Distrust Trap: When Changes in Beliefs About the Behavior of Others are Not Paired with Welfare Variations

Abstract

This study contributes to the understanding of trust as a determinant of welfare from the perspective of a microfunded model. In a competitive setup, agents decide whether to participate in a productive interaction and an action therein. Matching in the interaction is anonymous. The maximization of expected utility requires individuals to forecast the endogenous distribution of agents' actions in society. The equilibria are Pareto rankable, which permits characterizing the relation between beliefs and welfare through comparative statics by parametrizing economies with respect to the payoffs of the interaction. It is possible for an economy to reach two different equilibria with equal welfare and different beliefs. This study contributes to explaining dispersion in beliefs for a given welfare measure. In particular, we argue that these beliefs can be measured by generalized trust. The analysis identifies a distrust trap, i.e., a situation where changes in beliefs do not imply welfare improvements; estimations with data from the World Values Survey provide a 6% average threshold variation in trust where no welfare improvements are present. This finding implies that public policy toward the improvement of social capital could be ineffective below that threshold, and that the characteristics of social interactions are essential for welfare improvement.

Keywords: Welfare, Beliefs, competitive equilibria, generalized trust