Algorithm Aversion in Joint Decision-Making: The Role of Preference for Neutrality

Given the widespread integration of algorithms into consumer decision-making, a growing body of research has examined how individuals respond to algorithmic versus human recommenders. However, most existing work has focused on individual decision-making, leaving open the question of how recommender preferences unfold in joint decision-making contexts. This study addresses that gap by investigating how consumers choose between algorithmic and human recommenders when making decisions individually versus jointly. Across five studies, we find that consumers exhibit lower algorithm aversion in joint (vs. individual) decision-making contexts. This effect is mediated by an increased preference for neutrality in joint decision-making contexts and is moderated by the presence of symmetries between decision partners, with algorithm use more likely when partners are asymmetrical in their goals. These findings contribute to the literature on algorithm aversion by extending it into social decision contexts and offer practical insights for the design and deployment of recommender systems in collaborative settings.