English  Sprachen Icon  |  Gebärdensprache  |  Leichte Sprache  |  Kontakt


I lie? We lie! Why group are more likely to be dishonest than individuals

von Lisa Spantig, IDK Evidence-Based Economics

Unethical behavior such as dishonesty, cheating and corruption occurs frequently in organizations or groups. Recent experimental evidence suggests that there is a stronger inclination to behave immorally within groups than individually. In a new study we ask whether this is the case, and if so, why. To this end, we use an experimental setting to study how individual behavior changes when deciding as a group member.

In recent years, several prominent examples of unethical behavior in groups and organizations have come to light. For instance, fraudulent accounting methods and malpractice of groups of executive officers have led to the marked bankruptcies of WorldCom and Enron. More recently, it has been discovered that inter alia the German car producer Volkswagen has sold diesel cars with emissions certificates based on potentially faulty information. In this new study which is part of my PhD work in the evidence-based economics program, I investigate together with Martin Kocher (IHS Vienna and University of Vienna) and Simeon Schudy (LMU Munich) the circumstances that facilitate such dishonesty. In particular, we wanted to know whether disregard for moral norms is a matter of individual choice or rather a product of structural factors within groups.

We conducted laboratory experiments with 273 participants, using an innovative computerized version of the so-called "die-roll-paradigm", a simple reporting task in which they trade off honesty and real monetary payments: participants see a video of a die roll, are asked to report this number, but are paid according to the number they report, not the number they see. The higher the reported number, the higher would be the payment. Even though the task is very simple and abstract, other studies have shown that comparable versions of it are related to dishonest behavior in real life, such as corruption of public workers or business men cheating on customers.

In the study, groups were formed randomly and were allowed to chat via the computer for five minutes before every participant enters her report privately. This opportunity to communicate is the only difference between groups and individuals in the experiment. Participants receive detailed instructions on the decision-making situation, but remain anonymous in the chat as anonymity helps to exclude possible uncontrollable factors (like personal ties or gender) from influencing the results.

Prior research on dishonesty has shown that dishonesty tends to be more prevalent in groups than among individuals, but it remained unclear why this is the case. It has been shown that groups behave more rationally than individuals and thereby may lie more when lying is profitable. Further, deciding in groups allows individuals to disguise dishonest acts. Finally, lying in groups often refers to a situation in which immoral individual behavior benefits other group members. We provide novel evidence that even in the absence of these group features, individuals are about 50 percent more likely to behave dishonestly, when they communicate in a group before deciding on a (dis)honest action. The size of this "dishonesty shift" does not depend on whether other group members benefit from lying or not. Implementing additional experimental conditions, we can also rule out that deliberation alone explains the dishonesty shift.

We show that communication within groups or organizations is likely to explain why unethical behavior such as lying, dishonesty and corruption prevails in many real world organizations: it allows group members to adjust their own moral standards. The exchange of arguments for and against dishonest behavior within a group makes group members more dishonest than individuals and also more pessimistic about other people's honesty. Since naturally occurring groups are usually allowed and often even encouraged to communicate, institutions might want to take into account the downside of communication. The study paves the path for future research that could investigate mitigating effects of codes of conduct, exogenous monitoring (and punishment) and their interactions with communication in the short run and, in particular, the long run.



Lisa Spantig, IDK Evidence-Based Economics

veröffentlicht am