How to convince someone that you can be trusted? The role of 'hostages'

We consider how credible commitments ('hostages') in the sense of Schelling affect trust and trustworthiness. Using the Trust Game as a metaphor, based on game theoretical modeling and previous results with the Trust Game, we derive several hypotheses on which hostages are more likely to effectively influence the degree to which actors are willing to trust others. Two main characteristics of hostages are highlighted. To overcome a lack of trust, hostages should credibly promise to compensate when trust is abused or sufficiently reduce the probability that trust will be abused (or both). Our experimental data show that compensating the trustor is more important than reducing the probability of abuse by the trustee. Moreover, counter to our theoretical predictions, the data suggest that hostage posting by itself has positive "signaling value." That is, posting a hostage may induce the other party to trust for two different reasons. Trust increases not only because the incentive structure changes in a favorable way, but also because of the mere fact that the actor has chosen to post a hostage instead of refraining from it. It also works the other way around. Not posting a hostage when one could have, has negative signaling value. Unfortunately, these signals are deceptive. Those who post a hostage are not more likely to reciprocate trust by trustworthy behavior, and those who do not post a hostage are not more likely to abuse trust.