One of the main challenges of today’s analytical geography is the back-linking of observed phenomena to individual actions, as not determined by individuals’ structural situation in a one-goal society but by their very choice of spatial goals, i.e. by what shall be called individuals’ actor-dimension. In this paper we present an actor-based model of urban residential mobility. Results show that public policies in urban development must take into account and act on this actor-dimension if they are not to obtain effects opposite to their aims.