All of the same breed? A networking perspective of private-collective innovation

Much research on open source software development has highlighted the best-of-both-worlds benefits generated by private and collective contributions from a broad base of developers. However, these studies have tended to overlook the heterogeneous nature of various developer groups including firm-sponsored developers. To unveil the behavioural differences between developer groups, we expand upon the private-collective innovation model using a networking approach, linking network closure and positional embeddedness with technical contribution. We tested our predictions using a lagged analysis based on communication and networking behaviours on the Linux kernel mailing-list in the production of the Linux operation system over a three-year period. Our findings support our predictions, showing that network closure has an adverse impact on technical contribution. Additionally, the relationship between positional embeddedness and technical contribution follows an inverted Ushape. In addition, high positional embeddedness counteracts the negative influence of extensive network closure on technical contribution. These effects are partly moderated by respective developer groups. Our model and results offer important theoretical and practical implications for community management within the framework of private-collective innovation.

Automated summary

Much research on open source software development has tended to overlook the behavioral differences between various developer groups. This study expands upon the private-collective innovation model using a networking approach and reveals the influence of network closure and positional embeddedness on technical contribution. The empirical findings show that network closure has a negative impact on technical contribution, while the relationship between positional embeddedness and technical contribution follows an inverted U-shape. Moreover, high positional embeddedness can counteract the negative influence of extensive network closure on technical contribution. These effects are moderated by different developer groups.