中心性
经济地理学
长江
市场细分
缩放比例
多维标度
代表(政治)
聚类分析
相似性(几何)
空间生态学
三角洲
空间异质性
经济
计量经济学
业务
地理
分布(数学)
城市空间结构
空间分布
比例(比率)
分割
共同空间格局
自然(考古学)
计算机科学
透视图(图形)
产业组织
层次聚类
作者
Zihan Xu,Yunjie Feng,Qi Song,Yao Shen
标识
DOI:10.1177/23998083261460686
摘要
Urban housing markets exhibit pronounced spatial heterogeneity shaped by multiple spatial factors, resulting in scaling characteristics and hierarchical structures typical of complex systems. These properties provide an alternative perspective for delineating housing market areas and interpreting inter-market linkages. This study introduces the concept of natural submarkets, defined as spatially contiguous housing clusters exhibiting both hierarchical scaling in price patterns and consistent hedonic price determinants. We implement this concept through a bottom-up segmentation framework that combines an improved head/tail breaks with hedonic model fit, enabling data-driven delineation of market boundaries and comparative regional analysis. Applied to 26 cities in the Yangtze River Delta (YRD), China, the proposed method yields a more objective and interpretable representation of price differentiation and scaling properties than conventional classifications. The results indicate that economically advanced cities may exhibit either multi-level submarket structures with well-defined hedonic relationships or more complex hierarchical structures with weakly defined pricing mechanisms. Further submarkets clustering by shared pricing mechanism reveals two dominant spatial patterns, comprising interconnected development corridors and isolated market structures. These morphologies are generally driven by network centrality and accessibility to amenities. As a result, the YRD housing price distribution is characterized by a polycentric structure and ongoing cross-boundary integration. The proposed framework complements traditional regional analyses by revealing structural patterns of mechanical similarity and supporting differentiated policy interventions aligned with housing market dynamics.
科研通智能强力驱动
Strongly Powered by AbleSci AI