Working Paper: CEPR ID: DP6716
Authors: Gatan de Rassenfosse; Bruno van Pottelsberghe de la Potterie
Abstract: This paper investigates whether patent counts can be taken as indicators of macroeconomic innovation performance. The empirical model explicitly accounts for the two components of patenting output: research productivity and patent propensity. The empirical analysis aims at explaining the `correct' number of priority filings in 34 countries. It confirms that the two components play a substantial role as witnessed by the impact of the design of several policies, namely education, intellectual property and science and technology policies. A major policy implication relates to the design of patent systems, which ultimately induces, or allows for, aggressive patenting strategies.
Keywords: education policy; patent policy; propensity to patent; R&D productivity; science and technology policy
JEL Codes: O30; O38
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Education policies (I28) | Research productivity (O47) |
| Research productivity (O47) | Number of patents per researcher (O34) |
| Education policies (I28) | Number of patents per researcher (O34) |
| Intellectual property policies (O34) | Patenting practices (O34) |
| Patenting practices (O34) | Number of priority filings (G33) |
| Higher administrative fees (K23) | Number of patents filed (O34) |
| Technological specialization (O39) | Patenting practices (O34) |