Homophily describes the tendency of individual actors to associate with others who share their social characteristics like religion, culture, political orientation, etc. Homophily may lead to the segmentation of society into ideologically homogeneous filter bubbles with relatively low internal conflict. On the other hand, homophily can also increase the level of inter-group conflict. The present paper analyzes both aspects of homophily by means of virtual encounter simulations. This is a microsimulation method that randomly matches dyads of persons and compares their opinions and attitudes, which were collected in standardized mass-interviews. By aggregating the value conflicts between such randomly matched pairs it is possible to determine and compare virtual conflicts at the group level. This allows to rank groups with regard to their homophily.

As an illustration of the mentioned method the author uses the Swiss data of the National Identities II module of the International Social Survey Programme, which contains information about desirable criteria of national citizenship. The virtual encounter simulations reveal that the German and Italian speaking citizens and the left and right political partisans are relatively homophilic with regard to the social definition of "Swissness". To the contrary, the French speakers and the different religious communities are substantially less homophilic with regard to Swiss national identity. Not so unexpectedly, the different criteria of "Swissness" considerably vary with regard to their importance for the homophily of these groups.

The present paper uses agent-based simulation to compare how rational and emotional decision architectures function in interethnic trust. Our main research question asks how emotional decision architectures explain additional variance in emerging social cohesion in the context of changes in the ethnic heterogeneity of a neighborhood.

Radicalization is often explained through isolated mechanisms, limiting a comprehensive understanding of how extremist behavior emerges. This paper builds on the Individual, Group, and Ideology (IGI) model and proposes an integrated agent-based model to examine how moral value satisfaction and intergroup dynamics jointly shape radicalization through the emergence of ingroup superiority, defined as the alignment of ingroup favoritism and outgroup derogation.

Simulation results show that these mechanisms can develop independently or align to produce reinforcing feedback loops, generating distinct trajectories of radicalization. The order in which ingroup favoritism and outgroup derogation emerge determines whether groups radicalize or not.

These findings demonstrate that radicalization depends not only on the presence of key mechanisms, but on how and when they interact, highlighting the importance of sequencing for understanding variation in radicalization processes and potential de-escalation.