The cultural impact on social cohesion: an agent-based modeling approach

Social processes in modern multicultural societies require a better conceptual understanding of the mechanisms of cultural events’ impact on social welfare. Due to a number of objective reasons, one of the critical challenges in this complex research area is the lack of empirically based predictive models. The current paper provides an alternative approach—a bottom-up (from agents to social systems) modeling how cultural events can shape social cohesion measured by social capital and cultural features probabilistic clustering in the population. In this paper, based on prior empirical observations, proposed agent-based modeling can help (i) understand and interpret some empirical findings, and (ii) foresee outcomes of otherwise very costly real-life social experiments. To this end, this paper presents an agent-based simulation model to demonstrate the simple mechanism of how cultural events can impact the empirically observed complex dynamics of social capital. Presented model is implemented in the NetLogo simulation environment, where simple agents’ behavioral properties are simulated following basic empirical observations. Implemented simulation approach upgrades Axelrod's classical model of cultural dissemination in three main ways. First, it models agents' neighborhood interactions not only in the simulated agents' physical space but in the cultural features space as well. Second, the model simulates the dissemination of cultural events’ impact (not only) through pair-based neighborhood interaction but also through wide-range social media and networks broadcasting. Third, it implements some agents’ inherent propensity toward differentiation (uniqueness) that generates divergence in the virtual space of characteristic behavioral cultural features. Simulation results provide not only proof of concept but also reveal underlying cultural conditions for the emergence of different behavioral patterns of social capital cohesion or fragmentation.