To avert dangerous climate change global CO₂ emission has to be reduced by 50% until 2050 (IPCC report). We simulated this collective risk social dilemma in a nutshell: Will a group of people reach a fixed target sum through successive monetary contributions, when they know they will lose all their remaining money with a certain probability if they fail to reach the target sum?

Groups of 18 players play 3 consecutive independent games of 10 rounds each. They are subdivided in 6 groups, “countries”, of 3 players each. Only one player, the representative, per “country” decides for her group how much to invest from the money (3x 40 €) of her group mates in each of 10 rounds, i.e. €0, €6, or €12. If the target sum of 360 € is not assembled after round 10, each group member loses all her remained money with a high probability. If the target is reached, they receive their remained money in cash. The less a representative invests, the higher is the gain for her group members, if the other representatives have compensated and invested more than their fair share, i.e. €6 per round. After the first and the second game, each representative can be voted out or re-elected by the members of her group dependent on her decisions during the previous game and her election pledge how to behave if elected (again).

We found that fair representatives are often voted out and selfish players are elected. Players who announce in their pledges to act selfishly for the group are significantly preferred. Elected players decide according to their pledges. Thus, representatives that invest less than the fair share are increasingly dominating the climate change game. However, fair representatives compensate for the low investments of the selfish representatives increasingly from the first to the third game, so that most groups reach the target sum and avert simulated dangerous climate change in the third game.