The Volkswagen Beetle is a best seller of the brand. This well-known car is easily recognizable even by those who are not car enthusiasts. But do you know the year in which Volkswagen launched the Beetle's first version (back then called the Volkswagen Type 1)? Take a few seconds to pick your brain and write it down. Now, take a few extra seconds and estimate how many years away your answer is from the correct answer to the question. In other words, how wrong do you think your answer is. Write it down again.
The first version of the Beetle was introduced in nineteen thirty-eight. If you did not get the answer right, do not worry; 88% of the people we surveyed recently do not know the answer either. However, what about your answer to the second question? How good was your estimation of how wrong your answer was? Were you overconfident about how well you could guess the date we asked you about, or, on the contrary, did you think that your error would be larger than it was? Again, do not worry; most people we surveyed thought their guess was more accurate than it really was.
In a recent study (Bosch-Rosa, Kassner, and Ahrens, 2021), we exploit this two-questions procedure to study how overconfidence in the accuracy of one's own information (also known as overprecision) impacts behavior. We do so by asking participants of the Innovation Sample of the German Socio-Economic Panel (SOEP-IS) to guess the year in which different events took place and, next, to estimate how wrong they think their answers are. By comparing the accuracy of participants' answers to their error estimation, we can establish how "overprecise" a person is.
Exploiting the large SOEP-IS dataset, we show that overprecision strongly correlates with people's financial and political behavior. Those who estimate the errors in their guesses to be smaller than they are (that is, those who are overprecise), make worse predictions about future stock prices, worse predictions about the future rental prices of houses, and inferior investment decisions. Furthermore, those who are overprecise tend to hold more extreme political views and are less likely to vote. However, overprecision does not determine whether a person is more left or right-leaning in her political views.
Our research highlights the importance of personal traits in predicting people's behavior as it shows that overconfidence in one's knowledge results in infelicitous behaviors across different dimensions of life. After all, it takes a very wise person to know that they know nothing.
Ciril Bosch-Rosa (TU)
Bernhard Kassner (Ludwig-Maximilians-Universität, München)
Steffen Ahrens (Freie Universität, Berlin)
References: