The Science of the Game

The economist Martin Dufwenberg

Dieser Artikel in Deutsch

Martin Dufwenberg

Anyone who decides to attend a Monopoly party at the Dufwenbergs should not expect that players will follow the usual rules of the game. Instead of two dice, players may use just one, or even three, when it is their turn at play. These alterations open up possibilities that are not found in the instruction booklet. As if that were not enough, nine year-old Martin has been crowned Monopoly King. More than one friend and scholar has walked out of the Bonn apartment in frustration. Yet Martin senior remains unfazed and continues to tamper with the rules of the game-after all, he is a professional. Professor Martin Dufwenberg is a renowned researcher of game theory, which is highly relevant not just in economics, but beyond.

"Using game theory and experimental methods, we create mathematical tools for economic research," explains the 38 year-old native of Sweden. At the same time, he emphasizes that these mathematical tools can be applied in many other areas as well, like sociology, psychology, or political science. Game theory deals with situations in which a number of people make decisions and thereby influence one another. This is the case with price wars in the business world, auctions, and even marriages. "I do not see myself first and foremost as an economist," says the passionate bridge, tennis, and golf player. "I was always most fascinated by strategic interaction: how individuals behave in situations where they make decisions, which strategies they use, what predictions can be made about their behavior, and finally, how they assess one another."

Today Martin Dufwenberg is conducting research at the Laboratory for Experimental Economic Research at the University of Bonn. He is the recipient of the Alexander von Humboldt Foundation's Friedrich Wilhelm Bessel Research Award. The award allows outstanding young scholars from around the world to pursue projects of their own design in Germany. Martin Dufwenberg's host is the 1994 winner of the Nobel Prize in Economics, Reinhard Selten. "He is one of my heroes," admits the scholar from southern Sweden. "It is a great adventure, to be working here with him."

By the time Martin Dufwenberg discovered his "heroes," he was almost finished with his doctoral degree. After completing his matriculation examination, he studied for a time in Alabama (United States). As a result of that stay abroad, he decided to study economics, rather than mathematics or linguistics. But that was not quite right for him. "I was lacking enthusiasm. Without it, you cannot do really good work." It was only during a course in game theory at the University of Uppsala that it dawned on him, "This is it." The topic for his dissertation, in the area of taxation, was thrown overboard, and Martin Dufwenberg jumped headfirst into his new field. He went on to spend an inspiring year at Princeton University. Later he spent two years as a post-doc at the Center for Economic Research in Tillburg (Netherlands). Shortly before he headed to the German Rhine River, he was made a tenured professor at the University of Stockholm.

The father of two travels often-the long list of conferences that he will attend speaks to his energy and level of engagement. "The most important thing is to talk and to exchange ideas with colleagues who have similar interests. That makes for good science. Ingenious ideas can come about unexpectedly, over a cup of coffee with colleagues between conference sessions." The fact that he values collaboration is reflected in the many publications he has written with co-authors.

Martin Dufwenberg

The economist has a variety of research interests. They range from pure theory all the way to practical experimentation. This is how Martin Dufwenberg develops fundamental tools for game theory. With the help of mathematical methods, he determines optimal, or predominate, game strategies, which, for example, can allow one to make better predictions about the outcome of the "game."

In other projects, the Humboldtian allows sociological and psychological aspects to flow into his theoretical analyses. "A large part of my work involves integrating these perspectives into the usual economic models." These are factors that cannot be disregarded, for example, in the determination of salaries, where there is always a connection between esteem and motivation. Traditional economic models barely consider the variety of factors affecting human behavior. Typically they work under the assumption that in economic processes, such as price wars, people act rationally and according to their own best interests.

"Motivation and emotions are not take into consideration." Martin Dufwenberg starts from there and tests his ideas with experimental research. "Experiments often show that traditional models, applied to the selfish human, are insufficient." For example, the image that individuals have of one another plays an important role in their decision-making. People are nicer to people whom they estimate to be nice and when they think this could be to their advantage, rather than to people who seem negative. Dufwenberg is trying to capture this concept of "reciprocity" in a mathematical model. "It is not enough to observe what people do. Instead, we have to recognize the intentions behind their behavior in, for example, economic transactions."

Together with Reinhard Selten and his Spanish colleague José Apesteguia, Dufwenberg is testing the possible effects of new anti-cartel laws. They were recently introduced in the United States and in the European Union, and grant immunity to those who break out of cartels and inform the authorities about the illegal price-fixing. "These laws change the rules of the market. We want to see how they influence behavior," explains the researcher. In an experiment at the laboratory in Bonn, students have to announce prices in a common market. Prior to that, three test subjects have the opportunity to correspond online for ten minutes. Do they try to form a cartel? Will one of them go to the authorities? Dufwenberg thinks it entirely possible that the new laws could have the opposite effect, namely a stabilization of cartels. That could be the case when members of the cartels threaten to report each other if they feel they have been treated in an underhanded manner.

Will the knowledge gained from this cartel experiment or other investigations find entrée into the household Monopoly game? No one would be surprised if it did.

Uschi Heidel (translated by Susanne Wunner)

28.05.2003