To Speak up or Shut up? Revealing the Drivers of Crowdworker Voice Behaviors in Crowdsourcing Work Environments

This study examines worker voice behaviors in the microtask crowdsourcing work environment (CSWE) where voice channels are absent. Informed by employee voice research, this study adopts the revealed causal mapping method to analyze the detailed narratives of 60 workers from Amazon Mechanical Turk. Our data analysis shows that the crowdworkers did engage in voice behaviors, but their voices were not always heard, depending on recipients. The crowdworker voice was directed to three different recipients (worker community, job requester, and platform) and influenced by six antecedents (duty orientation, efficacy judgment, workgroup identification, anger/frustration, futility, and achievement orientation). Based on the findings, we propose a model of worker voice antecedents and moderators in the CSWE. This study extends employee voice research by presenting a moderator perspective in the CSWE. Moreover, our study provides a nuanced understanding of crowdworker voice behaviors from two major aspects - antecedent and recipient - contributing to crowdsourcing research.