Understanding the effect of individual decision-making strategies on collective decision outcomes

Collective decisions are central to societies, from committees to governments. An important open question is how collective decisions are affected by individuals who follow the choices of others (social learners) instead of evaluating the merit of options on their own (individual learners). The three central questions guiding this project are: 1) Will a collective choose the best available option when many of its members are social learners? 2) How does the collective outcome depend on the properties of individuals and the options? and 3) How will social learners affect society’s ability to adjust to new evidence? This work expands the current understanding of collective decisions by examining a variety of factors simultaneously rather than in isolation. It helps anticipate and manage instabilities in important real-world collective decision-making systems such as democratic elections, jury decisions, and teamwork in companies. Moreover, the most pressing challenges facing humanity, such as climate change, technological disruption, and pandemics, are critically subject to global collective decisions. The project investigates the effect of social learning on collective decision-making using a transdisciplinary combination of three methods. First, the program of research develops a dynamical-system model that studies how the interactions of social and non-social factors affect collective decisions and adjustments to new information. A preliminary model predicted a critical transition in opinion composition as the proportion of the social learners crossed a threshold. The current project expands this framework to include the presence of committed minorities, the introduction of new evidence about the merit of options, and the effect of network structures. The model derives conditions under which groups can settle on the option with the most merit. Second, the project tests the predictions of the model by collecting and analyzing observational and survey data and estimates the effect of social learning on beliefs and behaviors of real-world societies. Third, group-level and individual-level human participant experiments test model predictions in controlled conditions. The individual-level experiments uncover factors that influence individuals to adopt social or individual learning. This research informs a long-lasting debate on how individual learning strategies (individual or social) affect collective decisions. It combines and synthesizes social psychological principles and experiments, mathematical modeling methods, and statistical analysis. Results inform a wide range of audiences with the purpose of providing key insights for improving collective decisions in many real-world domains. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.