Transparency Fallacy: Perceived Fairness in Algorithmic Management

Abstract Algorithmic management (AM) can pose serious challenges for workers on digital labor platforms (DLPs), such as exploitation and a lack of transparency. Prior information systems research has characterized these challenges as unfair practices, particularly in terms of platform functionalities and procedures (e.g., for salaries). This work examined how disclosing information about AM’s functionalities and procedures affects its perceived fairness. Building on organizational justice theory (OJT), which distinguishes between distributive, procedural, and informational justice, perceived fairness as a measure of justice was used. The authors conducted an online experiment with 234 DLP workers on a self-developed DLP, on which an algorithm allocated suitable tasks. The workers were assigned to three groups with different types of transparency – distributive, informational, and both distributive and informational – and one control group without enhanced transparency. The results suggest that the provision of transparency in AM practices and decision-making processes neither affects the perception of informational nor both distributive and informational fairness. Instead, it influences how AM’s distributive fairness is perceived on DLPs. Although, according to OJT and research on human–computer interactions, individual differences influence perceived fairness regarding transparency measures, factors such as technology affinity and human–computer trust have not been found to be suitable moderators for their interaction. However, the results revealed that these factors were important predictors of perceived fairness, independent of the level of transparency provided.