August 31, 2006

Source: :
http://www.mta.ca/news/?id=1086

Overconfidence — too much of a good thing or not?

SACKVILLE, NB — Could that arrogant guy in your office, with his overconfident demeanour, be on to something? According to a Mount Allison University honours economics student who spent her summer studying the benefits of overconfidence, maybe so.

Kirsten Cornelson has carried out her own research program focusing on the "overconfidence phenomenon." She explains, "People have a tendency to overestimate their chances of success, especially when success or failure depends on their own talents and abilities. I am attempting to explain why overconfidence might be beneficial when considered in conjunction with several other biases that systematically influence the way people make decisions. I wanted to examine some of the research done on overconfidence and see, from an evolutionary perspective, whether overconfident behaviour might generate some advantage that could be acted upon by natural selection." The research is funded by the Summer Research Program at Mount Allison University, and is conducted under the supervision of Neils Anthonisen, a faculty member in the department of economics.

The Toronto native came up with the idea for her research in a game theory class. Through her research, Kirsten found that overconfidence characteristics vary in different cultures and between the genders. Western cultures and males tend to display this characteristic more, but Kirsten explains that these findings might not be as clear cut as they seem. "There are different values in Eastern and Western cultures. Thinking well of ourselves is universal; it’s how we illustrate these thoughts and explain events that happen to us (ie: if something bad happens, it was the circumstances, if something good happens, it was the individual’s doing) that allows us to see the differences in perception."

Dr. Anthonisen says, "Kirsten has spent her summer thinking and reading both widely and deeply, and, as a consequence, has generated several interesting ideas about why people behave the way they do. Economists have long been aware that their predictions about the behaviour of economic agents often do not square with experimental results generated by experimental game theorists and psychologists. This has led to an exciting new body of research where tools from economics and game theory are used to explore phenomena that have traditionally been confined to the realm of psychology. Kirsten’s research makes a nice contribution to this new trend in economic thinking."

Several economists and psychologists have pursued research along these lines and have created formal and informal models of rational overconfidence. These models show that overconfidence can be beneficial in specific types of situations. In order to account for the importance of self-esteem in mental well-being, a model of overconfidence needs to predict this kind of behaviour across a much wider range of circumstances. Kirsten is attempting to build a more general, comprehensive model that accommodates a broader range of applications.

Kirsten’s model relies on research involving "learned helplessness," a phenomenon that occurs when someone experiences an uncontrollable event and subsequently stops exerting effort in other situations. This is problematic because it frequently extends to situations that are, in fact, within that person’s control. Kirsten says, "I believe that confidence may operate in a similar fashion; if we receive one negative piece of information about ourselves, we interpret it as being reflective of our general abilities or worth. This means that in the long term, we may be better off avoiding all negative information about ourselves, even if that information could help us in specific circumstances."

A compelling model of overconfidence could help to understand certain anomalies that are currently outside the scope of economic theory. For example, research Kirsten has found shows that overconfidence has been used to explain the high rate of failure among new businesses; entrepreneurs overestimate their chances of success, leading them to make imprudent decisions. Research studies also indicated that overconfidence in world leaders has been linked to more aggressive foreign policy. By helping to expose the root causes of overconfidence, such a model can help to correct this kind of behaviour.

Kirsten is going into her fourth-year at Mount Allison and plans to continue her studies at the master’s level. She hopes to eventually work in either economic policy or international development.

—30—