Abstract
A large number of studies have used both an economic experiment and surveys to measure trust. There is some evidence in the literature on how behaviour in the experiment is related to socio-economic characteristics (for example, age, gender, income levels, educational attainment, marital status and group memberships). However, the relationship between survey trust and such characteristics has not been explored in the current literature. This paper explores this relationship. Generally, the extent of trust declines as the radius of trust widens, suggesting that social distance is important. The results show some evidence that survey trust is correlated with socioeconomic characteristics. However, the correlates of context-specific and non-context specific trust are different. The number of years lived in the village is the key determinant of non-context specific trust. ROSCA membership is important for noncontext specific trust in fellow ROSCA members only. Age and marital status are significantly negatively correlated with non-context specific trust in other village members. Income is what really matters for context-specific trust; however, years lived in the village and whether someone has ever lived in an urban area are also correlated with trust in fellow villagers.