Leading-by-example in public goods experiments: What do we know?

Abstract

This meta-analysis explores whether leading-by-example increases contributions in public goods experiments. I find that leadership overall improves public good provision significantly, whereas leaders benefit less than followers. The reason is that followers match the leader’s contributions imperfectly and, on average, only contribute about 79% of the leader’s contribution. Consequently, only a small proportion of group members are willing to bear the burden of leadership. Despite this, leaders are usually not worse off compared to an average player in a simultaneous game without a leader. In general, leadership becomes more effective when one adds transactional elements like punishment and reward, whereas it becomes less effective if there is frequent change (rotating leadership).