Experimental Design
In the experiment, we implement the one-shot threshold public goods game with four teams, each with three team members, repeatedly in multiple periods. We employ a two-by-two between-subject design. The first treatment variable varies the team endowments. In Homogeneous Endowment treatment, each team has an equal endowment. While in Heterogeneous Endowment treatment, there are two rich teams and two poor teams. Across the Homogeneous and Heterogeneous Endowment treatments, we keep the global endowment constant for a fair comparison.
For the second treatment arm, we vary whether the delegates in each group are chosen randomly by the computer program or elected through a democratic procedure. In Delegation Voting treatment, one delegate is elected within each team to make the team’s contribution decision. The election procedure involves a Pledge stage, in which participants indicate their intended amount of contribution to the public good if elected as the team delegate, and a Voting stage which uses the majority voting mechanism. On contrast, in Random Delegation treatment, team members are informed about the random delegate’s ID directly.