Effects of city size, shape, and form, and neighborhood size and shape in agent-based models of residential segregation: are Schelling-style preference effects robust?

Abstract

In 1971 Thomas Schelling outlined a model in which group segregation emerged out of the residential decisions of individual agents who were guided by relatively weak preferences for same-group contact, preferences that on first consideration appeared to be compatible with integration (Schelling, 1971). This agent-based model has proven to be an important tool in the study of segregation dynamics and the role of individual preferences in producing patterns of segregation. Rigorous formal studies (Young, 1998;Zhang, 2004a) have examined the underpinnings of important theoretical insights obtained from the model. An even larger number of empirical studies have explored the role of selected model parameters. Schelling initiated this line of research by examining several variations on his initial model. Subsequent work by others has explored different aspects of the model in more detail. For example, later studies have focused on varying implementations of agent vision (Laurie and Jaggi, 2003), refining decision rules that agents use when choosing their residential location on the basis of their neighbors (Pancs and Vriend, 2003; Sander et al, 2000), introducing competing residential preferences (Wasserman and Yohe, 2001), using unbounded torus landscapes

Segregation That No One Seeks

“…3 Schelling concluded that small preferences for similarity can lead to massive segregation. This conclusion is quite robust across many changes to the model, including different utility functions (Zhang 2004;Bruch and Mare 2006;Pancs and Vriend 2007), different rules for updating (Bruch and Mare 2006), differing neighborhood sizes, and different spatial configurations (Fossett and Dietrich 2009). These studies show that it is extremely hard to avoid segregation when agents have some preference for like neighbors.…”

Una Exploración Conceptual de la Formación de Patrones en Sistemas Sociales Autorganizados

“…Dado el carácter material de los métodos computacionales empleados para la aproximación a los fenómenos complejos, la comprensión computacional también depende de los avances en materia ingenieril. Varias de las extensiones con las que se ha mostrado la robustez del mecanismo subyacente de asimetría espacial del modelo de Schelling hacen uso de modelos en donde hay una representación espacial con alto nivel de realismo [28]. Generar estas representaciones, sin embargo, requirió inicialmente de un avance ingenieril en términos del desarrollo de sistemas híbridos que combinan modelos basados en agentes y sistemas de información geográfica [29].…”

“…No es claro de qué forma, ni hasta qué punto, diferentes métodos puedan capturar las intuiciones y creencias más importantes sobre la ontología de fenómenos sociales complejos. El modelo de Schelling, por ejemplo, es un autómata celular cuyos resultados han sido confirmados por diferentes replicaciones y extensiones que han incluido una cantidad mucho mayor de información empírica [28], [32]. Si bien el modelo original es extremadamente sencillo, esútil porque captura el mecanismo subyacente de asimetría espacial, que es el mismo con el que operan modelos mucho más robustos e intricados de segregación espacial.…”

Why We Cannot Learn from Minimal Models

“…After the publication of Schelling's model, several studies (see e.g. Fossett and Dietrich, 2009, Pancs and Vriend, 2007, and Zhang, 2004 have documented the robustness of Schelling's result across a wide range of modifications in his model (e.g. think of changes in neighbourhood sizes, spatial configurations, and individuals' tolerance thresholds for individuals of a different type).…”