The impact of anticipated discussion on cooperation in a social dilemma

We study the impact of anticipated face-to-face discussions among group members after they have made an anonymous contribution to a public good in an experimental setting. We find that the impact of anticipated discussions depends on how we frame the public good game. When framed in non-evaluative language, anticipated ex post discussions lead to a sharp reduction in contributions to the public good. This effect reversed when evaluative language was used to underscore normative expectations. In contrast, there was no framing in the no-discussion baseline version of our game. We offer an explanation that centres on the idea that the announcement of ex post discussions reinforces both normative and predictive expectations.