Are All Contributions Equal? Investigating the Role of Community Participation in Crowdwork

Abstract

ABSTRACTA characteristic of crowdwork (i.e., online labor) is that many workers actively participate in online communities. Despite some anecdotal evidence that these communities help sustain workers, crowdwork struggles with high turnover rate. Meanwhile, it is unclear whether the acclaimed benefits of online communities materialize in earnings. Our objective is to understand what forms of active participation behavior in crowdwork communities affect workers’ earnings and continuance, why, and how. Our empirical probe to these questions is guided by a theoretically grounded tripartite taxonomy of active community participation behavior. We adopt simultaneous equations modeling with instrumental variables and analyze a multisource dataset of 283 crowdworkers who are members of a crowdwork community. We show that decomposing posting behavior into different types reveals its controversial relationships with work outcomes. Specifically, discussing crowdwork skills increases earnings. In contrast, socializing with other workers results in lower earnings. Somewhat counterintuitively, contributing knowledge to the community is negatively related to crowdwork continuance intention. Overall, the effects of the disaggregated community posting behavior imply that online communities can have profound impacts on crowdwork.