Beliefs as a Means of Self-Control? Evidence from a Dynamic Student Survey

By Tobias König, Sebastian Schweighofer-Kodritsch and Georg Weizsäcker - Version in Deutsch

The development of a behavioral model capturing how individuals put excessive weight on pleasure and pain in the present, relative to the future, is one of the most influential contributions of behavioral economics. This excessive weight on the present leads to a self-control problem. A typical example of this is studying for an exam, which imposes costs now for benefits later. Here, overvaluing the present means studying too little.

Numerous studies show that people’s beliefs are biased toward wishful thinking. Could it be, then, that people form systematically biased beliefs in order to overcome self-control problems, such as that of studying? In a recent study, BCCP Fellows Tobias König, Sebastian Schweighofer-Kodritsch, and Georg Weizsäcker present the first field evidence that this is indeed the case.

The three authors regularly surveyed about 120 first-semester students throughout the semester, asking about their study plans and expectations regarding one particular class. The authors assumed that students might motivate themselves to study more by subjectively exaggerating the returns of this effort. In the hot phase close to the exam, studying and, hence, self-control are particularly important, such that the students might then be particularly optimistic about the returns to studying. Upon having completed the exam, they might become realistic because the motive for distorting one’s beliefs is gone. This yields the testable hypothesis that students believe that the returns to studying are higher before the exam than afterwards.

Testing this hypothesis faces the challenge that students’ beliefs may change in various ways as they consume information that the researchers do not observe (for example, as a result of seeing a TV discussion about the German higher education system). However, the fact that respondents wrote their exams  at two different times (separated by almost 8 weeks) allows the simultaneous comparison of the beliefs of one group, which is about to write their exam, with those of another group, the "control group" whose exam is still in the future or has already passed. This controls for information that systematically affects all beliefs, whereby the crucial difference between groups remains valid: whether or not a respondent is close to an exam or not. In this comparison, the authors’ hypothesis is confirmed. Relative to the control group, the expected returns to studying of those students who are about to write their exam are significantly higher than those in the control group, with expectations dropping significantly after the exam.

This result suggests that biased beliefs need not be irrational, per se, but may have a material (economic-instrumental) value. The next important step is to more deeply understand the psychology of such biased belief formation in order to optimally design information provision that best supports people in motivating themselves and achieving their goals.

The full paper is available as WZB Discussion Paper SP II 2019-2014 (open access pdf download).