Do people prefer being replaced by a robot over being replaced by a human?

November 2019: We are compiling summaries of state-of-the-art research in ethics at the frontier of technology, following the theme of our 2019 Susilo Symposium. Today, we review insights on robotic job replacement from Armin Granulo, Christoph Fuchs (both from Technical University of Munich) and Stefano Puntoni (Erasmus University Rotterdam).

Given the advances in artificial intelligence and robotics, organizations are increasingly able to replace human workers with machines and algorithms. Across the OECD countries, 9% of the jobs are automatable and new technologies could affect millions of workers in different occupations. While there is an intense debate about this development in the business world, little is known about how workers react to this change.

Granulo and his colleagues conducted research exploring the psychological reactions to being replaced by a robot or a human. In several online and laboratory experiments, participants were randomly assigned to one of two experimental conditions.

In the “other-perspective” condition, participants read a scenario in which a firm needed to cut costs and therefore had the option to replace existing employees either by new employees or by robots. The majority of respondents preferred other employees to be replaced by human workers rather than robots.

In the “self-perspective” condition, participants adopted the perspective of the employee about to lose their jobs. When respondents read that they could be replaced in their own jobs, the majority preferred to be replaced by a robot rather than by another human worker.

The thought of being replaced by a robot (rather than a human) generates a conflict by simultaneously reducing self-threat while increasing future economic concerns. The authors consistently find that the feelings of self-threat are stronger than future economic concerns. This result elucidates the preference reversal in the experiment since the feelings of self-threat are more salient when the situation is self-relevant (versus relevant to someone else).

Granulo and his colleagues recommend policymakers to design unemployment interventions targeted at retraining and upgrading skills: “such retraining interventions could not only provide workers with new skills that are difficult to automate but could also alleviate feelings of future economic concerns by reducing perceptions of skill obsolescence” (Granulo et al. 2019, p1067).

The published academic paper can be found here:

Granulo, Armin, Christoph Fuchs, and Stefano Puntoni (2019), Psychological reactions to human versus robotic job replacement. Nature Human Behaviour, 3, 1062-1069.

View all posts