I develop a model of an R&D race with knowledge accumulation. My model does not inherit the memorylessness property of the exponential distribution that troubles existing models of R&D races. Hence, firms' knowledge stocks are no longer irrelevant to their behavior during the R&D race, and knowledge accumulation has strategic implications. In this more general setting, I obtain results that stand in marked contrast to the previous literature. In particular, under some conditions, the firm that is behind in the race engages in catch-up behavior. This pattern of strategic interactions (action-reaction) is consistent with empirical research. Copyright 2003 by the RAND Corporation.
