![Sample our Geography journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5937/TOC-Subject-Sample-2021-Geography.jpg"})
Abstract
In the past decade, Beijing has experienced dramatic spatial restructuring. In this paper, we test whether the conventional monocentric model of urban spatial structure can explain recent developments in Beijing’s land market. Using official land transaction data from 2008 through 2012, we find that (1) for commercial, mixed-use, and residential land parcels, both land prices and development intensities do not decline with distance from the urban center; (2) for industrial land parcels, land prices increase with distance from the urban center. These insignificant and positive land-price gradients defy the predictions of the monocentric model; such results are explained by examining Beijing’s evolution from a monocentric to a polycentric spatial configuration, which has been driven by various urban planning efforts and institutional reforms. Our analysis of post-Olympic Beijing land market dynamics leads to a fresh understanding of the city’s emerging polycentric metropolitan structure.
Acknowledgments
The authors would like to thank two anonymous referees for their constructive comments, which have resulted in an improved manuscript.
Notes
1. “Sealed bidding” means a land parcel goes to the highest bidder, while the “two-stage auction” means the government leases a land parcel to a chosen bidder based on a comprehensive evaluation (Cai et al., Citation2013; Yang et al., Citation2015).
2. For some land parcels, the government may ask land bidders to assist in acquiring the adjacent land (Dai Zheng Yong Di) for the purpose of public use, such as a road, parking lot, or public green space. The cost of the adjacent land is not factored into land transaction prices. Usually, the government bears the majority of the cost of acquiring the adjacent land.
3. From 2008 through 2012, Beijing had constructed three, five, one, one, and one new subway lines each year, respectively. Using GIS software, we calculated the distance for each land parcel to the existing, nearest subway station by the year when the land parcel was leased. For instance, for the land parcels leased in 2008, we calculated the distance to the existing, nearest subway station in 2008. This method can relatively accurately reflect how subways impact the land market in a specific year.
4. For some land parcels, the government has provided only a rough address, such as a particular village or industrial zone. As a result, some parcels’ addresses may be counted more than once and thus are duplicated in .
5. Anselin (Citation2005) delineates the rule for deciding whether to employ a spatial error model or a spatial lag model, based on the Lagrange Multiplier test statistics.
6. Geoda is available at http://geodacenter.asu.edu/.
7. To calculate the Moran’s I values, we employ a row-standardization of distance-based weight matrix (Anselin, Citation2005; Chakraborty, Citation2009; Kissling & Carl, Citation2008). After conducting a set of experiments, we find 1 km is the optimum distance in the spatial weight matrix. In this paper, we also employ 0.5 km and 1.5 km as alternative distances, respectively, and find they will not change the significance of the Moran’s I values.
8. In China, UMP is a tool to regulate urban spatial structure, with an implementation term of 5–20 years. Each UMP provides a land use map to guide land use planning in the near future, through assigning land use types to each land parcel.
9. Usually, the leasing terms are 50 years for commercial use, 40 years for industrial use, and 70 years for residential use.