La ségrégation spatiale selon Schelling : la perversité est ailleurs

In the 1970s, Thomas C. Schelling proposed a model which claimed to show that a high degree of spatial segregation can result from individual preferences which do not in themselves aim to achieve such a degree of collective segregation. A perverse effect seems therefore to occur. However, the present paper demonstrates that this model contains several biases which totally explain the obtained levels of segregation. A real weak individual demand for similar neighbours leads to low social segregation and a strong demand leads to high segregation. The link between the two is linear. The levels of entropy that are globally achieved for a given threshold of individual satisfaction are not at all surprising in relation to the laws of chance and to what each individual actually wishes