Because of its relation to economic growth, there is a policy interest in measuring social capital and average trust as its currently most important proxy.In this paper we measure societal trust and trustworthiness by combining the virtues of laboratory experiments and survey data and present results from an economic experiment conducted using a representative sample of the Dutch population.By combining both types of data, we are able investigate the determinants of trust and trustworthiness at the population level, to contrast the inferences which can be made on trust propensity using stated and revealed measures, and to test for self-selection bias through voluntary participation in the experiment.Our results can briey be summarized as follows.Contrary to existing laboratory based experiments, we find that stated trust measures correlate with experimental trust.The effect of education and religion is shown to be depend enormously on the trust measure used.We find that the age and education profiles of experimental trust are the complete opposite of those of trustworthiness, which contrast with the findings of the social capital literature.Finally, we do not find evidence of a participation selectivity bias, which is a serious concern for laboratory experiments which rely almost exclusively on volunteer participants.
