What Drivers Want vs. What Drivers Take: A Worker-Centered View of Fairness in Ridehail Platforms

Abstract

Digital labor platforms increasingly mediate work and disrupt traditional employment norms. Since many workers are now classified as independent contractors, established workplace fairness rules, like the Fair Labor Standards Act in the U.S., do not cleanly apply. Frameworks for what constitutes fair work in these contexts remain underdeveloped. Over the past decade, machine learning research has offered mathematical formalizations for defining and operationalizing fairness in multistakeholder platforms, including digital labor platforms. In this paper, we respond to calls to explore what fair work looks like on digital labor platforms and contextualize mathematical formalizations of fairness in ridehail platforms by examining, from the workers' perspective, how they define fair work and fair compensation. We interview nine ridehail drivers using simulated task offer screens to elicit how drivers define fair work and operationalize these definitions in practice when deciding whether to accept or reject offers. We find that drivers define fairness in task offers based on six features: location, duration, distance, take rate, anticipated passenger behavior, and vehicle operating costs. Rather than reducing fairness to a single metric, drivers construct context- and task-dependent judgements about the work, cost, risk and opportunity each ride requires. We further identify a gap between what drivers consider fair and what they accept, driven by both algorithmic pressure to maintain acceptance rates and the economic reality of market scarcity. Our work contributes a worker-centered operationalization of fairness in digitally mediated work platforms that connects multisided platform fairness with the judgments workers construct in practice.