Coordination Under Threshold Uncertainty in a Public Goods Game

Many natural resources involve threshold effects. Using these resources beyond a tipping point can have disastrous consequences for the environment and human well-being. Prominent examples are related to catastrophic climate change, such as the collapse of the Atlantic Thermohaline Circulation or the decay of the Greenland ice sheet, and the collapse of natural resources, such as fish stocks, grassland, or forests. The potentially dangerous consequences have led to a political consensus about the urge of avoiding such thresholds. However, these natural tipping points entrench high uncertainty which may seriously affect people's willingness to cooperate in order to prevent catastrophes. In this work, we explored the effect of uncertainty on agents’ ability to coordinate their cooperative efforts in order to prevent a collective damage. To this end, we conducted a laboratory experiment involving a threshold public goods game and compared how coordination success was affected by whether the threshold was known or not. In particular, we employed four different forms of threshold uncertainty. Whereas two experimental treatments involved risk, as the threshold was a random variable with known probability distribution, two other treatments involved ambiguity, as the probability distribution of the threshold was unknown. Our experimental data indicate that threshold uncertainty was detrimental for the provision of the public good. Whereas all groups succeeded in preventing the public bad when the threshold was known, this result was not replicated in the presence of threshold uncertainty. Although the contribution pattern differed depending on how uncertainty was configured, contributions were generally lower when players did not know ex-ante the exact threshold value. Critically, contributions were particularly low and erratic in the treatments involving ambiguity. We also found that early signaling of willingness to contribute and share the burden equitably made groups more likely to reach a high public good provision level, even in the presence of threshold uncertainty.

Dannenberg, Astrid, Andreas Löschel, Gabriele Paolacci, Christiane Reif and Alessandro Tavoni (2011), Coordination Under Threshold Uncertainty in a Public Goods Game, ZEW Discussion Paper No. 11-065, Mannheim.