基于Bayes空间模型的区域金融发展影响因素分析

An Analysis of Factors Influencing Regional Financial Development based on Bayesian Spatial Econometric Model

本文基于区域金融发展的综合指标体系,衡量了我国31个省市的金融发展水平,并利用Moran’s I指数从全局和局部层面揭示了区域金融发展的空间关联性,从Bayes空间计量视角分析了区域金融发展与经济、政治、社会环境和地理位置等因素的关系。研究结果表明:我国金融发展表现为外生性引起的空间误差自相关;在同时将空间相关和空间异质性引入模型后,经济、政治和社会环境对区域金融发展都有一定的正向影响,但普通回归模型高估了空间相关和经济因素对金融发展的贡献程度,低估了政治和社会环境的贡献程度;区域金融发展存在明显的集聚性差异,东部地区的金融发展显著领先于其他区域,中部和西部地区之间无显著差异。

Based on the comprehensive index system of regional financial development, this paper measures the financial development level of 31 provinces of China, uses Moran′s index to reveal the spatial relevancy from overall and partial levels, and investigates the relationship between financial development and economy, politics, social environment and position by exploiting a series of spatial econometric models from a Bayesian perspective.The results show that the spatial correlation between financial development is caused by external shocks, which behaves as spatial errors autocorrelation. By including the spatial correlation and heterogeneity into the model, it′s found that economy, politics and social environ-ment have the positive effects on regional financial development, however, ordinary regression models overestimate the contribution of spatial correlation and economy to financial development, and underestimate the contribution of politics and social environment.Regional financial development exhibits agglomeration effects difference, which is shown by the fact that the development of the eastern region significantly exceeds the other areas, while there is no distinct difference between the middle and the western areas.