Doctoral researcher Özgür Gürerk  first encountered experimental economics--a young field anchored in the close observation of human behaviour--as a subject in an experiment himself. Today he is formulating hypotheses, designing experiments, and publishing high profile work.
Gürerk and his supervisor Bettina Rockenbach , based at the University of Erfurt in Germany, are broadly interested in why--and under what conditions--individuals decide to cooperate in communities even when they have to sacrifice their own self-interest. Trained as economists, Gürerk and Rockenbach both say that to engage successfully in the field you have to embrace a highly interdisciplinary, open-minded, and cooperative approach. Before entering the experimental lab, however, they say they must invest lots of time into planning their research, asking the right questions, and pre-testing their hypotheses. The work may be demanding, they say, but it leads to career opportunities outside as well as inside the academic ivory tower.
Learning the ropes
Gürerk started studying economics at the University of Bonn in 1995. But it wasn’t until 3 years later he got his first experience of experimental research, when he joined Reinhard Selten's experimental economics lab, as a student assistant. It was a particularly exciting time; a few years previously Selten, along with John Harsanyi and John Nash, won the Nobel Prize in Economics  for pioneering work in game theory. In 2001, Gürerk undertook his own research project with Selten, for his diploma thesis (roughly equivalent to a masters), studying the effect of information provision on an oligopoly market. Gürerk summarises the main question in his research at this time: Within an oligopoly "if one group has more information [than another] on the structure of the market, what effect does this have on collusion?" Gürerk says that after years of reading and studying theory, he was happy to be an experimentalist: "It was fascinating having the possibility to test economics theories."
After finishing his diploma thesis in 2001, Gürerk decided to move with Rockenbach, also based in Bonn at the time, to the University of Erfurt where Rockenbach had been offered a professorship. In Erfurt, Gürerk became a research associate--a full salaried position--and is presently working on his doctoral thesis. He divides his time among doing his own research, taking care of the laboratory, and doing some teaching.
Finding a research niche
Gürerk says his greatest career challenge to date was searching for a solid topic for his doctorate. He says he first considered going further with industrial oligopoly, but after investigating the literature, he felt "there wasn't too much more to explore." Together with Rockenbach, Gürerk turned his attention to the mechanism of cooperation and punishment, specifically how punishment might encourage people to cooperate. "I thought it was a hot topic," he recalls.
Still, it took about 6 months to fully investigate and plan his doctoral project. "The hardest part is finding a good idea, a good research question," he says. Economist Ernst Fehr of the University of Zurich agrees and also believes having the skill to pin down what can be answered experimentally is just as important: "an instinct for the important unsolved questions and tackling those that can be successfully tackled," says Fehr.
The next challenge, says Gürerk, is creating--or finding--a mathematical model to pretest hypotheses electronically and make predictions about experimental outcomes. Before you enter the experimental setting, according to Gürerk, you need to have clear predictions. Rockenbach agrees. "You need to find a design where you can make clear predictions: what are the causes, what are not the causes. And what's doable," she says.
Having nailed down his plan, Gürerk canvassed university students to sign up for experiments. "I go out to the campus and advertise," he says. Subjects receive a nominal payment but much more important, says Gürerk, is that the subject understands the rules and behaves as if they were in a real life situation.
Into the lab
Using an experimental approach, Gürerk--under the supervision of Rockenbach--has investigated related issues: Under what conditions individuals decide to cooperate, whether they are likely to sacrifice their self-interest for the sake of this cooperation, and whether they choose a community that punishes (sanctioning institution) or one that does not punish (sanctioning free institution) non cooperators. In their study, cooperation was measured as the amount of money an individual contributed to a "public good" fund, one designed so that the entire group--even individuals who had not contributed--profited from the returns. Noncontributors were termed "free-riders." Members could impose a "punishment" on free riders, but only at a cost to themselves. Punishment of free-riders involved deducting a fixed sum from their endowment; the punishers also had to pay a sum (albeit a smaller sum than a punishment) if they chose to punish. After the experiment subjects were paid according to the success of their financial returns in the experiment.
