Introducing a Signaling Institution: An Experimental Investigation

Trustworthiness can be conveyed by sending a costly signal in advance. In an experiment, we find that the smaller the percentage of trustworthy people in a group, the higher is their percentage of signaling. The introduction of signaling has strong distributional effects. It may be efficient because only under this institution can large-scale justified trust develop. In other circumstances, it may be (and in this investigation is only slightly) inefficient. The social product may be decreased, not only because of signaling costs but because of a lack of trust for people who, although trustworthy, are not ready to send the costly signal.