Abstract: Based on the dynamic panel data for 31 provinces in China between 2000 and 2018, this paper examines the heterogeneous impacts of population ageing on technological innovation. We apply the system GMM method and reveal a humpshaped relationship between population ageing and innovation. Moreover, as compared with incremental innovation, population ageing poses a more noticeable negative impact on radical innovation at an earlier stage. We further employ the dynamic threshold regression to explore the nonlinear relationship between ageing and technological innovation. The results confirm the threshold effect of urbanization. In other words, once urbanization is below 86%, population ageing inhibits innovation.Otherwise, it could promote innovation. Therefore, the urbanization process could efficiently attenuate the discouraging effect of ageing on innovation.

Keywords: Population Ageing, Innovation, System GMM, Threshold Regression

摘要： 采用2000~2018年中国31个省份的动态面板数据，系统考察人口老龄化对科技创新的影响以及影响的异质性。运用系统GMM方法发现，人口老龄化与科技创新、根本性创新及渐进性创新之间均呈驼峰型关系，即老龄化先促进后抑制创新水平；相较于渐进性创新，人口老龄化对根本性创新的负面效应将更早显现。进一步运用动态面板门槛模型，揭示了人口老龄化与科技创新之间的关系存在城镇化率门槛效应，城镇化率的门槛值约为86%，即当城镇化率低于该门槛值时，人口老龄化将抑制创新；而当城镇化率高于门槛值时，老龄化转而有利于促进创新。由此可见，以人为核心的新型城镇化将有助于在人口老龄化的背景下推进创新驱动发展战略，推动中国经济行稳致远。

关键词: 人口老龄化, 创新, 系统GMM, 门槛模型