Abstract: Most researches on R&D subsidies focus on the evaluation of the effect of subsidies,but the research on the allocation of R&D subsidies is insufficient.We can observe the Matthew effect in the distribution of R&D subsidies which means some companies take away most subsidies while others never get once. This paper analyzed the causes of Matthew effect,including Pick the winer system,Learning accumulation effect,Reputation advantage,Interest group capture and Government audit factors.We divided Matthew effect into vicious Matthew effect and virtuous Matthew effect by different causes.The empirical results show that there was obvious Matthew effect in the allocation of R&D subsidies and the enterprise economic indicators and R&D indicators had significant positive effects on persistence of subsidy, which indicated that the Matthew effect of R&D Subsidy in China is virtuous Matthew effect.We tested the influence of Matthew effect on enterprise R&D by PSM,finding that virtuous Matthew effect had a positive influence on Enterprise R&D activities.

摘要： 当前企业R&D补贴研究大多关注补贴对企业R&D活动的效果评估，对R&D补贴的分配问题研究较为不足。现实中R&D补贴分配存在明显的马太效应：曾经获得补贴的企业更容易获得后续补贴，以前未能获得补贴的企业未来也较难获得补贴。本文分析了马太效应的内在机理，提出了形成马太效应的五大根源：竞争择优机制、学习积累效应、纯粹声誉优势、利益集团捕获及审计反腐效应，并根据成因的不同将其分为“良性的马太效应”和“恶性的马太效应”。通过转移概率矩阵法考察了企业R&D补贴分配中的持续效应，证实了企业R&D补贴分配存在明显的马太效应；通过Probit模型分析了造成马太效应的关键变量，发现企业经济指标、R&D指标对企业持续获得补贴存在显著正影响，说明我国企业R&D补贴存在的马太效应为良性马太效应；通过倾向得分匹配模型实证考察了马太效应对企业R&D活动的影响，发现马太效应对企业R&D投入、专利申请、新产品销售收入方面均为正向影响。