Hierarchies and decision-making in groups: experimental evidence

While individual decision-making is a principal research area in theoretical and experimental economics, many decisions are actually made by groups, and not by individuals. In addition, groups are often characterized by hierarchical organizational structures, but little is known about their possible effects on outcomes in economic experiments. This exploratory study investigates differences in decision-making outcomes for individuals and groups under different hierarchies using data from an incentivized economic experiment using a sample of South Korean college students and Bayesian hypothesis testing, frequentist regression analysis, and analysis of predicted probabilities for "ideal types". It compares the choices of groups of three in lottery choices and intellective tasks under five different hierarchy types: a group without hierarchy, a hierarchy by age, by merit, by chance, and by election. A within-subjects comparison of choices as individuals and as leaders allows to control for individual differences that might matter for choices beyond the hierarchy type. Regression results show that there are no differences in the number of safe choices between the different hierarchy types, but groups with a leader assigned based on merit perform better in intellective tasks, compared to leaders assigned based on a vote, age, or a random mechanism. These findings suggest that the type of hierarchy and its perceived legitimacy might matter for the quality of outcomes in intellective tasks, with no effects found for risk attitude.