Spatial distribution of rural population using mixed geographically weighted regression: Evidence from Jiangxi Province in China.

Abstract

On the basis of the spatial panel data of 2000, 2005, 2010, and 2015, this study uses a mixed geographically weighted regression model to explore the spatial distribution characteristics and influencing factors of the rural (permanent) population in Jiangxi Province, China. Results show that residents in the county area have a significant spatial positive autocorrelation, especially in the lake and mountain areas and the global Moran' I index is more than 0.05. The influence of social and economic factors presents spatial homogeneity. The effect of urbanization and per capita disposable income is negative, whereas that of agricultural output value and rural electricity consumption is positive. The influence of climate factors presents spatial heterogeneity. The influence coefficient of rainfall in 2015 ranges from [-0.061, 0.133], which has a negative effect on the southwest mountain areas and a positive effect on the northeast lake areas., The influence coefficient of temperature in 2015 ranges from [-0.110, 0.094], which has a positive effect on the southwest mountain areas and a negative effect on the northeast lake areas. The influence coefficients of wind speed and relative humidity range from [-0.090, 0.153] and [-0.069, 0.130] in 2015 respectively, which further reinforce this effect. Therefore, scholars should pay attention to the universal adaptability of economic and social factors. Moreover, they should consider the spatial difference of climatic factors to promote urbanization following the local conditions. Finally, policymakers and concerned non-governmental institutions should fully understand the sensitivity of the rural population in underdeveloped mountain areas to climate factors to promote their rational distribution.