Cooperation and punishment
Gürerk's and Rockenbach's data--which was published together with Bernd Irlenbusch of the London School of Economics in Science this year--showed that initially the majority of individuals chose to join a fund with no punishments, but halfway through the experiment, the vast majority had switched to a community that imposed sanctions. Why? Because the participants in the sanction-free institution--even free riders--realised that punishment exerts an incentive to cooperate, and the overall profitability of the sanctioning community was much higher to their own.
"This paper is a fascinating study of the spontaneous emergence of social order through migration to groups with sanctioning opportunities," says Fehr. Researchers in fields from philosophy to evolutionary biology "have written--speculated--much about the problem of social order. But why and how humans have been able to escape a state of anarchy is still largely unknown," says Fehr. This work, he says is "an important step forward."
Joseph Heinrich, of Emory University in Atlanta, agrees it is an important finding. He also acknowledges the study was not trivial to set up. "Allowing different groups possibilities to simultaneously interact is hard to set up [experimentally]. And the results," continues Heinrich "are clean."
Rockenbach says this work falls into the broader context of understanding "how can we enhance the provision of cooperation in our society?" Although Rockenbach stresses their recent findings are of academic interest, understanding what motivates people to cooperate for the sake of the "public good" is important in many societal contexts. One example is pollution. "Clean air is a public good," says Rockenbach, but individual sacrifices, and perhaps sanctions, may be needed to achieve it.
Skills and rewards
Apart from the philosophical questions their work raises, Rockenbach and Gürerk say there are plenty of professional challenges and rewards involved. Working as an experimentalist since 1988, Rockenbach is no stranger to understanding a new research territory. As a doctoral student of Reinhard Selten at Bonn, she combined rational game theory, which is firmly anchored in mathematics, with behavourial game theory, which is routed in psychology. Rockenbach says at the time "it was unusual to combine the two." It meant she invested a lot of time learning psychology from scratch. "I went right back to the orginial literature of the 1950s," she says.
Rockenbach notes that the work is highly interdisciplinary in nature, so researchers in the field find they must cooperate. Since her graduate student days, she has worked with biologists and anthropologists and has thoroughly enjoyed it. "You have to interact with a lot of people. It’s very enriching, but it's a struggle too," she says. Additionally, "You need to be open-minded to your own discipline, to be able to rethink things that were so clear to you." For example, a collaborator might ask her why she approaches an experiment in a certain way. If your answer is: "because we always did it that like that," Rockenbach says, it won't convince your collaborators. "You need to be willing to rethink a lot of things. But it's fascinating [suddenly] seeing different facets of your own discipline." That's not to say that they aren't frustrating moments too, when you "realise you have being using some word in a different context [to your collaborator], and you have doing this for 6 weeks," she says.
In addition to bridging the interdisciplinary divide, Rockenbach says, it is important to develop analytical skills. "It doesn't have to be maths," she says, "but you need to be able to make sharp distinctions between factors." And last but not least, "You have to enjoy observing and analysing social interactions."
Rockenbach's position is tenured, but both she and Gürerk intend to stay in academia. Gürerk's goal is to complete his doctorate by the end of this year and find a research position, but he knows that academic posts in Germany are rare. "An academic career is risky," he says. But he feels his training would also be useful outside the ivory tower. "With research training you think strategically. You can link an action to an outcome." These skills, Gürerk says, are relevant to sectors such as consultancy and finance.
But it’s back in the experimental lab where Gürerk is happiest. His ultimate reward is seeing the outcome of his experiments live. After carefully experimental planning and predicting, he wonders, "what are the subjects going to do?"
"When you see that they follow your prediction, your own hypothesis, it's a special moment," he says.
Anne Forde is European editor North and East for Science's Next Wave.
Comments, suggestions? Please send your feedback to our editor .