Is the voter model a model for voters?

Area of Science:

• Complex Systems
• Computational Social Science
• Statistical Physics

Background:

• The voter model is a standard for opinion dynamics but lacks real-world applicability.
• Existing models do not fully account for social context or agent mobility.

Purpose of the Study:

• To develop a more realistic opinion dynamics model by including social influence and agent mobility.
• To assess the model's ability to replicate empirical data from U.S. presidential elections.

Main Methods:

• Introduced a noisy voter model with recurrent agent mobility.
• Incorporated spatial and population diversity as model inputs.
• Analyzed the model's dynamics as a noisy diffusive process with anisotropic coupling.

Main Results:

• The model successfully captures stationary vote-share fluctuations across U.S. counties.
• It reproduces long-range spatial correlations decaying logarithmically with distance.
• Model behavior is consistent across different geographical scales (counties, districts, states).

Conclusions:

• The noisy voter model with mobility provides a robust framework for understanding opinion dynamics.
• Mobility range and decision-making randomness are key factors consistent with empirical observations.
• The model offers insights into the statistical features of political opinion spread.