The main result of this thesis is that people follow trust norms when making the trust decisions that underpin economic performance. Specifically, mean trust is shown to be a statistically significant determinant of individual trust. This...
moreThe main result of this thesis is that people follow trust norms when making the trust decisions that underpin economic performance. Specifically, mean trust is shown to be a statistically significant determinant of individual trust. This empirical result, which implies that there are endogenous social interactions effects in trust decisions, is found in a Random Utility Model with Social Interactions (RUM-SI), in which people have rational expectations and make binary decisions, to trust or not to trust others.
A theoretical model of trust as a recurrent 2 × 2 coordination game with Stag hunt payoffs, played adaptively by randomly matched players, explains how boundedly rational agents make coherent, mutually consistent decisions despite the knowledge and information constraints of the trust game. The theoretical model therefore provides justification for the model closure assumption in the RUM-SI, while its social learning dynamics allows boundedly rational agents to behave ‘as if’ they were fully rational, as was assumed in the RUM-SI. Hence, there is consistency in the behavioural assumptions of both the theoretical and empirical models.
These models have multiple aggregate trust equilibria in high, medium and low trust, which explains both the local (national) conformity and the global diversity in trust decisions evident in the data.
High trust and low trust equilibria are stable norms, while medium trust is unstable and thus is not a norm. The theoretical model implies that in the (very) long run all countries would be caught in stochastically stable low trust traps. Institutions prevent both the short run collapse and the long run persistence of low trust.
A tipping mechanism or informal institution would be required to transform low trust states into high trust states. Formal institutional reform, which itself cannot effect the initial transformation, will then be required to prevent reversion to the stochastically stable state in the longer run.