Abstract: Based on firm level data in Shanghai from year 2008 to 2014, and using Propensity Score Matching method to analyze whether government R&D subsidies can promote the state-owned enterprises and private enterprises of innovation and technological progress, we found that: As for SOEs, R&D subsidies can neither significantly affect the number of patent applications nor significantly affect their total factor productivity; As for private enterprises, R&D subsidies can significantly increase the number of patent applications, but also significantly reduce their total factor productivity. Those reflect that SOEs lack the incentive and have low innovation efficiency, and information asymmetry between the government and firms causes principal-agent problems. Therefore, the government should establish a scientific assessment mechanism and reasonable incentive mechanism in R&D subsidy policy, while to further promote the professional manager system and improve the market mechanism is needed for SOEs.

摘要： 本文基于上海市2008-2014年的企业数据，使用倾向得分匹配法对政府R&D资助是否促进了国有企业及民营企业的创新产出和技术进步这一问题进行实证分析。研究发现：R&D资助既没有显著影响国有企业的专利申请数量，也没有显著影响其全要素生产率；R&D资助显著增加了民营企业的专利申请数量，但是也显著地降低了其全要素生产率。这反映出了缺乏激励的国有企业创新效率低下的现状，以及政府与企业之间的信息不对称等导致的委托代理问题。因此，应当在R&D资助政策中建立科学的评估机制和合理的奖惩机制，同时加快推进国有企业的职业经理人制度并健全市场化机制。