The Nonlinear Influence of Street Quality on Housing Prices Based on Random Forest Model: A Case Study of Guangzhou

Abstract

Streets are important spaces for transportation and life. It is important to clarify the economic effects of street quality on the real estate industry to create high-quality street spaces and meet the environmental needs of residents. Based on factors influencing traditional housing prices, such as street quality, this study used the random forest model and an accumulated local effects plot to explore the nonlinear relationship between street quality and housing prices, with its threshold effect in Guangzhou. The results show that the random forest model's fitting accuracy is 0.145 times higher than that of traditional linear methods and that the random forest model effectively captures the nonlinear relationship and threshold effect between street quality characteristics and housing prices. Commercial location, global betweenness, and the years of community construction significantly impact housing prices. The total contribution of green vision rate, sky openness, and enclosure to housing prices reached 5.68%, and residents' preferences for a comfortable environment drew attention. When distance increased, the negative influence of the commercial and economic location on housing prices gradually decreased. When global betweenness is less than 1,715, it has an exponential positive correlation with housing prices. When local proximity exceeds 117, it has little impact on housing prices, revealing the alternative effects of transportation hubs at different levels. Positive externalities, such as convenient transportation and negative externalities, such as traffic congestion and environmental pollution, affect housing prices equally. Road construction should consider traffic efficiency and environmental livability. When the degree of road motorization is less than 23% and more than 26%, the construction ratio of protective isolation facilities is more than 4.27%, and the night light brightness is higher than 0.0037 W/(m2·sr·μm), they have a lesser impact on housing prices. The over-construction of safety facilities reduces the economic benefits of street quality. Moreover, to ensure fair street usage, the green vision rate of the living environment should not be less than 15%. The spatial distribution trends of sky openness and enclosures are opposite, but both negatively impact housing prices. In urban development and construction, the development intensity should be reasonably controlled to avoid psychological depression caused by space cramping. For future street-stock renewal, it is necessary to accurately understand the current situation of street quality and the needs of residents, specifically improving street quality. Avoiding street quality creates property premiums and strives to ensure the fairness of residents' access to and use of street spaces. It guides real estate developers to participate in street construction. This study compensates for the limitations of traditional characteristic price models and the lack of explanatory properties in traditional linear methods, thereby providing a scientific basis for improving residents' living environments and building a livable city.

Keywords

• street quality
• housing prices
• nonlinear relationship
• threshold effect
• guangzhou