Absorptive capacity and technological spillovers - simulations in an evolutionary framework

Already Joseph Alois Schumpeter (1952) emphasized that innovative activities are to be considered as an active search (or R&D) process for new possibilities and opportunities rather than a solely accidental event. Besides the investment of R&D-funds this search process also requires certain tasks such as screening, recombining, modifying and even consciously rejecting already existing know-how. An analytical treatment of these tasks obviously has to cover a broad range of topics such as learning, perception, decision under uncertainty, creativity, etc. The approach we follow in the proposed paper, however, is rather modest, and we want to emphasize two aspects, the ability to absorb outside knowledge and the knowledge creating effects of search procedures. Central to both of these points of investigation is the existence of technological spillovers which are of special importance in an evolutionary context. There the cumulativeness of technological progress and the boundedness of search routines imply that firms engaging in R&D develop along certain technological trajectories (Dosi (1988)). Those development paths obviously do not allow for a continous progress forever but their technological potential decreases with increasing progress. In economic reality, however, an unsteady but continuous progress is to be observed. This is to be interpreted in a way that single technologies are again and again influenced positively by their environment. An important source of such effects are the so-called technological spillovers. Those can come from other firms, but also from universities and other research institutes. Technological spillovers are of course nothing new for traditional economic theory. There they are dealt with as positive externalities which, however, reduce the entrepreneurial's incentive to engage in R&D. Therefore, from a welfare-theoretical point of view, the total level of R&D is sub-optimal. Our approach want to modify this view twofold. First, we consider technological spillovers not primarily as incentive-reducing, but as ideas-creating. Second, we want to take account of the fact that technological spillovers - contrary to some traditional approaches - cannot easily be adopted by imitators but that they have to satisfy certain prerequisites and technological capabilities. In this respect, the so-called absorptive capacities gain special importance (Cohen/Levinthal (1989)). They put a firm in the position to understand and adopt know-how generated elsewhere - or in other words, they make technological spillovers work as they are supposed to work, i.e. making knowledge created by A useful for B. Miyazaki (1994) shows that this concept of absorptive capacity is not an entirely theoretical one but it is a strategy which is actually relevant in reality, especially in high-tech industries. On this basis our paper is concerned with alternative firm strategies where the capacity to use technological spillovers plays a central role. At first we distinguish between two firms forced to invest in absorptive capacity in order to utilize spillover effects. Strategy I, labelled with ad hoc strategy invests in absorptive capacity only when the opportunities of the current technology decreased on a certain level. Contrariwise, strategy II invests continuously in absorptive capacity. In a second simulation we additionally take into account that strategy II firms may decide whether to use the beneficial effects of spillovers or not. This third strategy is called selective. Besides these basic strategies, we also take account of effects which in our view are important for the understanding of innovative processes. We consider the stochastic character of innovative activities, learning effects, and the loss or depreciation of technological know-how. With the help of simulations differently nested firm strategies are compared. As the most important result of our study we find that strategy II continuously investing in absorptive capacity is very likely to dominate a conservative strategy. This becomes even enforced when both strategies are assumed to follow an identical decision rule. Here, both agents compare the new and the old technological opportunities. And, under certain circumstances spillover effects will not be used