pp. 599-628

This paper is an introduction to social capital effects at the top of the firm. I contrast social capital with human capital, and discuss the information and control benefits provided by contact networks rich in structural holes. Higher returns to human capital are expected for people rich in social capital. I use network data on a probability sample of senior managers to illustrate two conclusions : 1) Social capital matters for the relative success of managers. 2) Social capital matters more where individuals matter more. Where many people do the same kind of work, peers are a frame of reference, and legitimacy is established by the number of people doing the work. Where few people do the same kind of work, there is no frame of reference, and legitimacy has to be established. The information and control benefits of structural holes are more valuable for the second kind of work. Results on the senior managers show that social capital is more valuable toward the top of the firm, and more valuable for managers in unique jobs than for managers with many peers. In fact, the value of social capital decreases exponentially with the number of managers doing the same work.