Nonparametric Estimation and Analysis of Regional Correlation Structure——empirical Research Based on China's Per Capita GDP Growth Rate

WANG Lili,MIAO Baiqi
DOI: https://doi.org/10.3969/j.issn.0253-2778.2011.12.004
2011-01-01
JUSTC
Abstract:Nonparametric and quantitative estimation of the spatial correlations between the 31 provinces in China was made based on the data of annual per capita GDP growth rates from 1978 to 2008 by SG method which considers the effect of both time and space,overcoming the shortcoming of mere qualitative description in the past.The rationality of this method has been proven compared with sample correlation.The empirical results show that there is a long-term,significantly positive correlation between economic developments,and this trend exhibits a slight weakening characteristic after a high degree of concentration with the progression of time.It also verifies the existence of spatial heterogeneity.Finally,cluster analysis was conducted based on the SG correlations between the provinces.
What problem does this paper attempt to address?