Should social media platforms differentiate their content moderation policies for minority communities in the presence of competition?

Social networks are intricate ecosystems comprising diverse communities with distinct laws, customs, and preferences. Social media platforms sometimes set content moderation policies tailored to individual communities but other times impose a uniform policy across multiple communities, leaving smaller communities dissatisfied. In this article, we propose a game-theoretical model that considers platform competition for users with diverse preferences within a complex network. Users allocate their time to platforms, while platforms optimize their moderation policies to maximize user engagement. Our result provides the methodology of optimizing the moderation policies. The aggregated user preference and the user engagement are determined by the community imbalance indicator, which encapsulates the network structure and user affiliations. We conducted simulations for a social network with over 60,000 users across seven country-based communities to identify the aggregated optimal moderation policy and assess the impact of differentiating the policy for minority countries. This research contributes to a deeper understanding of the relationship between moderation policies and platform competition.