地理研究 ›› 2011, Vol. 30 ›› Issue (9): 1592-1605.doi: 10.11821/yj2011090005

• 经济与区域发展 • 上一篇    下一篇

地理与认知邻近对高技术产业集群创新影响——以我国软件产业集群为典型案例

李琳, 韩宝龙   

  1. 湖南大学经济与贸易学院,长沙 410079
  • 收稿日期:2011-03-02 修回日期:2011-06-01 出版日期:2011-09-20 发布日期:2011-09-20
  • 作者简介:李琳(1965-)女,湖南涟源人,教授,博士,主要从事产业集群与区域创新研究。 E-mail: lilin963@vip.sina.com 韩宝龙(1986-)男,河南信阳人,硕士生,专业方向为区域经济与产业集群。 E-mail: baronhan@foxmail.com
  • 基金资助:

    教育部人文社会科学规划课题(08JA790038);湖南省社科基金课题(2010YBA049)

An empirical research on how geographic proximity and cognitive proximity work on the innovation performance of high-tech industrial cluster

LI Lin, HAN Bao-long   

  1. College of Economics and Trade, Hunan University, Changsha 410079, China
  • Received:2011-03-02 Revised:2011-06-01 Online:2011-09-20 Published:2011-09-20

摘要: 多维邻近性是近些年国际学术界在区域创新及产业集群方向新的研究视角。本文首先从多维邻近视角出发探讨了地理邻近、认知邻近对高技术产业集群创新的影响机制,并据此提出4个待验假设;进而以我国国家级软件产业园产业集群为典型案例进行实证分析,并创造性地使用人工神经网络为前导的OLS回归分析方法对待验假设进行双重递进检验。实证结果显示:在高技术产业集群的发展和成熟阶段,地理邻近对集群创新绩效产生负的影响,但负影响递减;认知邻近对集群创新绩效产生正影响;集群外部知识的获取有利于集群创新绩效提升;集群直接创新投入也促进创新绩效的提高,但边际报酬递减。

关键词: 高技术产业集群, 地理邻近, 认知邻近, 创新绩效, 人工神经网络

Abstract: With the rise of knowledge-based economy, high-tech industry clusters and their ability to innovate become the key reason for regional development. According to the tacit knowledge and knowledge spillover theory proposed, more and more academia begin to pay attention to the new view of Dimensions of Proximity in order to explore the essential factors to the innovation of high-tech industry cluster. From the view of Dimensions of Proximity, this article analyzes how those proximities usually work on the innovation of high-tech industrial cluster, and serve a theoretical mechanism for this process. After a discussion on the mechanism between high-tech industry cluster innovation and proximities, this article proposes four hypotheses of the relationships between the geographical proximity, cognitive proximity and cluster innovation, and each of these relations is transformed into mathematical formula expression. Based on the data of national software industrial parks of China in recent five years, two methods are used in the empirical tests: artificial neural network and ordinary least squares (OLS). According to the comparison among theoretical mechanism and two empirical analysis results, this paper finally draws four conclusions as follows. Firstly, during the development period and mature period of high-tech industrial cluster development, geographic proximity has a positive influence on innovation performance of high-tech industrial cluster, but the marginal effects for this are decreasing with the development of cluster. Secondly, cognitive proximity has an active influence on cluster innovation. Moreover, the learning on external knowledge can promote the increase of innovation very much. Fourthly, the direct investment on research and development can enforce the capacity of innovation, but the marginal return for this is decreasing.

Key words: high-tech industrial cluster, geographic proximity, cognitive proximity, innovation performance, artificial neural network