Communication in a threshold public goods game with ambiguity: Anomalies and regularities

This paper offers evidence on the impact of communication on public good provision within the context of ambiguous public good values and noisy information. We run a laboratory experiment with two treatments, where the control variable is pre-play communication in the form of unrestricted text chat. A binary threshold public goods game with four-person groups and threshold of three is at the core of the design, whose main novelty lies in the provision mechanism with ambiguity. Moreover, a private signal for the actual value of the public good is provided, before the contribution decision. In accordance with related literature, we find that communication significantly increases public good provision by reducing inefficiency that comes from wasteful undercontribution. Nevertheless, the players in the chat treatment seem to neglect the free-rider issue and often end up overcontributing, in contrast with previous scientific findings. After chat analysis, we propose the common fate hypothesis as original explanation of the massive overcontribution, in addition to group identity generated by the partner matching. Since the players show greater concern for ambiguity than for the free-rider problem, we finally speculate about possible crowding out effect of the former on the latter.