Social Capital, Values and Trust: A Conceptual and Analytical Clarification.

Trust does not feature in the values or motivation literature as a determinant of social capital, except in some cases as a personal trait. In the social capital literature, however, shared norms are discussed as a key ingredient in building trust but no link is made to the values literature. At the same time, trust is often also considered to constitute social capital. The link between trust and structural social capital (social relations as memberships in organizations or networks, volunteering, informal relations), however, has been shown to be rather tenuous. In structural models of social interaction, on the other hand, trust re-emerges as a product of social interactions and is linked to shared norms of reciprocity. It is thus considered to be endogenous to social interactions. This paper argues that depending on the level and the nature of social interactions, it is personal values, not levels of trust, that influence prosocial behaviour. There is, for example, an important difference between helping and cooperation. Helping involves individual or unilateral action, while cooperation often involves action within or between groups. One of the key differences between the two in terms of individual-level and group-level mechanisms, is the expectation of reciprocity involving trust. Thus trust is relevant in situations of interdependence, such as in relationships within or across networks, and it is endogenous to social exchange. Values, on the other hand, influence individual behaviour such as unilateral helping or social volunteering. As structural social capital can involve both individual and group relations, trust can, but does not have to, be involved. Most social capital authors, however, despite aiming to explain individual-level action such as volunteering, discuss trust as a foundation of social capital and in the context of shared norms/values. This paper thus aims to conceptually and empirically separate determinants and consequences of social interaction at different levels. Cross-sectional data (WVS, ESS) is used to show that individual level helping behaviour, such as volunteering, is influenced by personal values, rather than levels of trust.