The groundbreaking work in Econometrics performed by EUR alumnus Guido Imbens and his colleague Joshua Angrist has set new standard for economics research. The methods he developed are now being taught in lecture halls where he himself once studied Econometrics. “Guido Imbens’ work was focused on making research as convincing as possible so that you could obtain credible evidence as to whether a certain intervention was working or not”, explains Dinand Webbink, professor of Policy Evaluation at Erasmus School of Economics. He uses Imbens’ methods to determine the causal effect of education policy, such as what impact higher salaries for teachers has on children’s learning achievements.
Credibility Revolution
Webbink is not at all surprised that Imbens has been awarded the Nobel Prize. “Guido Imbens and his colleague Joshua Angrist have set the standard for how we now conduct empirical research in economics.” Imbens is viewed as a key figure in what is known as the ‘Credibility Revolution’ in the field of Econometrics. The main reason he won the Nobel Prize is due to his studies on causal effects in economics research. He asked himself: how can you demonstrate whether a specific intervention such as new policy is having its intended effect? Or, looking at it another way: how can you identify causal relationships and what is the most convincing way to do this?
Imbens received acclaim for the paper he published together with Angrist in 1994 in the journal Econometrica. “In this paper, they show how experiments using observational data can still be used to credibly identify cause and effect relationships”, says Webbink. Imbens encountered the following problem in ‘old school econometrics’: existing methods could be used to obtain various estimates of an effect, but it was not clear which estimate was actually caused by the intervention.
However, this problem can be resolved by using experiments, especially natural experiments. For example, income limits are commonly used when granting subsidies. By comparing the findings for people close to the income limit, it is possible to identify the effects of receiving a subsidy, just like in a randomised experiment with a treatment group and a control group. In the same way, he compared lottery participants who had won the lottery with other participants who had not won in order to work out what effect the extra money had on the lives of the winners.
Broad application
“By using phenomena that occur in natural experiments, Imbens and Angrist were able to devise methods that made research in the economic sciences more transparent”, explains Webbink. “That’s how they laid the foundation for the current standard of research that uses empirical data. It makes it easier to review economics research that focuses on finding causal relationships and makes it much less susceptible to fraud.”
According to Webbink, the greatest achievement of Imbens’ work is making research more convincing. “For example, in economics we want to know what effect policy has, but Imbens’ methods can be used in any area of science where researchers are looking for causal relationships.”
EUR alumnus
Webbing believes that winning the Nobel Prize is more than just the crowning achievement of Imbens’ hard work. It also confirms that his methods are now the standard for empirical research. He is especially proud of the fact that Imbens studied in his programme at EUR. “This is where it all began and he received a great education here. And now his methods are being taught in the Economics and Econometrics programmes. I should also mention that he is a remarkably nice man. A long time ago, I once defeated him in a game of chess at the Dutch Youth Chess Championships. Whenever I run into him nowadays he never fails to mention that I’m a better chess player.”
