PI proposes experimental test of a new model of coalition for- mation in cooperative exchange networks. Several earlier models predict coalitions depending upon available communication chan- nels. This model adds several game theory considerations - most importantly, side payments and concerns with equity - to ratio- nal actor considerations. The experiments will enable com- parative assessment of predictions from this model and from earlier, competing models. Potential applications include calculating the power of actors as a function of their organiza- tional positions, and designing communication patterns to create desired power positions in organizations.