This paper experimentally studies peer punishment under three alternative technologies. We find that the choice of peer punishment technology has a substantial impact on group performance. First, under technology where at least two subjects in the group must agree before another group member can...

Punishing the free-riders of a team can promote group efficiency but is costly for the punisher. For this reason, economists see punishment as a second-order public good. We show in an experiment that subjects do not value punishment for its deterrence but instead for the satisfaction of...