This paper re-visits the empirical failure to establish a clear link between R&D efforts and patent counts at the industry level. It is claimed that the "propensity-to-patent" concept should be split into an "appropriability propensity" and a "strategic propensity". The empirical contribution is...