Understanding the role of land finance in economic and infrastructure development in Chinese cities: new evidence from a novel structural equation modelling approach
46 Pages Posted: 28 Nov 2021
Date Written: October 29, 2021
The rapid urbanisation in Chinese cities features a distinctive land finance model, where land market, local economy, government revenue, and urban development are intertwined. Quantifying the interdependence between land market and other parts of the social, economic, and political systems has been a challenging undertaking, and the task is further complicated by the great cross-city heterogeneity in natural endowment and local socio-economic conditions. Few existing studies succeeded in capturing both the complexity of the system and the nuance of cross-city variations at the same time. We propose a novel structural equation modelling (SEM) method, integrated with the latent class analysis (LCA), to address this challenge. The LCA is used to identify distinct city groups based on two purposely constructed land-use efficiency measurements. The categorical latent classes of cities are then incorporated in a series of structural equation models, capturing the non-linear heterogeneity across cities. Based on data for 272 prefecture-level Chinese cities between 2012 and 2017, we found quantified evidence on both the direct channel (i.e., one-off revenue from land conveyance fee) and indirect channel (e.g., sustainable tax revenue from the business and employment growth enabled by land development) through which land supply drives urban development. The study also quantifies the significant gap among Chinese cities in terms of land-use efficiency. Our findings highlight the importance of developing and implementing reliable land-use efficiency measurements, the need to shift policy focus from one-off income to long-term sustainable revenue, and the potential of lower-tier cities in the next stage of urbanisation in China.
Keywords: Land use, Land finance, Structural equation model, Latent class analysis
JEL Classification: R
Suggested Citation: Suggested Citation