Mathematical Sciences Department
Colloquium
Speaker: Vicky Chuqiao Yang, MIT
Friday, February 3, 2023
11:00 am - 12:00 pm
Stratton Hall 203
Title: Dynamical system models for voting and collective decisions
Abstract: I will give an overview of two projects using quantitative behavioral models to study voting and collective-decision making, leveraging dynamical-system models. The first project proposes a mechanism for the polarization of the US Congress despite the lack of clear evidence for voters taking more extremist policy positions. The model considers voters using the bounded-rational, "satisficing" decision-making---one chooses a candidate who is "good enough" instead of finding the best option. The model suggests that party polarization can be an outcome of increasing ideological purity within the parties without changes in the voters' policy positions. We also compare models' predictions with historical data. The second project addresses whether a collective can arrive at the better of two options when some voters learn from others (social learning) instead of evaluating the options on their own (individual learning). The mathematical model predicts a critical proportion of social learners above which either option can win the majority regardless of merit.