You were awarded the Nobel Prize. How does that feel?
“It’s a strange feeling. Firstly, I was awarded the prize mainly for work I did about thirty years ago. That’s a long time ago. Secondly, it’s not as if the way I’ve worked in the last few weeks has changed a lot because of the prize, although it is suddenly getting a lot more attention, even from media that don’t normally write about economics. All things taken together, this is a very unusual situation. I’ve never gone through anything like this before.”
Does winning a prize like this come as a great surprise or were there any signs you were going to get it?
“Rumours do fly and obviously people talk about it. But generally, people aren’t awarded prizes such as this one until they reach a much more advanced age. I wasn’t lying awake at night, wondering if I’d get it this year. There was a chance I’d get it some day, but there are plenty of other people who are also in with a chance and will never get it.”
You were awarded the Nobel Prize for your Local Average Treatment Effect Model. You’ve created a wonderful video in which you explain to your children what that means. Can you explain it to me in even more basic lay terms?
(Laughs) “That video is a very good start. The children asked very good questions, even though the video was recorded at around 4am local time, when they’d been up all night. You see, I got my phone call about the Nobel Prize at around 2am. It’s remarkable they were so clear-headed at that time of night.
“Basically, my research focuses on estimating causal effects. In the social sciences, we are more interested in causal relationships than in correlations, as they are useful for policymakers, who want to know what kind of effect certain measures will have. The work I was doing with Josh Angrist (who was also awarded a Nobel Prize – ed.) was about estimating causal relationships in cases where the sample was not entirely random, but also depended on the choices people made. We made certain. assumptions about how people came to make those decisions. They turned out to be applicable in many more fields. It’s inspired by Tinbergen’s work, which is about the fundamental principles of economics.”
Jan Tinbergen is one of your predecessors as a Nobel economics winner, and a big man in Erasmus University history. Do you have any memories of your time here?
“I have very fond memories of it. The Econometrics department was small. We started out with sixty students, and after a year, only twenty were left. The degree programme did an excellent job of preparing me for a doctorate in either economics or econometrics. In the end, I chose not to complete it. In my third year I went to England on an exchange programme. They found I was so advanced I was allowed to do a master degree there. Afterwards I went to America and never returned to the Netherlands. Erasmus University was hugely influential in my career, though. I learned a lot in a brief period of time.”
What type of student goes on to win a Nobel Prize?
“I was a pretty good student. I worked hard, but also managed to have a good time. I was a member of the RSG student society, where I made a lot of friends. I definitely didn’t spend all my time working.”
Tinbergen has a square and building named after him on campus. Would you like to have your contribution to science commemorated on campus in any way?
(Laughs) “Tinbergen is a major role model. I don’t want to raise myself to his level. When I was attending secondary school, my economics teacher gave me a book written by Tinbergen, which was very inspiring. When I was a student in Rotterdam, I sometimes attended his lectures. But he’s not just famous for his academic work. He also made a great contribution to society. For instance, he founded the Netherlands Bureau for Economic Policy Analysis, as well as the field of econometrics. He really did a lot. I definitely don’t compare as yet, I feel.”
At present you are conducting a lot of research on machine learning and artificial intelligence, in association with companies such as Facebook and Amazon. Can you tell us more about that?
“Machine learning and artificial intelligence are increasingly important in economics. Those techniques are capable of a lot more than some of the traditional methods we use, but in themselves they generally don’t suffice to answer the kinds of questions we economists are interested in, which tend to be causal questions. Machine learning is good at predicting things but not at identifying causal relationships. The challenge lies in combining these new techniques and our models in such a way that they present plausible answers to causal questions. That’s something for which there is a great demand, both in academia and in the private sector.”