为了探讨广深科技创新走廊与美国硅谷的智能效率差距,采用两地的上市公司数据,运用DEA-Malmquist模型和固定效应回归分析法测量其智能效率,并实证其智能效率的来源和影响因素,发现：考察期内两地的智能效率、技术进步、纯技术效率和规模效率均得到优化。硅谷的智能效率主要来源于技术进步；走廊的智能效率主要来源于纯技术效率改善。影响因素对两地智能效率的影响迥异；相对硅谷,影响因素对走廊的影响更为强烈；他们主要通过对规模效率和纯技术效率来影响硅谷的智能效率,而通过技术进步和规模效率来影响走廊的智能效率。

In order to explore the intelligence efficiency gap between Guangzhou-Shenzhen Science and Technology Innovation Corridor and the US Silicon Valley, intelligence efficiency of the Corridor and the Valley are measured and the sources and influencing factors of the intelligence efficiency of the Corridor and the Valley are tested by using the listing companies’ data and adopting the DEA-Malmquist model and fixed effect regression analysis. It is found that the intelligence efficiency, technological progress, pure technical efficiency and scale efficiency of the Valley and Corridor have been optimized. The former"s intelligence efficiency mainly comes from its technology progress change, while the latter"s mainly comes from pure technology efficiency, the influencing factors on the intelligence efficiency of the Valley and Corridor are quite different; their influences on the Corridor is stronger than the Valley; they mainly influence the intelligence efficiency of the Valley through scale efficiency and pure technology efficiency, while they influence the intelligence efficiency of Corridor through technological progress and scale efficiency.

F 061.3

本文得到国家留学基金资助(2018),教育部人文社会科学研究一般项目“破解中国工业增加值率的“递减之谜”：成因、机理与对策研究”(项目号待补充)、广东省普通高校特色创新类项目“大广海湾海洋产业调查及发展路径研究”(批准号：2016WTSCX114)