Spatiotemporal Evolution, Agglomeration Heterogeneity, and Poverty-Causing Factors of Poverty-Stricken Counties in China

Poverty persists as a pervasive global challenge, posing important barriers to the advancement of both social equity and economic development. Against this backdrop, this study focuses on rural poverty in China, with the specific aim of investigating its dynamics through a spatiotemporal analysis of impoverished counties across multiple time periods. To achieve this goal, the DBSCAN (density-based spatial clustering of applications with noise) spatial clustering algorithm is first utilized to identify the spatial clusters of poverty and associated risk areas; subsequently, the MGWR (multiscale geographically weighted regression) model is used to evaluate the spatial characteristics of poverty-causing factors and unpack the underlying mechanisms driving poverty. The study identifies 663, 832, and 52 state-level impoverished counties in China in 2001, 2016, and 2021, respectively. Moreover, the counties are categorized into 6 types based on the principal factors causing poverty, namely, frontier terrain constraints, infrastructure limitations, geological disaster vulnerabilities, minority aggregation industry constraints, underdeveloped old revolutionary areas, and ecological fragility. Among these factors, topographic relief and healthcare institutions exhibit a negative correlation with poverty, while average slope and nighttime light index display a positive correlation. Furthermore, location conditions, farmland production potential, vegetation index, residents’ deposit balance, and local government revenue demonstrate a 2-way influence on gross domestic product per capita in impoverished counties. This study offers targeted and actionable insights to local governments, supporting the localized and effective implementation of targeted poverty alleviation strategies.