The purpose of this research is twofold: First, it aims to clarify the public subsidy effect on Small–Medium Enterprises’ (SMEs’) R&D expenditures by contrasting total R&D investments and net R&D investments excluding the amount of public funding. The second aim is to determine whether government subsidies induce firms to nonlinearly invest their own funds in R&D. To control for potential selection bias, we estimate the policy effect through matching techniques and a dose–response function with a generalized propensity score, using panel data that link the 2015 through 2019 waves of the South Korean Survey on the Technology of SMEs. We were not able to reject the hypothesis that uses a linear effect of public subsidies on SMEs’ R&D investments, suggesting a complementary effect on total R&D expenditures and a partial crowding-out effect for the net R&D case without a specific minimum threshold or a saturation point. Further investigation of a possible heterogeneous effect based on a subsample shows a U-shaped relationship, with a minimum threshold in the manufacturing industry. Our findings imply that public financial support for SMEs in South Korea has not completely prevented market failure beyond a certain level of public intervention. Importantly, the absolute amount of direct public support for individual SMEs should be increased in order to efficiently promote R&D spending in the manufacturing sector.

本研究的目的有两个：首先，通过对比研发总投资和不包括公共资金数额的研发净投资，阐明公共补贴对中小企业研发支出的影响。第二个目的是确定政府补贴是否会导致企业将自己的资金非线性地投入到研发中。为了控制潜在的选择偏差，我们通过匹配技术和具有广义倾向评分的剂量反应函数来估计政策效果，使用面板数据将韩国中小企业技术调查的 2015 年至 2019 年的浪潮联系起来。我们无法拒绝使用公共补贴对中小企业研发投资的线性效应的假设，这表明对总研发投资的互补效应 D 支出和净 R&D 情况下的部分挤出效应，没有特定的最低阈值或饱和点。基于子样本的可能异质效应的进一步调查显示 U 型关系，在制造业中具有最低阈值。我们的研究结果表明，除了一定程度的公共干预外，对韩国中小企业的公共财政支持并没有完全防止市场失灵。重要的是，应增加对个别中小企业的直接公共支持的绝对数量，以有效促进制造业的研发支出。具有制造业的最低门槛。我们的研究结果表明，除了一定程度的公共干预外，对韩国中小企业的公共财政支持并没有完全防止市场失灵。重要的是，应增加对个别中小企业的直接公共支持的绝对数量，以有效促进制造业的研发支出。具有制造业的最低门槛。我们的研究结果表明，除了一定程度的公共干预外，对韩国中小企业的公共财政支持并没有完全防止市场失灵。重要的是，应增加对个别中小企业的直接公共支持的绝对数量，以有效促进制造业的研发支出。