-cost, a finding that runs counter to the notion that you must charge something for a product to make people value it. For this kind of product under these conditions, charging nothing is demonstrably better than charging next to nothing.
Christina Brown, a Saieh Family Fellow at the Becker Friedman Institute, says that reading Kremer’s research as a master’s student showed her that she didn’t have to choose between academic research and making a difference in the world. Brown, who is slated to join the economics department as an assistant professor in 2023, will work along with Kremer to further bolster development economics at the University.
“While there have always been fantastic development economists at Harris and Booth, the economics department did not have as many scholars in that area as of a few years ago,” Brown says. Now, she believes, the University of Chicago is “in the top two or three places in the world to study and do development work.”
As Kremer inspires young scholars in his field—and funds them through the Development Economics Center—he continues pursuing his own research into the causes and conditions of poverty. Commentators tend to have a favorite go-to solution for the problem, he notes: more education, more physical capital, more democracy, and so on. In his experience, though, the solution depends critically on the country, the local community, and the sector of the economy. “We know a lot less than we think we know,” he says, and focusing on the specifics can “break you out of some theoretical straitjackets.” Deeper understanding is often just a well-designed experiment away.
—This story was originally published on the University of Chicago Magazine website.