Data for: Avoiding the Cost of Altruistic Punishment Through Principle-Evading Delegation

While altruistic punishment sustains social norms, its
costs encourage responsibility shifting via strategic delegation. Across two
laboratory experiments, we examine whether delegation also serves as a principle-evading
means to resolve the conflict between upholding fairness norms and avoiding the
costs of altruistic punishment. In study 1, after experiencing betrayal in a standard trust game,
trustors could choose to punish their counterpart through a costly lose-lose
allocation, or to cooperate in a win-win outcome, with the option to delegate
this decision to a random device. Allowing delegation reduced direct punishment
after defection. Study 2 compared delegation to a human agent versus to a
random device. Human agents applied principals’ standards, leaving acceptance
thresholds unchanged—as participants anticipated. By contrast, the random
device provided credible cover and was used to accept lower offers. Delegation
thus changes how enforcement is carried out: impersonal procedures diffuse
responsibility and temper sanctions, whereas accountable human delegation
preserves sanctioning standards. These findings clarify when delegation
undermines norm enforcement and inform the design of organizational and
algorithmic governance: mechanisms that anonymize responsibility (e.g., opaque
algorithms) can erode deterrence, while retaining identifiable human
accountability helps sustain fairness norms.