2009 – 5 Department of Economics Royal Holloway College University of London Egham TW 20 0 EX © 2009

Trust as a concept and empirical measure has attracted significant attention in economic theory and research within the last two decades. However, gaps remain in our understanding of what exactly is measured by the different experimental and survey measures of trust, and how they relate to each other. In this paper, we demonstrate that the commonly used trust game captures trust in strangers, but is not related to other standard survey measures of trust. To achieve this, a comprehensive examination of both survey and experimental measures of trust was conducted. Using nationally representative data, we demonstrate that the trust game is robust to a number of interferences. Further, we show that selection into the experiment can be a problem if people are unfamiliar with the situation, and that students exhibit higher trust than the general population – a result which needs to be taken into consideration by researchers working with student subjects only. Inspired by criticism of the widespread trust question used in many surveys, we created a new, improved survey trust scale consisting of three short statements. We demonstrate that our new scale is a valid and reliable measure of trust in strangers. The analysis shows that the experimental measure correlates with the survey measure of trust in strangers but neither with trust in institutions nor with trust in known others. The survey measure of trust captures mainly the expectation part of the decision to trust in the experiment. The survey and the experimental measure correlate similarly with related factors such as risk aversion, entrepreneurship and holding shares. We conclude that the experimental measure of trust is a robust and valid measure which refers not to trust in a general sense, but specifically to trust in strangers. JEL classification: C83, C91, D63 and Z13