Scholarlink

<p>This study employs an integrated methodology, synthesizing vector error-correction modeling (VECM) with directed acyclic graphs (DAGs), to conduct a thorough investigation into the changing interdependencies observed within monthly residential property values across six major urban centers in Fujian Province: Xiamen, Quanzhou, Fuzhou, Sanming, Putian, and Longyan. Our empirical examination covers an extensive timeframe exceeding nine years, specifically from April 2015 to July 2024. To construct the foundational DAG framework underpinning the causal analysis, we implement a sequential computational process. First, the PC (named for Peter Spirtes and Clark Glymour) algorithm generates an initial, tentative network depicting plausible causal linkages between the cities’ housing markets. Following this, the LiNGAM (Linear Non-Gaussian Acyclic Model) technique is applied; it utilizes inherent non-Gaussian properties within the dataset to resolve any remaining uncertainties regarding causal directionality, thereby deriving a definitive causal ordering structure. Utilizing the unambiguous causal sequence established by this DAG procedure, we then perform detailed innovation accounting analyses. This incorporates impulse response function assessments, which precisely measure the dynamic interactions and quantify the propagation effects of economic shocks throughout the interconnected price system. The investigation uncovers complex transmission pathways and heterogeneous adjustment velocities across the provincial real estate market when responding to external perturbations. Significantly, empirical outcomes demonstrate that policy interventions intended to elevate property values in Longyan and Xiamen possess a disproportionately positive impact, capable of stimulating broader market recuperation throughout Fujian. This effect stems directly from these cities’ pivotal positions within the identified spatial network. Conversely, for the other four urban areas—Quanzhou, Fuzhou, Sanming, and Putian—which exhibit comparatively lower systemic influence, the analysis yields critical insights. Findings strongly suggest that localized revitalization strategies, tailored specifically to each city’s unique circumstances, offer a more direct and efficient path to recovery than either province-wide stimulus packages or policies concentrated solely on core metropolitan centers. This underscores the essential requirement for policy differentiation, meticulously grounded in an understanding of each location’s distinct functional role within the overarching spatial economic network.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Inter-city housing price causality in Fujian: contemporaneous analysis via vector error-correction and directed acyclic graph models

  • 期刊论文

  • 发表日期：

    2026年4月6日

查看全文
  • Bingzi Jin,
  • Xiaojie Xu

摘要

This study employs an integrated methodology, synthesizing vector error-correction modeling (VECM) with directed acyclic graphs (DAGs), to conduct a thorough investigation into the changing interdependencies observed within monthly residential property values across six major urban centers in Fujian Province: Xiamen, Quanzhou, Fuzhou, Sanming, Putian, and Longyan. Our empirical examination covers an extensive timeframe exceeding nine years, specifically from April 2015 to July 2024. To construct the foundational DAG framework underpinning the causal analysis, we implement a sequential computational process. First, the PC (named for Peter Spirtes and Clark Glymour) algorithm generates an initial, tentative network depicting plausible causal linkages between the cities’ housing markets. Following this, the LiNGAM (Linear Non-Gaussian Acyclic Model) technique is applied; it utilizes inherent non-Gaussian properties within the dataset to resolve any remaining uncertainties regarding causal directionality, thereby deriving a definitive causal ordering structure. Utilizing the unambiguous causal sequence established by this DAG procedure, we then perform detailed innovation accounting analyses. This incorporates impulse response function assessments, which precisely measure the dynamic interactions and quantify the propagation effects of economic shocks throughout the interconnected price system. The investigation uncovers complex transmission pathways and heterogeneous adjustment velocities across the provincial real estate market when responding to external perturbations. Significantly, empirical outcomes demonstrate that policy interventions intended to elevate property values in Longyan and Xiamen possess a disproportionately positive impact, capable of stimulating broader market recuperation throughout Fujian. This effect stems directly from these cities’ pivotal positions within the identified spatial network. Conversely, for the other four urban areas—Quanzhou, Fuzhou, Sanming, and Putian—which exhibit comparatively lower systemic influence, the analysis yields critical insights. Findings strongly suggest that localized revitalization strategies, tailored specifically to each city’s unique circumstances, offer a more direct and efficient path to recovery than either province-wide stimulus packages or policies concentrated solely on core metropolitan centers. This underscores the essential requirement for policy differentiation, meticulously grounded in an understanding of each location’s distinct functional role within the overarching spatial economic network.