摘要
AbstractHousing prices are significantly influenced by the presence of public facilities, such as schools, parks, and transport infrastructure. Whereas existing literature has mainly focused on the proximity of public facilities, this study goes beyond proximity and introduces the concept of quality metrics to evaluate public facilities. By employing the gradient-boosting decision trees approach, we analyze the nonlinear relationships between public facilities and property values in Shenzhen, China. Our study not only quantifies the extent to which the quality of these facilities is capitalized in housing prices, but also examines the interaction effects of quality and proximity on housing prices. Our results reveal that quality variables exhibit a greater relative importance than proximity variables in determining housing prices, and this relationship follows a nonlinear pattern. Furthermore, we investigate the moderating effects of quality on the relationship between proximity and housing prices. We find that the amplifying effects of higher quality are particularly evident in metro stations and public middle schools, whereas the impact of park quality on housing prices is less pronounced. These findings highlight the need to consider both quality and proximity in the supply of public facilities, as they have synergistic effects on housing prices. The nonlinear effects observed in our study can serve as a valuable tool for identifying deficiencies in the supply of public facilities. Additionally, the distinction between proximity and quality, as well as their interaction effects, contributes to our understanding of how the value of public facilities is capitalized in housing markets.住房价格在很大程度上受到学校、公园和交通基础设施等公共设施的影响。文献主要关注公共设施的邻近性。本研究在邻近性的基础上, 引入了质量指标的概念来评估公共设施。采用梯度提升决策树方法, 分析了中国深圳市公共设施与物业价值之间的非线性关系。不仅量化了这些设施的质量在房价中的资本化程度, 还探讨了质量和邻近性的相互作用对房价的影响。结果表明, 在决定房价方面, 质量变量比邻近性变量有更大的相对重要性, 并且这种关系呈非线性模式。本文还研究了质量对邻近性和房价关系的调节作用。高质量设施的放大效应在地铁站和公立中学尤为明显, 但是公园质量对房价的影响并不显著。这些发现表明, 由于公共设施的质量和邻近性对房价有协同增效作用, 需要同时考虑质量和邻近性。本研究发现的非线性效应, 是识别公共设施不足的宝贵工具。邻近性和质量的区别及其相互作用, 有助于我们理解公共设施价值在住房市场中的资本化。Los precios de las viviendas están muy influidos por la disponibilidad de ciertos servicios o equipamientos públicos, como escuelas, parques e infraestructura del transporte. En tanto que la literatura existente se han centrado principalmente en la proximidad de los equipamientos públicos, este estudio va mucho más allá de considerar la proximidad e introduce el concepto de la métrica de calidad para evaluar aquellas conveniencias. Empleando el enfoque de árboles de decisión de gradientes ascendientes, analizamos las relaciones no lineales entre los servicios públicos y los valores de la propiedad en Shenzhen, China. Nuestro estudio no solo cuantifica la extensión a la cual la calidad de estos elementos se capitaliza en los precios de las viviendas, sino que examinan también los efectos de la interacción de la calidad y la proximidad sobre los precios de las viviendas. Nuestros resultados revelan que las variables de la calidad exhiben una importancia relativa más alta que las variables de la proximidad en la determinación de los precios de las viviendas, y que esta relación sigue un patrón no lineal. Además, investigamos los efectos moderadores de la calidad sobre la relación entre la proximidad y los precios de la vivienda. Encontramos que los efectos amplificadores de una más alta calidad son particularmente evidentes en lo que concierne a las estaciones del metro y las escuelas públicas medias, mientras que el impacto de la calidad de los parques sobre los precios de la vivienda es menos pronunciado. Estos hallazgos destacan la necesidad de que se tome en cuenta tanto la calidad como la proximidad en la provisión de los equipamientos públicos considerados, en la medida en que estos tienen efectos sinérgicos sobre los precios de la vivienda. Los efectos no lineales observados en nuestro estudio pueden servir como herramienta valiosa para identificar deficiencias en la provisión de equipamientos públicos. Adicionalmente, la distinción entre proximidad y calidad, lo mismo que sus efectos de interacción, contribuye a nuestra comprensión de cómo el valor de estos equipamientos es capitalizado en los mercados de la vivienda.Key Words: Chinagradient-boosting decision treeshousing pricesnonlinear effectspublic facilities关键词: 中国梯度提升决策树房价非线性效应公共设施Palabras clave: árboles de decisión de gradientes ascendientesChinaefectos no linealesequipamientos públicosprecios de la vivienda AcknowledgmentXin Li served as the corresponding author for this article. We would like to thank the editor and anonymous reviewers for their constructive comments.Notes1 Whether an apartment is decorated or not is typically defined in the policy as having finished interior apartment walls, basic facilities such as cabinets and bathroom fixtures, running water, and electrical service.Additional informationFundingThis research was funded by the Key Program of Shenzhen Social Science Federation (Grant No. SZ2022A008) and the Stable Support Program Project of Shenzhen Municipal Science and Technology Innovation Committee (Grant No. GXWD20201231165807007-20200812133137001).Notes on contributorsDe TongDE TONG is an Associate Professor in the School of Urban Planning and Design, Peking University Shenzhen Graduate School, Shenzhen, Guangdong 518055, China. E-mail: tongde@pkusz.edu.cn. Her research interests include housing studies, urban regeneration, and land use policy and governance.Yue ShenYUE SHEN is a PhD candidate in the School of Urban Planning and Design, Peking University Shenzhen Graduate School, Shenzhen, Guangdong 518055, China. E-mail: shenyue@stu.pku.edu.cn. Her research interests include equity in public facilities, informal settlements governance, and socio-spatial segregation in China.Xiaoguang WangXIAOGUANG WANG is a Professor in the Department of Geography & Environmental Studies, Central Michigan University, 287 Dow Science Complex, Mount Pleasant, MI 48859, USA. E-mail: xiaoguang.wang@cmich.edu. Her research interests include land use and transportation planning, spatial analysis, and GIS applications in planning.Yiyu SunYIYU SUN is a PhD candidate in the School of Urban Planning and Design, Peking University Shenzhen Graduate School, Shenzhen, Guangdong 518055, China. E-mail: 18813051707@163.com. His research interests include park accessibility, rural transformation, and urban-rural linkages.Ian MacLachlanIAN MACLACHLAN is a Professor in the School of Urban Planning and Design, Peking University Shenzhen Graduate School, Shenzhen, Guangdong 518055, China. E-mail: maclachlan@pkusz.edu.cn. His research interests include housing development for rural migrants and talent workers, supertall buildings, and informal urban agriculture in China.Xin LiXIN LI is an Associate Professor of Urban Planning at the Department of Architecture and Civil Engineering, City University of Hong Kong, Hong Kong SAR, China. E-mail: xinli87@cityu.edu.hk. Her research interests include urban issues caused by socio-economic changes, brownfield redevelopment, land use regulations, and public housing policies in different institutional settings.