Förderjahr 2018 / Stipendien Call #13 / ProjektID: 3844 / Projekt: Essays on Communities
For our identification strategy we deployed a refined degree centrality measure for code‐exchange ties stipulated between OSS repository founders and other programmers. Here, by discerning the effect of the diversity of helping behaviors that a repository founder provides to and receives from other participants on his or her projects’ success, we are capable of tracking non-equivalent reciprocity among participants in general. By adopting a two-dimensional measure for outdegree centrality based on contributions that repository founders make by maintaining their own projects and those by gifting to other projects, we can show more specifically that non-equivalent reciprocity manifests itself in specialized exchanges, and as such, facilitates division of labor.
Our baseline tests using GitHub individual contribution and project-level data provide us with support for our Hypotheses 1 and 3. Meanwhile, we do not find an effect of equivalent reciprocity in the expected (from the social networks literature) direction (H2) – we find an opposite effect, suggesting that GitHub is more of a professional, goal-directed community with norms of division of labor prevailing (rather than a social network like YouTube or Twitter). This finding shows that communities (i.e. OSS projects in particular) can be efficient and thus can provide complementary assets to firms' traditional innovation activities.
The next question that then arises is - which communities can be more efficient and bring the greatest value to firms, what are the boundary conditions which define such projects, or sub-communities? Our further theoretical investigation is to test whether different regimes in norm enforcement across projects within the GitHub community explain differences in projects' efficiency levels (and ultimately amount of output being produced). The expectation is that the projects with higher norm enforcement, would have the baseline effects being exacerbated. Currently, we are testing whether the effects we obtained would differ based on the number of watchers a project has or difference in tenures among project members (i.e. proxy for internal cohesion within a project team). The analysis is ongoing.