Nobel Prize laureate Guido Imbens. Photo by Elena Zhukova


How can you measure the unmeasurable? That’s the question Guido Imbens and Joshua Angrist wanted to answer. And so they did, in a manner of speaking.

In 2021, the two college professors won the Nobel Prize in Economic Sciences for their pioneering research on new methods of using econometrics and statistics to simulate policy experiments that might be unethical or too expensive to test in real life. Or, as the New York Times explained it, “They have developed research tools that help economists use real-life situations to test big theories, like how additional education affects earnings.”

The duo won the prize jointly with David Card, who used natural experiments to analyze the labor market.

We recently spoke with the Dutch-American Guido Imbens, a professor of economics at Stanford University, about his thoughts on economics, winning the Nobel Prize, and how he sees the research environment in the United States.

The idea for the research first came to Imbens and Angrist in a laundromat in the 1990s, when they were both assistant professors at Harvard University.

1. What happened when you won the Nobel Prize?

When I received a call from Sweden in the middle of the night, my three children got up and started to make breakfast for the Stanford people who came over to help with the media that showed up at my door. We have chickens at home, and my kids made scrambled eggs and pancakes for the media crew. It was a very proud parenting moment.

It feels great that the sort of work that I had done was being recognized, and I am thrilled to share the prize with two very good friends: (economist) Josh Angrist had been the best man at our wedding, and David Card is also a very close friend, we’ve known each other since the early ‘90s. Our work wasn’t always viewed very positively, and there was lots of pushback early on. All of us have gone through this whole journey together.

2. Who inspired you to do research in econometrics and statistics?

My older brother had decided to study mathematics. I was kind of leaning toward mathematics as well, but I was looking for something different: math that could be applied and have an impact on society. One day my high school economics teacher gave me this little econometrics book by Nobel laureate Jan Tinbergen, one of the founders of econometrics. It appealed to me. Tinbergen was an inspirational figure. He was doing lots of government advising, as well as great academic work. He had founded an economics program at the Erasmus University in Rotterdam in the ‘60s that looked very exciting to me, and so I decided to enroll in the program.

3. How would you describe your research interests in econometrics?

I’m interested in the causal effects of policy changes. For example, what is the effect on earnings of getting more education? What is the effect of early childhood intervention programs? What is the effect of military service on the labor market? These are areas where we can’t run experiments, because it would be too expensive, or it wouldn’t be ethical, so we need to come up with clever ways of teasing out these effects and use data to draw conclusions. My research is all about trying to figure out cases where we can credibly find causal effects that are of importance to policymakers. We exploit idiosyncrasies in the system to get at credible causal effects. For example rules about when children have to enter school, or how priority for various services is determined, or arbitrary cutoffs for being held back in school.

4. What are some interesting experiments that illustrate your methods?

One of my research projects was to see what would happen if you gave everybody some guaranteed income. Clearly, here you could not do an experiment giving a large number of people universal basic income for a number of years, as the long-term implications of that would be incredibly expensive. But what we did was look at the lottery, which was essentially doing that experiment for us by giving randomly selected people large sums of money over a long period of time. In that way it resembles a universal basic income  We found out that most of the people who won the lottery just kept working, may be just a couple of hours less. Looking at it this way gives you a credible way of understanding what the effect is of having unearned income, without doing an experiment.  

Another example of my research was to look at the effect of military service. Together with my Dutch colleague Wilbert van der Klaauw, we looked at the effect on future earnings of the military service in the Netherlands. In the late ‘70s, the military didn’t need quite as many people in the draft as they were getting given the size of the birth cohorts, so they decided that one birth year cohort was going to be completely exempt; the men born in 1959 did not have to do military service. These were not smarter or dumber or different from those born a year before or after, but we found they have slightly higher earnings than those born a year before or a year after. Our estimates suggest that the effect of serving in military service on yearly earnings was roughly equivalent to the cost of losing a year of work experience.

5. Could you give an example of a challenge in working with your students that you had to overcome?

Working with students is one of the most enjoyable parts of my job. The transition from students just taking courses to actually doing research is always challenging, especially with coming up with new things and ideas. I talk to the students during that process, as it is very easy to get frustrated and lose hope. I’ve had the same feelings. When I was in graduate school, after my first year, I wasn’t really sure I wanted to continue doing research. I wasn’t sure I was going to be cut out for this and started applying to a job in banking in the US. They needed someone who had a master’s degree in economics and was fluent in Dutch and I thought I was perfect for that job. They didn’t even interview me! I figured that I’d stick with the PhD program. Research is challenging, and at times it is a very lonely thing. It’s great when it goes well, but it is never going to go well all the time.

6. How can people outside of academia learn from your research?

There are an enormous number of tech companies, especially in the Bay Area, doing experiments using data science these days. They are so close by that I can just ride my bike over there if I want to talk to people from these companies — Apple, Google, Facebook, and more — and a lot of the questions they are interested in are very similar to the questions outside of the tech world. I am currently doing work for Amazon and we developed a very novel way of designing experiments that is particularly relevant for these companies. I think it is incredibly stimulating and inspiring to talk to people at these tech companies and see how they think about these problems and what type of questions they’re working on. For example, they’re often not interested in what the effect of an intervention is  next month, but they want to know what the intervention will do two to three years from now. The econometric research I do can help answer that. But that does not just answer that question, it is also relevant for many problems in other areas. For example, I worked with Raj Chetty, professor of economics at Harvard, who’s very interested in the long-term impacts of childhood intervention programs, like the Head Start pre-K program in the US. We’ve done a lot of experiments where we look at the effects of test scores a one or two years later, but what we really want to find out is if programs like these help children 20 years or more later in life? It’s the the same problem that these tech companies were thinking about, and the solutions I developed work in both settings..

7. What is it that you like best in the US research system?

There are incredibly strong research places with researchers that are very inspiring and stimulating, like Harvard, MIT or Stanford. Having a group of really high-quality people to talk to continuously in the next office or in the hallway is very inspiring. In addition, the top universities here are very broad. They try to be good at everything. It is also less siloed than in the Netherlands, where if you would come into a program, such as economics or law, there are fewer opportunities to take courses in other areas. Here in the US students specialize later, and that has some disadvantages, but it does make it easier to connect with people from other departments.

For the work I do, it is really important to interact with computer scientists as well as with people in political science and in statistics. At Stanford, all these groups are nearby, so it is very easy to get in touch with them. My students take classes in computer science and statistics. Interdisciplinary research here is very well organized. When I was at Harvard, I had the statistics department next door, so I started talking to people there and  I  taught classes together with people from the statistics department. And the universities here are good at integrating people from other cultures, both first and second generation immigrants. When I look around here in my business school, there’s someone from Iran in the next office or someone from Belgium or someone from France. There are people from India. There are people from China. It’s very diverse, and that is partly because the school system in the US doesn’t specialize quite so early, so it is easier for non-native English speakers to do well.

8. Now that you’ve won the Nobel Prize, what is it still that you hope to achieve?

The prize is going to make all the things that I wanted to do before easier. I just love my research and working with students. I’ll probably shift a bit more to mentoring junior people and doing things for the profession and part of that is that I’ll be doing more in the Netherlands. I’ll be going to the Netherlands in July to give talks and I hope to contribute more to the tradition started by Jan Tinbergen helping the Dutch economic research move forward.