RESIDENTIAL MOBILITY IN A CHANGING HOUSING SYSTEM:

Based on retrospective life histories derived from a sample of 1,500 households, the present study estimates the changing rate of residential mobility in Guangzhou, China, during the period 19802001. The estimated mobility rate shows a rapidly rising trend and this increase remains pronounced after taking into account the cohort effect inherent in retrospective histories. Marketization of the housing provision system has brought about increased residential changes. Discrete-time logit analysis shows that elements of the socialist redistributive economy. particu­larly the work-unit system, remains important in structuring residential change under market transition. More specifically, education and memhership in the Chinese Communist Party enhance mobility. Almost invariably, change in employment is associated with a move. Whereas owning an apartment purchased from the work unit is associated with substantially lower mobil­ity than are other ownership categories, renting from the work unit is associated with higher mobility than are other rent categories. However, as the case in the West, heightened residential mohility in Guangzhou is also a demographically-driven process. Change in marital status likely triggers residential change. Age exhibits the usual curvilinear effects, and gender is an important diffcrcnlialor of residential mobility.