| Abstract: |
This study investigates the effects of sponsorship-based funding on
open-source software contributors' behavior. We examine how sponsorship
impacts contributors' creation and maintenance activities on the host platform
where funding is received, as well as the spillover effects on a complementary
knowledge-sharing platform. While sponsorship is expected to increase overall
effort, its effects on different contribution types and spillover to other
platforms remain unclear. To address these questions, we analyze the impact of
Github's sponsorship feature, introduced in 2019, on contribution behavior.
Using individual-level panel data from both the host platform (Github or GH)
and the complementary platform (Stack Overflow or SO), we employ a
difference-in-differences estimation methodology. Our findings reveal that
sponsorship funding increases maintenance-related activities (e.g., pull
request reviews) while leaving creation activities (e.g., pull requests)
unaffected, indicating an effort reallocation effect. Moreover, we observe a
negative spillover on knowledge-sharing activities on SO, suggesting an effort
distortion effect. Interestingly, this negative spillover is limited to
knowledge-creation activities (e.g., answers and questions), while
maintenance-related activities (e.g., edits and reviews) remain unaffected,
evidence of an effort mirroring effect. We attribute these findings to
motivational differences between creation and maintenance activities.
Sponsorship funding appears to boost less rewarding activities such as
maintenance of software, while intrinsically motivated creation activities
remain largely unaffected. We further investigate the underlying mechanisms
and observe a crowding-out effect on creation activities for contributors who
are more involved with the OSS community. Our findings provide empirical
evidence that sponsorship funding can be a powerful tool to incentivize
maintenance-related activities in open-source software. |