On the Relational Aspects of Trustworthiness: Results from a Laboratory Experiment

Do we trust better-connected people more than others? Are those who are better connected more trustworthy? And how does interaction in social networks affect trust and trustworthiness? We address these questions in a laboratory experiment in which trustors decide on how to allocate their endowment to five trustees. The allocation decision is made twice, both before and after a network formation phase during which all subjects can establish links with one another. Our results show that trustors are influenced by the behaviour of trustees in the network formation game: when the allocation is chosen after networks have been formed, trust is directed towards better-connected trustees. However, when trustors’ offers are anonymous, such increased trust is not reciprocated by trustees. This suggests that trustworthiness is not just an individual characteristic but crucially depends on previous actual interaction between trustor and trustee. While the trustor expects well-connected trustees to reciprocate to a greater extent, this does not occur when the trustor is anonymous: the propensity to reciprocate seems therefore conditional on the relationships built in the society.