Modeling Reason and Time of Residential and Job Relocations: Application of a Simultaneous Nested Hazard Based Modeling Structure

Abstract

This study examines the assumption that the decisions on residential relocation and job relocation are made with respect to each other in an endogenous fashion. The interdependency between residential relocation and job relocation is captured with a focus on time of relocation and the reason triggering the relocation. A parametric proportional hazard based model is used to model the time of relocation. For every reason of relocation a univariate survival function is defined. The multivariate survival function is then approximated using the marginal univariate distributions by utilizing the joint survival function. The longitudinal data from the Household, Income, and Labour Dynamics in Australia (HILDA) Survey is used for developing the models. The results of the model estimation practice indicate a positive correlation between job relocation and residential relocation events. Therefore, the presumption that individuals contemplate changing residence and workplace together is confirmed and can be captured using the proposed model of this study.