摘要:In this paper we reexamine the relationship between patents and R&D using empirical likelihood estimation. Based on the data of Hall, Griliches, and Hausman (1986) and the specification allowing for endogenous regressors, we found that the contemporaneous effect of R&D is significantly positive, yet the first-lag effect is significantly negative. Moreover, the total effect of R&D is much larger than those found in the early studies.