Travail de recherche/Working paper
| Résumé : | This paper tackles one of the most persistent criticism of patent statistics. Because not all inventions are patented, the patent-to-R&D ratio reflects both a productivity effect (the number of inventions created per unit of research input) and a propensity effect (the proportion of inventions patented). We propose a solution to this identification problem. Our methodology uses information on the density of patent value and leads to results that are easy to interpret. It is applied to a novel data set of priority patent applications in which each patent is fractionally allocated to its inventors’ countries and to the technological areas to which it belongs. Interestingly, it is frequently observed that an industry may exhibit a low number of patents per unit of R&D in one country yet actually be more productive than the same industry in another country where the patentto-R&D ratio is higher. |
