科研经费均衡度度量偏差的机理分析与实证研究

Mechanism Analysis and Empirical Study on Measurement Biases of Allocation Inequality in Research Funding

现有的科研经费分配机制大多数是竞争性的,容易导致科研经费分配的不均衡现象。如何准确度量经费分配的均衡度成为科研管理领域的一个重要问题。基尼系数常用于刻画科研经费分配的均衡度,然而由于样本选择等原因,现有研究在使用基尼系数刻画科研经费均衡度时存在度量偏差。总结分析了度量偏差的2个来源,一是在研究对象层面上,现有研究多数是从机构层面来度量科研经费分配均衡程度,忽略了机构规模带来的影响;二是在统计口径上,现有的研究假设获资助科研人员即为机构科研人员的总体,科研人员基数的统计口径作了简易处理,造成了经费分配不均衡度的高估或低估。从理论层面阐述了度量偏差的形成机理和改进方法。以2011—2015年中国科学院科研人员获得资助的项目数据为例,计算了中国科学院研究所和科研人员的经费基尼系数,并对比分析了机构与个人层面、名义与实际科研人数对基尼系数产生的度量偏差。结果表明,中国科学院机构经费分配较为均衡,而个人科研经费集中程度较为明显。以机构层面的基尼系数来表示中科院科研经费分配均衡度,低估了科研经费在科研人员间分配的不均衡程度;而传统研究中仅研究获资助人员间的分配状况,严重低估了科研经费分配的不均衡程度。

The inequality of research funding does exist, and the quantitative measure of the inequality of funding allocation is the primary task when we discuss the inequality in science. Previous researches have tried to use the Gini coefficient to measure the inequality, but they mainly focused on the institutional-（regional-） level due to the limitation of data availability. A few of researches have studied into the inequality of funding allocation at the in- dividual level, while the total number of the scientific researchers is calculated in a simple way. This paper con- cludes two measurement biases in fimding inequality by Gini coefficient and provides a more accurate result of the distribution inequality of the research funding. In the empirical studies, the funding data of Chinese Academy of Sciences （CAS） from 2011 to 2015 are abstracted from Academic Resource Planning （ARP） system. The em- pirical results show that the Gini coefficient of CAS is relatively equal yet the Gini coefficient in individual level is high inequality. Furthermore, it is highly underestimated if ignoring the researchers who have not been funded.