Combining segregation and integration: Schelling model dynamics for heterogeneous population

The Schelling model is a simple agent based model that demonstrates how individuals’ relocation decisions generate residential segregation in cities. Agents belong to one of two groups and occupy cells of rectangular space. Agents react to the fraction of agents of their own group within the neighborhood around their cell. Agents stay put when this fraction is above a given tolerance threshold but seek a new location if the fraction is below the threshold. The model is well known for its tipping point behavior: an initial random (integrated) pattern remains integrated when the tolerance threshold is below 1/3 but becomes segregated when the tolerance threshold is above 1/3. In this paper, we demonstrate that the variety of the Schelling model steady patterns is richer than the segregation-integration dichotomy and contains patterns that consist of segregated patches for each of the two groups alongside patches where both groups are spatially integrated. We obtain such patterns by considering a general version of the model in which the mechanisms of agents’ interactions remain the same but the tolerance threshold varies between agents of both groups. We show that the model produces patterns of mixed integration and segregation when the tolerance threshold of most agents is either below the tipping point or above 2/3. In these cases, the mixed patterns are relatively insensitive to the model’s parameters.

💡 Research Summary

The paper extends the classic two‑group Schelling segregation model by allowing each agent to possess an individual tolerance threshold drawn from a prescribed distribution, thereby introducing heterogeneity that mirrors real‑world differences in preferences, income, culture, or openness. The authors retain the original relocation rule: an agent compares the proportion of same‑type neighbors within a defined neighbourhood (typically the Moore 8‑cell or a radius‑r set) to its personal tolerance τi; if the proportion falls below τi the agent moves to a randomly chosen vacant cell, otherwise it stays. Simulations are carried out on square lattices (e.g., 50×50, 100×100) with varying mean (μ) and variance (σ) of the tolerance distribution, as well as different neighbourhood radii, update schemes (sequential vs. synchronous), and initial random configurations.

The key finding is that the system no longer exhibits a simple binary outcome (integrated vs. segregated) determined by a universal tipping point at 1/3. Instead, two distinct regimes produce robust mixed patterns that combine segregated patches of each group with zones where the two groups are spatially integrated. When the majority of agents have low tolerance (μ < 1/3) but a minority possess high tolerance, the high‑tolerance agents cluster together, forming integrated islands, while low‑tolerance agents disperse into separate domains, yielding a “segregated‑integrated‑segregated” mosaic. Conversely, when most agents are highly tolerant (μ > 2/3) the opposite arrangement emerges: tolerant agents create large integrated blocks, and the few low‑tolerance agents occupy residual areas where they can tolerate a lower same‑type fraction, again producing co‑existing segregation and integration. Increasing σ (greater heterogeneity) accentuates these mixed configurations, but the qualitative structure remains stable across a wide range of model parameters, indicating that tolerance heterogeneity is the dominant driver of the observed dynamics.

The authors discuss the implications for urban policy: interventions that shift the distribution of tolerance—through education, housing incentives, or community‑building programs—could deliberately shape the spatial layout of cities, promoting mixed‑use neighborhoods or preventing the emergence of starkly segregated zones. They also note methodological contributions: the work demonstrates that a simple modification—allowing agent‑specific thresholds—generates a richer set of attractors without altering the underlying interaction mechanism.

Limitations are acknowledged. The current study restricts itself to two groups, ignores economic factors such as housing costs or moving expenses, and assumes static tolerance values. Future research directions include extending the framework to multiple demographic groups, incorporating network‑based neighbourhood definitions, adding dynamic adaptation of τi over time, and coupling the model with economic variables to capture more realistic urban dynamics. Overall, the paper provides a compelling argument that heterogeneity in individual preferences fundamentally reshapes the emergent spatial patterns in Schelling‑type models, moving the discourse beyond the traditional segregation‑integration dichotomy toward a nuanced understanding of mixed urban landscapes.