Knowledge complementarity and the volatility of relationships in the Dutch life sciences industry network seen from two perspectives

The life sciences industry is considered as very important for stimulating future economic development. In order to let this industry flourish and grow, collaborations between organizations in this industry are very important for the survival of organizations and will be considered in this study. This is due to the diverse specialties of the organizations within this industry and the rapid development of techniques, skills and resources. Organizations are thus forced to collaborate with other organizations. For one organization it is impossible to have all the competences and resources needed to develop a new product in-house. So, organizations are forced to collaborate in order to acquire technological knowledge complementarities. The organizations themselves also indicate that this is one of the most important motives for organizations to collaborate. The life sciences industry is a knowledge intensive industry and it takes a long time to develop a product. So, in order to innovate long term inter-organizational collaborations are expected. However, within the Dutch life sciences industry a highly volatile network was discovered. This resulted into the following research question: To what extent does knowledge complementarity influence the volatility of relationships in the Dutch life sciences industry network from 2002-2005 from both a relational and organizational perspective? This research question will be answered with the help of the Resource Based View (RBV) and the Resource Dependence View. Two perspectives are taken regarding this question indicating two different operationalizations of the concept knowledge complementarity. In the relational perspective knowledge complementarity will be operationalized as patent citations and within the organizational perspective a more Resource Based View is taken where knowledge complementarity is operationalized as the number of patents which an organization owns or has access to. For the relational operationalization no analysis could be done. Within the network there were so little patent citations between collaborating partners that a decent analysis was impossible. In order to give at least an indication of knowledge complementarity on a relational level a patent class analysis has been conducted. However, it turned out that only 40% of the organizations within the life sciences industry in the Netherlands own one or more patents. Again no analysis could be done. At the organizational level first more insight needs to be gained into the dynamics of the Dutch life sciences network. It turned out that the number of relationships has a two wave influence on the number of aborted and newly formed relationships. Organizations with a lot of relationships abort the less beneficial ones. This behavior continues over time. Organizations with only a few relationships will form new relationships and then again will form new relationships, but from the set of previously formed relationships they will abort the least beneficial ones. So it seems that having a few relationships is an incentive to form new ones. These two patterns are continued over time, however the effects deteriorate over time. After that, different RBV variables (the amount of acquired venture capital, R&D intensity, size, age, type and the total number of relationships) and knowledge complementarity have been added to the model. Then it turns out that knowledge complementarity indeed has a negative influence on the volatility of the relationships in the network. However, it turned out that it also had a positive influence (i.e. because of the subdivision of the volatility into the number of aborted and newly formed relationships). This seems to be the effect of knowledge leaking or the fact that expected knowledge complementarities were not really present. Knowledge leaking happens when one of the two collaboration partners also has relations with other organizations. The exchanged knowledge can then more easily leak to third parties. This can lead to network effects like shorter collaborations despite the fact that the organizations have a high rate of knowledge complementarity. Next to this some other conclusions can be drawn from this analysis. The explanatory variables are very time dependent and the effects of the explanatory variables are also volatile over time. It turned out also that the earlier found result that organizations with less relationships first form new ones and then select and abort the least beneficial, is due to the fact that their resources are changing over time which could be deduced from the volatility of the explanatory variables over time. From the time dependency of the explanatory variables it can be concluded that the needs of organizations within the industry change over time. These results have some implications for those obtained in other studies and give an important suggestion for further research. In this study it is shown that the effects from Resource Based View variables are time dependent and volatile over time. This demands a more dynamic model and explanation since there seems to be no sound structural Resource Based View model. Furthermore, it would be interesting to link this more dynamic model to the phase of product development and the phase of industry/network development