Abstract: Abstract Abstract This paper examines social capital and its relation with economic development. We focus on the role that interpersonal relationships play in social exchange, whether through the market or through the provision of public goods. By facilitating search and trust, social capital can increase the efficiency of social exchange where formal institutions are weak. But the benefits from social capital are likely to be unequally distributed. Given these features, documenting empirically the benefits of social capital is complicated by the presence of negative and positive externalities and by the existence of leadership and group effects. Lessons for development policy are drawn at the end. Acknowledgements The support of the Economic and Social Research Council (UK) is gratefully acknowledged. The work is part of the programme of the ESRC Global Poverty Research Group. Notes 1. This case does not rule out the presence of positive externalities on non-members, as long as social capital effects are stronger for members. 2. This may be true of the rural poor, who find little to do. But it need not be true of the urban poor. 3. In practice, employment offices often play a minor role in matching employers with workers. Perhaps a better comparison is between the US and European markets for academic economists or medical interns: in the US, these markets benefit from a strong coordinating device; in Europe, job matching is uncoordinated and probably less efficient. 4. The presence of credit reference firms such as Dun and Bradstreet is another example of formal information sharing device that can be partially substituted by informal networks. 5. This example is inspired of the work of Platteau and Seki (2002) on Japanese fishermen.