Can residential integration be a stable equilibrium?

Shelling’s model and its different versions predict that segregation is the unique stable equilibrium. However, cities around the world show different levels of intermingle, and it is very uncommon to observe fully segregated urban patterns even when segregation is high. We argue that individuals do not take into account only their neighbours characteristics for making a location decision. In particular, if residential segregation generates ghettos of low-skilled labourers, the societal level of production would diminish and, consequently, individuals’ level of consumption too, affecting all society members’ well-being. When agents care about the impact of segregation on their own consumptions, we prove that the set of system’s equilibria can be one containing states of residential integration. Besides, using simulations, we compare the aggregated utility related to these equilibria with the maximum level of aggregated utility. First published online: 03 Nov 2015