摘要
本文以马歇尔集聚理论为基础构建空间外部性指标和计量模型,运用2003-2011年城市面板数据检验了经济活动空间集聚对能源效率的影响。结果显示,城市能源效率受到空间中城市专业化劳动力、中间投入和技术溢出效应的影响,其中,中间投入空间可得性和空间技术外溢对能源效率具有显著促进作用,而邻区专业化劳动力却不利于当地能源效率的提高。从分地区的估计结果来看,空间集聚外部性对各地区能源效率的综合效应均为正,但各地区不同集聚外部性对能源效率的作用渠道和影响效果各异。进一步通过考察空间集聚外部性对能源效率不同分位点的边际效应及其变化趋势发现,专业化劳动力可得性、中间投入可得性和区际沟通的技术溢出对能源效率的边际贡献随分位数增加呈现先增后降的倒U型变化趋势,而区际研发的技术溢出则呈现U型变化趋势。
Based on the theory of Marshallian agglomeration externalities, this paper analyzes the effects of spatial externalities on urban industrial energy efficiency. The results find that, urban energy efficiency is not only related to the agglomeration factors of index city, but also affected by specialized labor force, intermediate inputs and technological spillovers of neighboring cities. Where spatial availability of intermediate inputs and spatial technological spillovers significantly improve the urban energy efficiency, while the specialized labor force of neighboring cities plays an negative role on energy efficiency. From the estimates by region, we can see that the combined effects of spatial agglomeration externalities on energy efficiency of each region are positive, but different types of agglomeration externalities of each region play different roles on regional energy efficiency in terms of action channels and impact effects. After examining the marginal effects of spatial agglomeration externalities on energy efficiency of different quartiles and their trends, we further find that the marginal contributions of availability of specialized labor force, intermediate inputs and technology spillovers of inter-regional interpersonal communication have an inverted U-shaped trend with quantile increases, while that of technology spillovers of inter-regional R&D has an U-shaped trend.
出处
《中国人口·资源与环境》
CSSCI
北大核心
2014年第5期72-79,共8页
China Population,Resources and Environment
基金
湖南省国际经济与国际工程管理研究中心基金项目"开放经济下生产性服务业集聚对城镇化的影响机制与实证研究"(编号:13IEPMZ1)
国家社会科学基金项目"劳动力供给变化影响制造业升级的机理及政策研究"(编号:13CJL058)
关键词
空间外部性
集聚经济
工业能源效率
分位数回归
Spatial Externalities
Economic Agglomeration
Energy Efficiency
Panel Data FGLS
Quantile Regression