An Agent-Based Statistical Physics Model for Political Polarization: A Monte Carlo Study

Area of Science:

• Computational Social Science
• Statistical Physics Modeling
• Political Science

Background:

• Global political polarization poses a significant threat to collective decision-making.
• Previous research utilized mean-field models for polarization dynamics, assuming universal interactions.
• A more realistic approach is needed to account for localized interactions in social systems.

Purpose of the Study:

• To extend previous polarization modeling by incorporating short-range interactions.
• To simulate political polarization trends in the USA across three groups: Democrats, Republicans, and Independents.
• To evaluate the effectiveness of agent-based Monte Carlo simulations for understanding polarization.

Main Methods:

• Agent-based Monte Carlo simulations were employed to model interactions between individuals.
• The study focused on short-range interactions, simulating a more realistic social network.
• A novel polarization index was used to quantify the degree of political division.

Main Results:

• Simulation results from agent-based models closely mirrored those from mean-field models.
• The study generated plausible scenarios for polarization trends over time in the USA.
• The proposed polarization index effectively measured societal divisions.

Conclusions:

• Short-range interaction models provide comparable insights to mean-field models in polarization studies.
• The agent-based approach offers a flexible framework for analyzing political dynamics.
• The methodology can be adapted to study polarization in diverse political systems.