This research empirically tests the Popper hypothesis (2012) by utilizing panel data that cover 132 countries over the time period 1980–2017 and via panel GMM estimation in order to investigate the causal link from democracy to innovation, which is captured by two comprehensive variables, patent application and trademark application. Aside from the relationship between democracy and innovation, we further explore whether or not the change of democracy significantly influences innovation, whether the influence of democracy on innovation varies among different kinds of democracy or autocracy, and hence pay more attention on factors such as economic development level, economic performance, globalization, and international trade, and how do they change the democracy's impact on technical innovation. To confirm credibility, we carry out the robustness test by changing the measurement of democracy, the measurement of innovation and deleting extreme values. In summary, while democracy promotes innovation performance, autocracy generally exhibits lower technological innovation performance. This study provides evidence on the influence of democracy on innovation being affected by other factors such as those stated above.

本研究通过利用 1980 年至 2017 年期间涵盖 132 个国家的面板数据并通过面板 GMM 估计对波普尔假设（2012 年）进行实证检验，以研究从民主到创新的因果关系，该因果关系由两个综合变量捕获，专利申请和商标申请。除了民主与创新的关系，我们进一步探讨民主的变化是否显着影响创新，民主对创新的影响是否因民主或专制而异，从而更加关注经济发展等因素。水平、经济表现、全球化和国际贸易，以及它们如何改变民主对技术创新的影响。为了确认可信度，我们通过改变民主的衡量标准、创新的衡量标准和删除极值来进行稳健性检验。综上所述，虽然民主促进了创新绩效，但专制通常表现出较低的技术创新绩效。本研究提供了民主对创新的影响受到其他因素（例如上述因素）影响的证据。