Abstract
Many Chinese cities are in a transition from industrial to post-industrial urban economies. In this process of urban restructuring, land use becomes polycentric and fragmented. More sophisticated models are needed to estimate the amenity effects of this complex residential environment. This article assesses the relative housing price effects of neighborhood characteristics and accessibility in Nanjing, China. This is achieved with a hedonic price model that incorporates detailed spatial measures, geographical contingency, and a modified version of Alonso’s (1964) general theory of land rent. A crucial finding is that the effect of job accessibility on house price varies depending upon the specific sector of employment. Accessibility to jobs in the public and private service sectors has strong positive effects. However, housing proximity to heavy industries has a spatially nonlinear effect: negative in close proximity, but positive at a larger distance. Second, when we control for job accessibility, access to public transport has an added positive effect. Finally, neighborhood “quality” (defined in terms of nearby amenities) is also relevant, but far less than access to service employment. This research shows that Nanjing’s housing prices are affected by different residential characteristics than those with dominant price effects in Western cities.
Acknowledgments
The authors would like to thank Niels van der Vaart, Tom de Jong, and Jianxi Feng (members of the Faculty of Geosciences at Utrecht University), and Tom Kauko (member of the Department of Geography at Norwegian University of Science and Technology) for their assistance at various stages of this research.
Notes
1. This sector has been growing most rapidly in the last two decades, despite the fact that the secondary sector still accounts for a substantial proportion of the city’s economy.
2. However, close proximity to these access points may incur disamenities like noise, pollution, or other nuisances (Hagoort, Citation2006).
3. Job numbers were drawn from the Statistical Bureau of Nanjing (Citation2011).
4. In this case, a usually varies between 1 and 2.
5. We included the Yangtze River in a preliminary test but found no significant effects, probably because approaching the river bank is not easy and the view of the river is considered much less attractive than that of other urban water bodies.
6. Our interpretation of dummy variables was conducted according to Kennedy (Citation1981).