城市流动人口居住自选择中的空间权衡分析——以成都市为例

Space trade-offs analysis in the urban floating population residential self-selection: A case study of Chengdu

基于成都市主城区各街道的流动人口数据,分析2010-2015年流动人口规模的空间分布格局及居住空间分异程度,并从流动人口行为决策的视角选取影响变量,分析其对居住自选择的影响程度及其空间差异,据此揭示居住自选择中的空间权衡过程,探讨流动人口的空间权衡对其居住自选择和居住空间格局形成的作用。结果表明:2010-2015年,成都流动人口在主城区南部和城市中心快速增加,其集聚态势为西高东低;流动人口相较于本地户籍人口表现出一定程度的居住空间分异性;流动人口占常住人口比、居住区面积、房租、公交、企业及生活服务设施对流动人口居住自选择有明显影响,且流动人口占常住人口比、居住区面积和房租影响显著,但影响关系受流动人口空间自相关影响显著;地理加权回归结果显示不同变量对居住自选择的解释能力存在空间差异性,流动人口通过不同变量空间分布的差异性权衡不同区域,以此完成居住自选择并最终形成居住空间格局。

A large floating population has entered urban areas under the rapid urbanization in China. However, their residential space pattern is strongly affected by residential self-selection,which has reconstructed the urban population distribution pattern and social space. This study examines urban floating population residential space pattern and its formation process, by using the floating population data of 2010 and 2015 in the yearbooks of Chengdu downtown block,and choosing influence variables from the perspective of behavioral decision made by the urban floating population. Therefore, the aims of this study are to analyze the influence of urban characteristic variables for residential self-selection and the influences’ spatial differentiation, and to reveal process of space trade-offs in residential self-selection and its role in the formation of urban floating population residential space pattern. The results show that from 2010 to 2015, the urban floating population of Chengdu increased rapidly in the southern part of the downtown area and in the urban centers, and a significant space agglomeration situation featured by low-west and high-east is obviously reflected. Moreover, residential space pattern of urban floating population shows that the level of residential segregation is partially related to the residential space pattern of the registered population, but a notable degree of residential segregation has reduced from 2010 to 2015. More importantly, the study proves that the proportion of urban floating population in permanent residential population, residential land area, housing rent, public transportation, enterprise, hospital, drugstore, restaurant and marketplace have influences on floating population residential self-selection. In addition, the proportion of urban floating population in permanent residential population, residential land area and housing rent are the three significant variables in the spatial error model, but it is dramatically impacted by spatial autocorrelation of floating population statistic block. A major contribution of this study is that the spatial differentiation of the variables’ influence on residential self-selection is verified by using geographic weighting regression(GWR), and it reveals the process of urban floating population space trade-offs on how to realize residential self-selection by weighting spatial variability of variables’ influence among different urban regions. That is to say, the spatial difference of living cost, employment opportunities, living environment and commuting costs have shaped the floating population residential space pattern, which is a complex reflection of the urban spatial perception, spatial trade-offs and spatial self-selection of floating population. It can help us to deeply understand the formation process of urban floating population residential space pattern, and provide references to promote community integration and urban management.