广州市制造业与生产性服务业协同集聚与空间相似性

基于企业微观大数据，从空间相似性的视角，运用核密度法、双变量空间自相关法和地理探测器等分析方法，对广州市6个制造行业和5个生产性服务行业的空间协同关系展开分析，在街镇尺度下从行业层面探讨生产性服务业与制造业在城市内部的协同集聚。研究发现：行业集聚中心的空间布局形态显示，广州的制造业与生产性服务业布局具有空间相似性。制造业与生产性服务业空间相似性最高的区域集中在城市的近郊区与远郊区，在城市中心城区两大产业的空间相似性较低。科技服务业与制造业的空间相似性最强，金融业与制造业的空间相似性最低。产业发展历史、行业的服务功能特性、民营企业的布局弹性等，是影响生产性服务业与制造业空间相似性的重要因素。

Co-agglomeration and Spatial Similarity:Based on the Analysis of Manufac-turing and Producer Services in Guangzhou,China

The symbiotic relationship between producer services and manufacturing leads to the co-agglomeration of two sectors. As most literature focuses on their functional connections, studies from a spatial perspective are limited. Relying on the firm-level data from tianyancha.com, this paper analyzes co-agglomeration patterns and characteristics between six manufacturing sub-sectors and five producer service sub-sectors in Guangzhou, China. We utilized kernel density, Moran’s I, and geo-detector to explore the spatial similarities between those sub-sectors and the underlying mechanisms. Kernel density was employed to visualize the collaborative agglomeration directly. Global and local bivariate Moran’s I statistics were employed to explore the spatial autocorrelation. By employing the geographic detector, we further examined the underlying mechanism of varying co-agglomeration patterns. Major findings are as follows: 1) Manufacturing in Guangzhou is more scattered relative to produce services, with several agglomeration centers in the peripheral regions. 2) The distributions of manufacturing and producer services are spatially similar. Such similarities also vary across sub-sectors. Among these six sub-sectors, technology service companies have the most similar manufacturing distribution, with a high level of polycentricity. On the other hand, financial companies have the lowest similarity to the spatial layout of manufacturing companies. 3) The high-high cluster areas of producer service and manufacturing are located near suburbs such as Panyu District and Baiyun District. The low-low cluster areas are concentrated in distant suburbs like Conghua District and Zengcheng District. The high-low and low-high clusters are located in central urban areas such as Tianhe, Yuexiu, and Haizhu districts, suggesting a spatial mismatch of producer service and manufacturing in Guangzhou’s urban center. 4) Technology service has the strongest impact on all six manufacturing sub-sectors’ spatial configurations, with the highest spatial similarity. In contrast, the cluster of financial firms is unlikely to leads to the agglomeration of manufacturing companies. This divergence is highly associated with the trajectory of the city’s industrial development, functional characteristics of service industries, and spatially varying flexibility of private enterprises.