| Abstract: | Product innovation is the result of a constant interaction between the in‐house research and development (R&D) department and knowledge exchanges with the firm's environment. Knowledge exchanges come in different forms. They break down into information gathering applied in new product development, research cooperation on particular innovation projects, and managing information outflows allowing the consequent appropriation of the results of product innovation through specific methods. The way firms handle knowledge exchanges affects their performance. This paper looks at three related indicators of performance: (1) research intensity (a measure of innovative input); (2) the share of revenue realized through innovative product sales (a measure of innovative output); and (3) their impact on the growth in total revenue. The bulk of the econometric literature looking into these matters only allows general statistical statements on the behavior of an “average” firm. This paper takes on another view by using the quantile regression method to stress the heterogeneity of innovative firms in their dealing with knowledge exchange and the effect this has on their performance. A first key finding is that research intensity is positively influenced by knowledge externalities, research cooperation, and appropriability, and it is through this that these variables affect innovative revenue and also the growth in total revenue. By using quantile regression these relationships are further refined to screen for differences in behavior between dynamic and lagging innovators. This refinement indicates that, in the case of research intensity, the knowledge externalities gain in importance in the higher quantiles and are insignificant in the lower ones. Next, research cooperation remains important in all quantiles, but a higher significance is observed in the higher quantiles as well. Finally, appropriability is extremely important for the lower quantiles, but it becomes insignificant in the highest. These findings corroborate the assumptions made in the literature on open innovation: knowledge externalities and research collaboration are vital for those opening up their firm for new ideas and who are, at the same time, reluctant to protect their findings through specific appropriation measures. In the case of innovative revenue all variables on knowledge exchange operate through the research intensity irrespective of the quantile, although the impact of research intensity on this type of revenue is higher in the upper quantiles. As for the growth in revenue, the effect of the innovative revenue is, again, higher in the higher quantiles. This suggests that dynamic product innovators have the most efficient R&D process and the strongest growers are so, especially, because they are successful product innovators. |
