Absolute β convergence results.

This article compares the population agglomeration characteristics of the Xi’an metropolitan area in western China with those of metropolitan areas in other regions officially approved by the Chinese government. The kernel density estimation method and Markov chain model were used to conduct the study. The results revealed that from 2010 to 2020, the population agglomeration level of the Xi’an metropolitan area showed a trend of first increasing and then decreasing. The absolute gap in the population agglomeration level between cities within the metropolitan area gradually narrowed, and the polarization phenomenon of population agglomeration was not obvious. Compared with metropolitan agglomerations such as Nanjing, Wuhan, Fuzhou, Changsha-Zhuzhou-Xiangtan, Chongqing, and Chengdu, the Xi’an metropolitan agglomeration had a lower population agglomeration level, with a significant gap. Moreover, there was an obvious “club convergence” phenomenon in the population agglomeration levels of different urban agglomerations. The probability of the population agglomeration level remaining stable was at least 53.85%, indicating that there was a “Matthew effect” in which the rich become richer and the poor become poorer. Through the convergence models of α and β, the analysis suggested that there was no significant α convergence between the population agglomeration level of the Xi’an metropolitan agglomeration and that of other metropolitan agglomerations. Instead, there was a significant β divergence, indicating that the gap between the Xi’an metropolitan agglomeration’s population agglomeration level and that of other metropolitan agglomerations is gradually widening. An integrated theoretical framework of population agglomeration was constructed from three dimensions: producers, consumers, and social people. An empirical analysis was conducted on the causes of population agglomeration in the Xi’an metropolitan area and other metropolitan areas. The multiple regression results showed that the income level, public consumption expenditure level, education level, comfortable living environment, and educational level were important factors leading to differences in population agglomeration. The geographic detector results showed that factors in the consumer dimension were the main reasons for population agglomeration in metropolitan areas.

本文以中国西部西安都市圈为研究对象，将其人口集聚特征与中国政府官方批复的其他区域都市圈展开对比分析。本研究采用核密度估计（Kernel Density Estimation）法与马尔可夫链模型开展研究。研究结果显示，2010至2020年间，西安都市圈的人口集聚水平呈现先升后降的变化趋势；都市圈内各城市间的人口集聚水平绝对差距逐步缩小，人口集聚的极化现象并不显著。相较于南京、武汉、福州、长株潭、重庆及成都等都市圈，西安都市圈的人口集聚水平偏低，二者差距较为显著。此外，不同城市群的人口集聚水平存在显著的“俱乐部趋同（Club Convergence）”现象；人口集聚水平维持稳定的概率不低于53.85%，这表明区域间存在“马太效应”，即富者愈富、贫者愈贫。通过α收敛与β收敛模型开展分析后，结果表明西安都市圈与其他都市圈的人口集聚水平之间不存在显著的α收敛特征，反而呈现出显著的β发散态势，意味着西安都市圈与其他都市圈在人口集聚水平上的差距正逐步拉大。本文从生产者、消费者与社会个体三个维度构建了人口集聚的整合理论框架，并针对西安都市圈及其他都市圈的人口集聚成因开展实证分析。多元回归结果显示，收入水平、公共消费支出水平、教育水平、宜居生活环境以及教育水平是引发人口集聚水平差异的重要影响因素。地理探测器（Geographic Detector）分析结果表明，消费者维度的相关因素是影响都市圈人口集聚的核心动因。