Perceptions of fairness in social exchange: A comparison of negatively and positively connected networks

Research within the exchange tradition has focused largely on the effects of network structure (in terms of the positioning of actors within a network) on the distribution of power. While network structure in this limited sense has received considerable attention in the literature, one aspect of network structure, which has remained largely unexplored, is the type of network connection. Emerson (1972) identified two types of network connection: negative connections and positive connections. In a negatively connected network, actors must compete with one another for valued exchanges and thus, these networks are inherently competitive. On the other hand, positively connected networks have the potential to be cooperative since exchange in one relation in the network facilitates exchange in a connected relation. In this dissertation, I argue that the type of network connection affects our expectations for the behavior of our exchange partners and thus affects our evaluations of the fairness of those partners. In addition to network connection, I also examine the role that information about a peripheral network member (one who is not a direct exchange partner) plays in affecting evaluations of the fairness of a direct exchange partner. Thus, this dissertation focuses on this concept of network connection and unites exchange and justice theories by examining the effects of network connection on perceptions of fairness. The research design involves a 2 x 2 x 2 factorial experiment crossing network connection (positive or negative), equality of an exchange partner's behavior (equal or unequal behavior) and information about the behavior of a peripheral network member (full information about this actor's behavior or no information). While the hypotheses were not supported in this experiment, some of the results concerning the underlying mechanisms indicate that network connection remains an important avenue for further research. Several possibilities for the lack of support for the hypotheses are discussed as well as the possible avenues for future research