Determinants of City Size of China： Development, Government and Location

刘涛,曹广忠

2011-01-01

Abstract:The classical Mills-Muth model explains city size based on growth of the observed city itself, without considering its regional environment and competition with other cities for capital and population. In this study, the authors established a new model from a regional perspective, which appears to be effective to explain the wide range of city sizes in China. Since China is experiencing rapid industrialization, the industrial development also plays an important role in urban growth, even more important than the growth of urban residents. All urban land in China is owned by the state and operated by city governments, so land-related policies will also deeply influence the urban expansion. In China, there are many cities located in places with bad natural conditions, which may also have an impact on their competition with other cities, and on their land use patterns. The general conceptual model was established in a land supply and demand framework in transitional China, in which influences of industrialization, rural-to-urban migration, interest competition between central and local governments in the fiscal decentralization era, and location constraint on cities were explored by comparative static analysis. An empirical model was then established, in which we firstly controlled the scale of urban population and economy; therefore the coefficients of other variables reflect that cities with higher values of these variables tend to obtain more or less land, given that they have the same scale of population and economy with others, i.e., the intensive or extensive use of city land. The results of the econometric model based on land sizes of 626 China’s cities show that 1） rapid growth of city population and industries significantly increase demand for city land. Roles of manufacturing and service sectors in city growth vary with development stage of cities. 2） The impact of per capita GDP or the development level of a city is not significant when we controlled its economic scales. Cities with the FDI-driven development model tend to use city land more intensively, but this effect becomes insignificant when policy and location variables are controlled. 3） Although the central government seeks to restrict urban expansion by the Construction Land Quota policy, local governments that consider fiscal revenue but do not seem to have motivation to land conservation prefer to transfer more cultivated land to urban land by establishment of development zones. This has almost offset the positive impact of the CLQ policy on intensive use of city land. 4） The scarcity of land resources and traffic disadvantage restrict urban expansion by increasing cost of land development or by reducing land demand. 5） Contrary to theoretical expectation, the disadvantage of the macro-economic location of cities stimulates local governments to sell more land with lower, sometimes even negative prices to attract more investment, rather than restricts the land size of cities.