科学学研究 ›› 2020, Vol. 38 ›› Issue (8): 1473-1480.

• 技术创新与制度创新 • 上一篇    下一篇

城市创新能力的空间分布及其影响因素研究

范柏乃1,吴晓彤1,2,李旭桦1,3   

  1. 1. 浙江大学公共管理学院
    2.
    3. 浙江大学控制科学与工程学院
  • 收稿日期:2019-07-26 修回日期:2020-02-26 出版日期:2020-08-15 发布日期:2020-08-15
  • 通讯作者: 范柏乃

Study on Spatial Distribution of Urban Innovation Capacity and Its Influencing Factors

  • Received:2019-07-26 Revised:2020-02-26 Online:2020-08-15 Published:2020-08-15

摘要: 本文利用中国285个地级及以上城市2016年的专利申请授权量数据及相关经济数据,采用探索性空间数据分析方法(ESDA)对城市创新能力的空间集聚进行全局和局部的自相关性检验。继而采用证实性空间数据分析方法(CSDA),引入空间滞后模型(SLM)空间误差模型(SEM),与普通最小二乘法模型(OLS)比较检验,对我国城市创新能力的影响因素进行了空间计量分析。研究表明, 我国城市创新能力的空间分布在全局和局部都存在显著的自相关性。空间计量模型相较于普通最小二乘法(OLS)模型估计具有更高的准确性和科学性,城市创新产生了明显的邻近空间溢出效应。制度因素、对外开放度、经济基础、研发投入和信息化程度均通过了显著性检验,其中制度因素、经济基础、研发投入和信息化程度表现为正向影响,对外开放度表现出负向影响。

关键词: 城市创新能力, 影响因素, 空间分布, 空间计量分析

Abstract: This paper uses the patent application authorization data and related economic data of 285 prefecture-level and above cities in China in 2016. A global and local spatial autocorrelation test of urban innovation output agglomeration was conducted using the Exploratory Spatial Data Analysis method (ESDA). According to the test results, the Confirmatory Spatial Data Analysis method (CSDA) was adopted. The Spatial Lag Model (SLM), Spatial Error Model (SEM) and Ordinary Least Squares model (OLS) were compared and tested, and the influencing factors of China's urban innovation capacity were spatially analyzed. The research results show that the layout of China's urban innovation output not only has overall spatial autocorrelation, but also has local spatial correlation. It is found that the spatial econometric model is more accurate and scientific than the Ordinary Least Squares (OLS) model estimation. Urban innovation has a significant spillover effect on neighboring space. The analysis of the influencing factors shows that the institutional factors, economic basis, R&D investment, and informationization degree have positive effects. The degree of opening to the outside world shows a negative effect.