Spatially varying relationships between land ownership and land development at the urban fringe: A case study of Shenzhen, China

Abstract The conversion of rural land to urban land, known as land urbanization or urban expansion, often follows two tracks in China: 1) the government becomes the owner of the land through land expropriation and performs primary land development and land transfer; 2) villagers maintain their ownership of the land collectively and convert their non-construction land into construction land by themselves. The dual land ownership system is a unique factor which influences urban expansion in China. However, this influence may be inconsistently distributed over space. Local variations in land use regulations, government controls, and stakeholder participation could all lead to varied relationships between land ownership and land development. This research aims to answer two main questions, 1) whether or not the influence of the dual land ownership system on land development is consistent over space (spatial stationarity), and 2) how such relationships vary over space. Capitalizing on the land parcel data collected in Longhua district in Shenzhen in the years 1996 and 2010, this research utilizes a geographically weighted logistic regression (GWLR) modelling approach to test and visualize spatial non-stationarity in the relationship between land ownership and land development. The study focuses on Shenzhen, China, because it has experienced extremely rapid urbanization and because of the notable dual land system there. The results demonstrate that land ownership (collectively-owned vs. government-owned) has statistically significant relationships to land development and that such relationships vary over space. There are areas where land ownership profoundly influences land development, whereas in other areas the influences are limited. GWLR proved to be a useful exploratory spatial analysis tool which may lead to new insights on the interconnections between dual land ownership and urbanization.