Managing internal embeddedness in multinational corporations’ R&D subsidiaries: An evolutionary perspective on the automotive industry in Silicon Valley

This paper analyses the processes through which multinational corporations' R&D subsidiaries conduct dual embeddedness from an internal embeddedness approach. While most studies that adopt a dual embeddedness perspective have analysed the influence of internal embeddedness on external embeddedness as a single effect (positive or negative), we offer a different stance by considering it through a double-edged effect. We conduct a multiple case study of R&D subsidiaries located in Silicon Valley and belonging to eight of the world's largest automotive manufacturing multinational corporations. By adopting an evolutionary perspective based on an analysis of a 23-year period (1995–2018), we find and unravel how to manage the double-edged effect of R&D subsidiaries' internal embeddedness on the development of external embeddedness by specifying two main levers: the autonomization of R&D subsidiaries and the headquarters' knowledge support of the subsidiaries. Our findings thus advance the understanding of R&D subsidiaries' dual embeddedness and external knowledge sourcing in technological contexts by proposing a model of the double influence of internal embeddedness on external embeddedness. © 2021 Elsevier Ltd