A systematic review on research methodologies for residential mobility

In this paper, we review possible research methodologies to analyze household residential mobility. Mobility occurs mainly due to the need of housing adjustments. Households compare the levels of utility each alternatives provides and select the alternative providing the highest level. Considering that mobility has discrete nature and is derived from consumer utility maximization problem, appropriate empirical methods should reflect these characterisites. Discrete choice models are well suited in empirical analyses and proved as consistent with consumer theoy. After introducing a simple theoretical framework for residential mobility decisions, we examined discrete choice models with four broad categories; binary choice models (logit or probit), multinomial logit models and conditional logit models, nested logit models, and dynamic discrete choice models, briefly reviewed characteristics and estimation procedures of each model, and discussed the results of existing empirical studies. Based on the review of existing studies and research trends, there has been increasing interest in dynamic discrete choice models because mobility and related housing decisions have dynamic aspects.