Second-order micromotives and macrobehaviour

Abstract

This paper deals with the gap between the preferences of individual agents and collective outcomes. The typical example of this topic is the Schelling–Sakoda segregation model where a first-order discrimination criterion is used to decide whether or not a person will leave his place. Based on the ratio of “acceptable” people around one individual, simple simulations show that even tolerant agents end up being spatially aggregated far beyond of the local requirement of their tolerance level. Here, we use a second-order discrimination criterion to compel people to leave their places, that is a criterion based on the result of the first-order one. According to the respective strength of the two criteria, the problematic will be to determine the spatial repartition of the agents resulting from their moves. This paper provides some answers to the question: does a second-order criterion contribute or not to reduce the gap between micromotives and macrobehaviour?