摘要
针对地级市区的金融发展与经济增长的关系,利用中国东部十个省市共87个样本地级市区2000-2006年的数据,基于分位数回归的统计分析结果表明,在被解释变量经济增长指标的不同分位数处解释变量金融发展和控制变量对经济增长影响的差异和波动是统计显著的。与经典的条件均值回归相比,条件分位数回归实证分析能够揭示数据生成过程更加丰富的信息,为区域金融发展与经济增长关系进行时空特征整合的统计建模提供了有力支持。
Targeting at the relationship between regional financial development and economic growth of prefecture-level cities, the paper applies data in 7 years (2000-2006) of 87 sample prefecture-level cities of ten provinces in east China, based on quantile regression method. Empirical analysis results show that on different quantile of dependence variable namely economic growth, there are remarkable statistical difference and fluctuation of effects of independence variables of financial development and other controlled variables. Compared with classical statistical methods, quantile regression method can mine more information from data generating procedure. The empirical analysis results can powerfully support statistical and econometrics modelling based on the integration of spac-time dependence characteristic of relationship between regional financial development and economic growth.
出处
《统计教育》
2009年第3期12-17,共6页
Statistical education
基金
湖南师范大学青年基金项目(06102)
湖南省教育厅基金项目(08C522)
关键词
东部十省市
地级市区数据
金融发展
经济增长
分位数回归
Ten Provinces in East China
Prefecture-level City Data
Financial Development
Economic Growth
Quantile Regression Method