Network Determinants of Interest Groups' Participation in Coalitions Over Time

Interest groups tied to other organizations by embedded, as opposed to arm’s length, ties are more likely both to participate and be leaders in coalitions. Specifically, I argue that an organization with a superior position in the policy domain network will participate in more coalitions, and an organization that participates in prior coalitions in the policy domain network is more likely to participate in future coalitions. Using a unique, longitudinal dataset (1998-2004) of lobbyists in the federal retirement policy domain, I analyze the level of participation in coalitions using network analysis and zero-inflated negative binomial regression. I find that coalition relationships are very stable over time and that organizations with a long-term presence within the policy domain will have more stable ties to other organizations through coalitions – That is, coalitions are more likely to replicate themselves with long-term organizations than with more short-term organizations. An important finding of the negative binomial regressions is the significance of prior coalition work on the level of coalition participation (as opposed to the initial decision to join in a coalition). The finding suggests that prior experience acts as a multiplier effect: Organizations are able to somehow leverage that experience into greater joint activity. Perhaps prior experience provides a level of comfort against the fear that others may free ride or otherwise take advantage of coalition partners