Analysis of a dynamic residential location choice model with transaction costs

This paper incorporates transaction costs in a residential location choice model. An individual's choice is assumed to be an outcome of a two‐stage process: a decision to change, or to undertake a transaction and, conditional on a change occurring, a choice of a new alternative. The dynamic choice model is aggregated to yield a Markovian model of residential location patterns. It is shown that recent contributions to dynamical urban modeling correspond to special cases of the deterministic version of this model. The Markovian model is used in a theoretical analysis of the influence of transaction costs on the properties of the stationary state. The effects of residential mobility rates and of interdependencies among individuals, caused by supply‐side and density‐related interactions, are also analyzed. It is shown how these dynamical factors modify the stationary state, thereby demonstrating the type of errors which may occur with static models that omit them.