An Empirical Test of the Asymmetric Models on Innovative Activity : Who Invests More into R&D, the Incumbent or the Challenger?

We study whether the incumbent or the (potential) entrant invests more into R&D, using data of about 3500 German firms from 1992 to 1995 that explicitly asks firms for their motives in undertaking innovative activity, making it possible to take account intended, not just completed, market entry. We find that challengers invest more into R&D to enter a new market than the incumbents, making the patent racing model by Reinganum and others seem more accurate than the auction model of Gilbert and Newbery. We estimate a tobit model with selectivity to take account of the econometric problem of double censoring