Supply-Demand Compatibility of Elderly Care Facilities and Its Driving Factors: A Random Forest-Based Study

The rapid development of new urban areas, coupled with an aging population, presents challenges for the equitable provision of public services. Focusing on elderly care facilities in Shanghai's five new cities, this research constructs a supply-demand compatibility evaluation framework with dual dimensions of quantity correlation and type correlation and employs a Random Forest model to identify the key drivers influencing the compatibility level. Based on multi-source data including facility points of interest (POI) and resident surveys, the study evaluates the match between facility supply and resident demand within 5-10 minute and 15-minute life circles. The results reveal a significant structural mismatch. This is characterized by a severe deficit of facilities within the high-demand 5-10 minute life circle, contrasted with a relative surplus at the 15-minute level. The model analysis further indicates that this spatial imbalance in the supply structure is the core factor affecting compatibility. Furthermore, compatibility levels vary greatly among the new cities, and the types of services offered often fail to align with residents' specific needs, particularly at the community level. The study suggests that future planning should shift towards a more precise, demand-responsive model, prioritizing the development of community-level facilities to better match the actual living requirements of residents.