创新集群发展的“东亚模式”:基于日、韩数据的实证检验

East Asian Mode of Innovation Cluster Development:Empirical Test Based on the Data from Japan and South Korea

文章以东亚创新集群发展的代表性国家——日本、韩国的2001-2010年数据为样本,利用全局空间相关性检验—Moran I指数检验了日本、韩国地区创新集群发展的基本形态及演变特征。为了避免由于模型设定所导致的推断错误,文章利用初始GMM估计方法、完全加权GMM估计方法和部分加权GMM估计方法,构建静态空间面板模型,检验大学、国有研究机构、风险资本、政府的政策倾斜以及大企业、中小企业各自在东亚创新集群发展中的作用,以此揭示东亚创新集群的发展模式。实证检验显示,创新集群发展存在"东亚模式",其与"美国模式"的本质区别在于创新集群产生与发展的关键驱动要素不同,"东亚模式"具有"跨越式"追赶的特征,但却未能发挥大学等知识部门的潜能。我国可以通过吸纳两种创新集群的发展模式,结合本国经济发展的实际情况进而采取灵活的政策措施,使两种模式相互促进并形成良性互动。

Based on the data from Japan and South Korea as representative countries of innovation cluster development in East Asia from 2001 to 2010, this paper uses general spatial correlation test, namely Moran I index, to study the basic forms and evolution characteristics of innovation cluster development in Japan and South Korea. In order to avoid inference error resulting from model setting, it employs initial, fully weighted and partially weighted GMM and constructs static spatial panel model to test the role of universities, state-owned research institutions, venture capital, government policy preference, large finns and SMF-s in East Asian innovation cluster development respectively, to reveal development modes of East Asian innovation clusters. The empirical test shows that there exists East Asian mode of innovation cluster development whose essential difference from American mode lies in key driving factors of emergence and development of innovation clusters. East Asian mode is featured by leapfrog catching-up, but does not exert the proper potential of knowledge sectors like universities. China could absorb the two innovation cluster development modes and flexibly take policy measures according to real situation of its own economic development to make a mutual promotion and form the positive interaction between the two modes.