When does distributed innovation activity make sense? Location, decentralization, and innovation success

Companies face an expanding set of choices about where to locate their innovation activity, both within their home countries and abroad. This location choice also requires firms to make a simultaneous choice about the organizational structure of innovation activity: almost by definition, multiple locations per firm imply some degree of decentralization. Using firm-level data on innovation output and the location of research and development (R&D) activity, we shed new light on the question of whether firms that have multiple locations also have greater innovation success. Our results indicate that, on average, having distributed R&D activity is beneficial in terms of the extent and breadth of innovation success, and the effect is strongly related to the knowledge sourcing strategies that firms employ. These results are consistent with the interpretation that R&D location decisions are driven by the desire of firms to access a broad set of external sources of knowledge for innovation activities. We also find that the benefits of multiple R&D locations do not apply to novel (new-to-the-market) innovations. Our results suggest that when analyzing technological innovation, it is important to distinguish between novel and imitative innovations, since their determinants may differ.