This is a study of collusion in exchange networks. A series of laboratory experiments, using a variety of networks of exchange and communication, will test hypotheses about the positions in social networks most likely to develop coalitions with each other, and the shapes of networks most conducive to coalition formation. This research springs from the theoretical work of Richard Emerson, who proposed that power in exchange networks tends to equalize. Those initially with less power have strategies available to increase their power, among which is the possibility of forming countervailing coalitions. Individuals with less initial power can increase their power by forming a single bargaining unit. The designs of previous experiments have precluded this possibility by preventing non-bargaining subjects from communicating with each other. The present research project suggests a model predicting which positions will form coalitions in exchange networks, and a series of experiments will test the model. The model is an advance over Emerson's in two respects. First, it implies that coalitions may form between weak positions that do not share a common trading partner, and it predicts that coalitions may form between classes of positions that are quite remotely connected in the network. Emerson's model was limited to weak positions that shared a common trading partner. Second, it describes a new incentive for forming coalitions: the reduction of uncertainty in unstable networks. Unlike coalitions of the weak, these are coalitions between exchange partners. %%% This project is fundamental research on the ways that coalitions may affect the power of individuals in social networks of various shapes, and thus it addresses crucial questions in theories of social capital. The work is connected to emerging new theories of latent classes that can be tested in the laboratory and then extended to help explain degrees of inequality in actual communities. The structure of social networks may turn out to be a major factor shaping the rewards individuals receive in social exchanges, quite apart from the effect of their skills or other previously recognized attributes of human capital.
