解析空间相关的来源:理论模型与经验证据

The Origin of Spatial Correlation:A Theoretical Model and Empirical Evidence

本文从需求和供给两个角度研究经济要素的空间流动,发现要素流动存在两种类型的空间交互作用,分别是流出地空间相关和流入地空间相关;更进一步,本文将要素流动方程转化为易于估计的空间相关模型。因此,本文为LeSageandPace(2008)类型的空间自相关计量模型提供了一种经济学解释。在此基础上,使用中国跨省人口流动数据进行经验检验,结果支持理论模型的结论,即资源流动过程中存在两种类型的空间相关,并且忽略空间相关会导致模型估计结果不可靠;大量稳健性检验结果表明存在空间相关的结论是稳健的。此外,空间相关的引入有助于更合理地解释"边界效应"、"全球化缺失之谜"等经济现象。

This paper considers regional movement of economic resources from both de- mand and supply. There exist two kinds of spatial interactions in the movement, origin-basedand destination-based spatial correlation, respectively. A spatial correlation model based on the equation of resources＇ movement is estimated. It offers an economic explanation for spatial correlation proposed by LeSage and Pace （2008）. Empirical studies using inter-province mi- gration data in China support the existence of two kinds of spatial correlations in the move- ment of resources, and ignoring them will lead to unreliable estimation results. Our approach of spatial e{fect helps to explain some economic facts, such as famous ＂border effect＂ and ＂missing globalization puzzle＂.