Key points are not available for this paper at this time.
This study employs a unified methodological approach, integrating vector error-correction modeling (VECM) with directed acyclic graphs (DAGs), to rigorously examine the changing interrelationships in monthly housing prices across five major Jiangsu Province cities: Nanjing, Xuzhou, Wuxi, Kunshan, and Suzhou. The investigation covers a substantial timeframe of nearly 13 years (January 2012–July 2024). Complementary computational methods infer the underlying DAG structure: first, the PC algorithm establishes an initial set of plausible causal patterns; second, the LiNGAM (Linear Non-Gaussian Acyclic Model) technique utilizes data non-Gaussianity to resolve directional uncertainties, pinpointing a definitive causal sequence. Using this DAG-derived causal order, thorough innovation accounting analyses — incorporating impulse response functions — are conducted to measure dynamics and shock influences within the price system. Results uncover complex transmission pathways and varied adjustment rates in the provincial property market after external shocks. Significantly, empirical evidence indicates that policy measures aimed at boosting property values in Xuzhou and Suzhou can significantly enhance broader market recovery across Jiangsu due to their pivotal network positions. Conversely, for other cities showing lesser systemic impact, findings suggest localized revitalization strategies offer more direct and effective pathways than province-wide stimulus or core-focused measures, underscoring the need for policy differentiation based on geographic network roles.
Jin et al. (Wed,) studied this question.