The economic literature on social capital usually argues that the economy can significantly benefit from social connections, norms, and trust. In fact, transaction costs may thus be reduced, and entrepreneurship may thus be encouraged.

By contrast, the sociological literature on social capital has distinguished between bridging and bonding social capital. Whereas the bridging type of social capital is characterised by the easy of social connections, thus enhancing civil engagement and democracy, the bonding type of social capital is characterised by close social attachment, which enhances secure actions within the strong ties, but possibly damaging the more general social welfare.

It has thus been suggested that social capital is not always a good thing. Connection, norms, and trust can develop within families and clubs, but possibly looking for rents. At the opposite, creativity and innovation often come out from bridging new social ties. But how can be distinguished the “good” type of social capital from the “bad” one? Strong ties are necessary for human development – the psychologists say – so that the strong vs. weak ties distinction is not a sound guide. Not even inferring the distinction within social capital from its good vs. bad effects on welfare can inform a sound theoretical analysis.

This paper attempts an answer to this issue by exploring the analysis of relational goods as a distinct concept from social capital. Relational goods are usually defined as produced and consumed at the same time by both parties of the relation. These goods are thus final goods, and imply bilateral actions. However, these goods do not necessarily imply symmetry of any sort from the two parties. Take the typical relational goods emerging between parents and children. Education and upbringing require, rather, asymmetry; the two parties exchange things which cannot be compared, and often they exchange things as externalities. Indeed, a good parent can be defined as her/him who enjoys the relationship with her/his children just from seeing them realising their own goals. This means, at the same time, that good parents realise their educational ability. But psychologists argue that realising competence and relatedness develops the individual’s self (or personal identity), and raises wellness. Therefore, relational goods can be enjoyed for their own sake, and this enjoyment can be learned.

Social capital, if simply defined as a set of norms and social conventions, plays a rather different role. It offers a public good to the individual (i.e. less risky and more pleasant environment), who may respond and contribute. Behavioural economics has observed that if the individual defects, s/he suffers a psychological cost. Evolutionary economics argues that the tendency to reciprocate, though this is of a specific kind, is a human characteristic.

The paper proposes a model which takes into account both relational goods, and social capital as discussed above. The individual’s choice setup is organised around three kinds of actions which are constrained by limited resources: (i) actions for exchanges on the markets, (ii) actions for their own sake, in order to capture enjoyment of relational goods, and (iii) actions for adjusting to the social environment, in order to capture the special reciprocation to social capital. Each kind of action is characterised by a reward depending, respectively, on the market stance, on individual’s ability, and on changing environment.

Since both the individual’s ability in enjoying relationships, and social environment change endogenously, a dynamic model is needed. Since actions are not strategic, the model can be parametric.

Three stable equilibria emerge as solutions of the model. In the first equilibrium (called A), the consumption of market goods is at a relatively low level, while both relational goods and social capital are at relatively great levels. At the opposite, in the second equilibrium (called C), the consumption of market goods is at a relatively great level, while both relational goods and social capital are at relatively low levels. In the third equilibrium (called B) social capital is at an intermediate level, while both the consumption of market goods and relational goods are at relatively low levels.

Equilibria A and C appear similar to those found by the models of the social poverty trap. However, those results depend on the assumption that social ties need reciprocity, and that trust is played strategically, so that either cooperation or competition in the social setting emerges as a stable equilibrium. My result of the first two equilibria depends rather on resources constraint only. If the benefit of social capital on economic efficiency and innovations were considered, these constraints would be even relaxed, and equilibrium A also tends to Pareto-dominate equilibrium C.

But the major result of my model regards the emergence of the third stable equilibrium (i.e. B). In fact, comparing equilibria A and B shows that social capital is fully beneficial for welfare, only if relational goods flourish. The distinction between “good” social capital and “bad” social capital, as pertaining to the bonding type, can thus be established.