To suggest is to commit? A common pool resource experiment with non-enforceable recommendations

•I ran a common pool resource game with simultaneous non-binding recommendations.•Subjects are highly consistent with the sent recommendation.•Subjects are not fully responsive to the incoming recommendation.•Consistency is correlated with the ratio between reported and satisfactory earnings.

I report the results of an artefactual field experiment conducted with Colombian artisanal fishermen. I set up a common pool resource game in which subjects exchange recommendations prior to the extraction decision. The classical tension of this game between individual and collective incentives opens the door for strategic communication. However, I find that subjects are highly consistent with their transmitted message and, to some extent, responsive to the incoming recommendation. Recommendations are efficiency enhancing: extraction levels are lower when the outgoing and the incoming messages are closer to each other. This is particularly relevant in treatments with a high proportion of cooperative recommendations. I link experimental behavior with survey data and I find that lying behavior is negatively correlated with the ratio between the (reported) satisfactory and realized earnings from the fishermen.