Based on the panel data of Chinese A-share listed manufacturing companies from 2009 to 2015, stochastic frontier model was applied to analyze the relationship between government R&D subsidies and enterprises' innovation efficiency. Unlike the previous studies, this paper finds that government R&D subsidies and enterprises' innovation efficiency follow a double inflection point inverted-U correlation and reveals the economic implications. Using marginal analysis method, the paper analyzes the marginal effect of incremental subsidies rather than the total effect of stocked subsidies, holding that the (marginal) total effect of subsidies is the superposition of (marginal) incentive effect and (marginal) crowding-out effect. In addition, the paper measures the incentive effect and crowding-out effect and judges the role of government R&D subsidies according to their change. To conclude, the government R&D subsidies are nearly inefficient. Further, we analyze the impact of environmental factors, protection of intellectual property, factor market and infrastructure, and find that factor market and infrastructure have a great impact on enterprises' innovation efficiency. Therefore, the paper gives some suggestions to improve the effectiveness of subsidy policy and the innovation efficiency of enterprises.%选取2009-2015年中国A股上市制造业工业企业数据,通过构建随机前沿模型对政府R&D补贴与企业创新效率的关系进行了实证研究.和已有研究不同,本文发现政府R&D补贴与企业创新效率呈双拐点倒"U"关系,并解释了双拐点倒"U"关系的经济含义;应用边际分析方法,考察政府R&D补贴增量的边际效应而非补贴存量的总效应,认为补贴的(边际)总效应是补贴的(边际)激励效应和补贴的(边际)挤出效应相叠加的结果;同时,测度了激励效应和挤出效应的大小,根据其增减变化,判断政府补贴的作用状态,发现政府创新补贴基本是无效率的;进一步地,实证分析了环境因素——知识产权保护、要素市场与基础设施对企业创新效率的影响,发现要素市场与基础设施这两大因素对企业创新效率的影响巨大.在此基础上,本文对增强政府创新补贴政策的有效性、提升企业创新效率提出了政策建议.
展开▼
