From Institutions to Networks to Organizational Outcomes: The Case of Open Source Innovation

That institutions and networks shape organizational outcomes is commonly accepted, but an integrative framework is still lacking. This research investigates how institutions can shape acquired (social) resources of innovative organizations, and how institutions and resources consequently shape organizational outcomes. A general theory is outlined that focuses on institutions that restrict the communication between actors across organizations, and may consequently impact the flow of innovative information.

Opposing direct and indirect effects of institutions are expected. Private investment models of innovation on the one hand suggest that restrictive institutions protect the (private) investments of organizational members against free-riding behavior, and consequently predict that restrictive institutions yield more beneficial organizational outcomes. Network models of innovation on the other hand suggest that restrictive institutions inhibit access to potentially relevant social resources, and therefore impact organizational outcomes negatively. 

Open Source Software (OSS) development communities provide an interesting case to study the aforementioned relations. In OSS development teams, software is distributed via licenses that differ in the degree to which innovative information can flow across projects. A composite dataset derived from a large repository of active Open Source Software is analyzed. This unique dataset contains organizational-level indicators of OSS project-performance, as well as indicators of social resources and licensing practices. Structural equation modeling (SEM) procedures are used to test the proposed relations, and to compare the model against alternative models. 

Results confirm the expected relations. A direct effect of licenses on outcomes was found, but also a negative indirect effect: projects with more restrictive licenses showed less brokerage value and consequently performed poorer. Subsequent exploratory analyses revealed differential effects of licensing and networks at different stages in the project's life course, and for different types of projects. This suggests that both endogenous and exogenous factors can moderate the optimal degree of restrictiveness, which warrants further research.