地理科学 ›› 2012, Vol. 32 ›› Issue (8): 905-912.doi: 10.13249/j.cnki.sgs.2012.08.905

• • 上一篇    下一篇

城市体系位序-规模特征的空间计量分析 ——以中部地区地级以上城市为例

程开明(), 庄燕杰   

  1. 浙江工商大学统计学院, 浙江 杭州 310018
  • 收稿日期:2011-08-14 修回日期:2012-01-12 出版日期:2012-08-20 发布日期:2012-08-20
  • 作者简介:

    作者简介:程开明(1975-), 男, 湖北广水人, 博士, 副教授, 主要从事城市与区域经济方面的研究。E-mail: chengkaim@tom.com

  • 基金资助:
    国家社科基金项目(09BTJ003)、国家自然基金项目(71173190)、浙江省高校优秀青年教师资助计划、浙江省社科规划“之江青年课题研究”、浙江工商大学研究生科研创新基金项目(1020XJ1511056)资助

Spatial Econometric Analysis of The Rank-size Rule for Urban System: A Case of Prefectural-level cities in China’s Middle Area

Kai-ming CHENG(), Yan-jie ZHUANG   

  1. School of Statistics, Zhejiang Gongshang University, Hangzhou, Zhejiang 310018, China
  • Received:2011-08-14 Revised:2012-01-12 Online:2012-08-20 Published:2012-08-20

摘要:

随着“中部崛起”战略的实施,中部地区的城市快速发展并引起广泛关注。为增进对中部地区城市体系发展规律的认识,根据1985~2009年城市人口数据,利用非参数密度估计方法得到该地区地级以上城市的Kernel密度图,发现城市规模的核密度曲线不断右移,城市整体规模日益扩大;采用空间计量模型对该地区地级以上城市的位序-规模特征进行分析,结果显示城市间的空间依赖性使空间滞后模型估计的Zipf指数值比OLS估计结果要小,表明考虑空间效应的城市体系规模分布更为集中。随后进一步解析空间依赖性影响城市规模结构演化的内在机制,认为要素集聚、产业调整和创新扩散等造成的空间相互作用提供了空间依赖性赖以存在的基础,影响城市体系的演化和位序-规模变动,使城市体系规模分布更趋均衡。结论对于明晰城市在区域城市体系中的定位,确定合理的发展战略具有参考价值。

关键词: 位序-规模法则, 空间计量模型, 演化机制, 中部地区

Abstract:

A remarkable empirical regularity is that the city size distribution in many countries is well approximated to a Pareto distribution. This claim is so widely accepted that it has gained the status of a law, Zipf’s Law, or a rule, the Rank-size Rule. It has also inspired extensive researches mainly in the fields of economic geography and regional science. The urbanization process has mainly taken place since 1978 with significant processes of industrialization and economic growth in China. Many researchers have studied the characteristics of city scale distribution in different regions according to Ordinary Least Squares (OLS). The spatial econometric analysis of urban system is still scarce. This article adopts spatial econometric models to analyze the Rank-size Rule and evolution mechanisms of urban system in China’s middle area through population data of prefectural-level cities from 1985 to 2009. Firstly, cross-sectional distribution of urban population is analyzed by means of nonparametric estimations of density functions. Evolution of the shape of urabn population cross-sectional distribution shows the existence of convergence trends. Secondly, the linear regression equation is estimated by OLS. Then spatial autoregressive model and spatial error model are estimated. The results show that the estimation over time of the q parameter displays an increasing trend from 1985-1990, then decrease until 1995, from which it starts to augment. Zipf’s exponent estimated by spatial lag model is smaller than OLS. It indicates that the size distribution of urban system is more convergent because of significant spatial dependence between cities. Thirdly, we explore the mechnism between the spatial dependence and city size distribution. Spatial interaction caused by the agglemation of production factors, industrial adjustment and innovation diffusion provides the basis of spatial dependence. It affects the evolution of urban system and makes urban size distribution more convergent. Finally, the article concludes with a summary of key findings and puts forward some recommendations.

Key words: Rank-size Rule, spatial econometric model, evolutionary mechanism, middle area of China

中图分类号: 

  • K902