Abstract: Based on the micro level of analysis, the paper takes the 143 researchers in the field of "technological innovation" as the sample, and investigates the science funding’s incentive effect under the condition of randomized simulation experiment，i.e., Using the method of propensity score stratification, we compare the output differences of researchers between two states of "not being granted with science funding" and "being granted with science funding". The empirical results indicate that compared with the researchers that are not granted with the science funding, the researchers who are granted with the science funding generally have higher research output. In addition, researchers with the science funding support under higher probability usually have higher research output. However, there are no obvious research output differences between researchers who are granted with the science funding support under low probability and those who are not granted with the science funding support under medium probability. In addition, compared with the researchers who are granted with the science funding support under medium probability, researchers who are not granted with the science funding support under the highest probability have higher research output. The research conclusions have implications for science funding support policy.

摘要： 基于微观分析层次，以“技术创新”领域的143位研究者为对象，采用倾向得分分层法，在模拟随机化实验的条件下考察科学基金的激励效应，即比较研究者在“获得科学基金资助”和“未获得科学基金资助”两种情形下的科研产出差异。研究结果表明，总体上来说，与未获得科学基金资助的研究者相比，获得科学基金资助的研究者普遍具有较高的科研产出，而且，在获得科学基金资助的研究者中，以较高概率获得科学基金资助的研究者具有较高的科研产出；然而，获得科学基金资助低概率的研究者与未获得科学基金资助中等概率的研究者在科研产出方面并未表现出明显差异；与获得科学基金资助中等概率的研究者相比，未获得科学基金资助高概率的研究者具有较高的科研产出。本文的研究结论对制定科学基金资助政策具有一定的启示意义。