Making the Most of Your Regret: Workers’ Relocation Decisions in On-Demand Platforms

Abstract

Problem definition : We have witnessed a rapid rise of on-demand platforms, such as Uber, in the past few years. Although these platforms allow workers to choose their own working hours, they have limited leverage in maintaining availability of workers within a region. As such, platforms often implement various policies, including offering financial incentives and/or communicating customer demand to workers in order to direct more workers to regions with shortage in supply. This research examines how behavioral biases such as regret aversion may influence workers’ relocation decisions and ultimately the system performance. Academic/practical relevance : Studies on on-demand platforms often assume that workers are rational agents who make optimal decisions. Our research investigates workers’ relocation decisions from a behavioral perspective. A deeper understanding of workers’ behavioral biases and their causes will help on-demand platforms design appropriate policies to increase their own profit, worker surplus, and the overall efficiency of matching supply with demand. Methodology : We use a combination of behavioral modeling and controlled laboratory experiments. We develop analytical models that incorporate regret aversion to produce theoretical predictions, which are then tested and verified via a series of controlled laboratory experiments. Results : We find that regret aversion plays an important role in workers’ relocation decisions. Regret-averse workers are more willing to relocate to the supply-shortage zone than rational workers. This increased relocation behavior, however, is not sufficient to translate to a better system performance. Platform interventions, such as demand information sharing and dynamic wage bonus, can help further improve the system. We find that workers’ regret-aversion behavior may lead to an increased profit for the platform, a higher surplus for the workers, and an improved demand-supply matching efficiency, thus benefiting the entire on-demand system. Managerial implications : Our research emphasizes the importance and necessity of incorporating workers’ behavioral biases such as regret aversion into the policy design of on-demand platforms. Policies without considering the behavioral aspect of workers’ decision may lead to lost profit for the platform and reduced welfare for workers and customers, which may ultimately hurt the on-demand business.