基于时变参数模型的中国研发投入回报率估算

发布时间：2018-10-31 21:09

【摘要】：研发投入回报率的估算对于科技政策指定具有重要意义。传统的研发回报率的估算一般基于恒定参数模型,恒定参数的假设不符合现实。文章基于傅里叶变换处理时变参数并扩展了Jones and Williams(1998)的研发回报率估算框架。使用我国1995—2013年的省际面板数据,估算我国的研发回报率。研究发现:生产部门生产函数的资本、劳动弹性具有明显的时变性特征;知识部门研发投入对知识生产的贡献也具有明显的时变性特征。在2002—2013年之间,基于时变模型的估算得到我国研发投入的平均回报率为11.8%,而基于恒定参数模型的估计结果为17.1%,即基于恒定参数模型的估算每年高估研发回报率约5%。因此,基于恒定参数模型考虑研发投入的回报率和最优规模可能产生误导性结论。
[Abstract]:The estimation of R & D investment rate of return is of great significance to the designation of science and technology policy. The traditional estimation of R & D return is based on constant parameter model, and the assumption of constant parameter does not accord with reality. This paper presents a framework for R & D rate of return estimation based on Fourier transform to deal with time-varying parameters and extend Jones and Williams (1998. Using the provincial panel data from 1995 to 2013, we estimate the R & D rate of our country. It is found that the capital and labor elasticity of the production function of the production department have obvious time-varying characteristics, and the contribution of the R & D investment of the knowledge department to the knowledge production is also obviously time-varying. Between 2002 and 2013, the average rate of return on R & D investment in China was estimated to be 11.8% based on the time-varying model, while the estimate based on the constant parameter model was 17.1%. The estimate based on constant parameter model overestimates R & D returns by about 5% a year. Therefore, considering the rate of return and optimal scale of R & D investment based on constant parameter model may lead to misleading conclusions.
【作者单位】：兰州大学管理学院;西北民族大学经济学院;
【基金】：中央高校基本科研业务费专项资金(15LZUJBWZY097;15LZUJBWZY118)
【分类号】：F224;F124.3

【相似文献】