Starting from the failure of the R&D–patents traditional relationship, when time-series and/or within-industry dimensions are included in the empirical analysis, the present work tries to contribute to the empirical literature from two directions. First, it performs a Granger causality test based on the theoretical presumption of a reverse patents→R&D link as an explanation for the failure of the traditional relationship. Second, assuming the reverse patents–R&D causality, we test and interpret the lag structure of such a relationship which shows the effective patent life that firms can expect within the two Schumpeterian patterns of innovations they belong to. In the light of the effective patent life, we offer a further explanation of innovation persistence which overturns the findings of the existing literature on persistence.