Emergent Complex Socio-economic Phenomena in Urban Environments. Hierarchies, Segregation and Cooperation

Urban environments are complex socio-economic systems in which inequalities, segregation, and cooperation emerge from the interplay between individual decisions, group dynamics, and institutional structures. This thesis explores how computational and data-driven tools from complexity science can be used to explain, simulate, and quantify such emergent large-scale social patterns. Grounded in a perspective specific to statistical physics, it integrates agent-based modelling, stochastic processes, network science, and game theory to investigate the micro-motives behind urban inequality, segregation, and collective action. The first part of the thesis models the spontaneous emergence of social hierarchies through agent-based simulations. It shows that even symmetric groups can evolve into highly unequal status distributions through self-reinforcing local interactions. The system exhibits identifiable phase transitions and scaling laws, independent of system size, highlighting how dominance structures can emerge endogenously through inter-group dynamics in competitive societies. The second part reconstructs co-residence network using high-resolution administrative registration data from the city of Vienna, covering local and migrant populations. The analysis reveals socio-spatial segregation driven by neighbourhood wealth, neighbourhood diversity, and nationality-based homophily. This network-based clustering uncovers latent group structures and offers a generalisable and novel method for mapping urban socio-economic segregation patterns in multicultural urban contexts beyond traditional spatial metrics. The last part draws on a citizen science experiment conducted in Barcelona, where participants engaged in a collective-risk dilemma for air pollution mitigation action under unequal initial resources to reproduce real-world socio-economic heterogeneity. The study analyses decision-making behaviour and finds that less endowed individuals contribute proportionally more to the public good, highlighting asymmetries in perceived responsibility and capacity. By combining real behavioural data with agent-based simulations led by conditional probabilities, the experiment challenges assumptions in classical models of cooperation and underscores the role of inequality in shaping strategic behaviour. Together, these case studies show how tools from complexity science can advance both theoretical and applied understanding of urban social dynamics. By bridging abstract modelling, empirical data, and behavioural insights, the thesis contributes to the study of statistical physics applied to modern cities, offering a deeper understanding of the structural mechanisms driving inequality, segregation, and cooperation. It proposes a multi-method framework for diagnosing feedback loops in urban complex systems and guiding policy interventions toward more inclusive and equitable urban environments, combining quantitative rigour with societal relevance. Agent-based modelling, network science, and game theory share key foundational principles with statistical physics, particularly the emphasis on emergence, collective dynamics, and bottom-up interactions. In this thesis, these frameworks are not treated as subsets of physics but as complementary perspectives aligned by a shared epistemological philosophy: that complex social phenomena arise from the interactions of many simple units. These approaches highlight the potential of physics-inspired models to support the co-production of knowledge among researchers, institutions, and citizens.