Innovation and Government Bureaucracy

This paper explores the nexus between innovation and bureaucratic performance where we focus on the effect of actual bureaucratic performance, rather than bureaucratic capacity. A conditional difference in differences estimation using an unbalanced panel of nations spanning the period 2004-2018, provides strong evidence that better bureaucratic performance underlies better innovation outcomes, ceteris paribus. At the median stock of knowledge capital, a one-unit improvement in bureaucratic performance raises the patent applications of a sample country by about 651, and ‘international’ patents by about 340, which constitute a 1.8% and 5.5% increase over the sample mean patent applications and sample mean ‘international’ patents, respectively. Second, this effect is heterogenous, with these responses becoming more pronounced at higher percentiles of the knowledge capital stock owned by a country. Thus, at the 95th percentile, a one-unit increase in bureaucratic performance raises patent applications by about 3.3%, and ‘international’ patents by about 6.7%, of their respective sample means. Third, the strong significance of bureaucratic performance for innovation is found to be fairly broad-based across technology groups such as Electrical/Electronics Technology, Professional and Scientific Equipment, Pharmaceuticals, Chemicals, and Machinery (Non-electrical), and is not driven by just one or two of these groups. The results are robust to several robustness checks. Key Words: Innovation, Bureaucratic performance, Nonlinear influence JEL Codes: O34, O38, O43