Polarization in increasingly connected societies

Area of Science:

• Social dynamics
• Computational social science
• Network science

Background:

• Traditional models of societal polarization focus on homophily, bounded confidence, and community interactions.
• These models often attribute polarization to a lack of interactions, which may be insufficient in highly connected societies.
• Existing explanations may not fully capture polarization dynamics in the context of increasing societal interconnectedness.

Purpose of the Study:

• To propose and analyze a novel voter-type model (I-voter) that incorporates individual involvement in opinion formation.
• To investigate how societal polarization is influenced by network connectivity.
• To understand the mechanisms driving polarization in increasingly connected social networks.

Main Methods:

• Developed an agent-based voter-type model (I-voter) incorporating an 'involvement' parameter.
• Analyzed the model's steady-state behavior using analytical techniques.
• Employed mean-field and moment-hierarchy approaches for analysis.
• Examined model behavior across various network topologies and extensions.

Main Results:

• The I-voter model demonstrates that increased network connectivity can exacerbate societal polarization.
• Involvement emerges as a critical factor influencing opinion dynamics and polarization.
• Analytical results provide insights into the steady-state conditions of the model.
• Findings are robust across different network structures and model variations.

Conclusions:

• Societal polarization is not solely driven by lack of interaction but can be amplified by increased connectivity when involvement is considered.
• The I-voter model offers a new framework for understanding opinion dynamics and polarization in complex networks.
• The study highlights the importance of individual involvement in shaping collective opinion and societal divisions.