Although policy and collaborative networks have been studied since the 1970s and 1980s, only recently has the management of these entities come under greater scrutiny. Studies of "network management" are designed to better understand the unique challenges of operating in a context where bureaucracy no longer provides the primary tool for "social steering".

These studies typically make three assumptions about networks, public managers, and the tasks of network management that empirical evidence from our casework in "Newstatia" suggests are suspect at best. If so, then network management theory needs to be reconsidered. The second half of this article begins this process. We have organized this article into six sections. The first defines policy and collaborative networks and discusses why analyzing them and their management independently is probably flawed. The second presents our data and justifications for believing the assumptions outlined above are oversimplifications. The third section reviews three perspectives and two partial models of network management and points out how the perspectives and models need integration. The fourth section develops a resource dependence framework for network management that can encompass the existing models and our new data on the environment in which network management occurs. The final section outlines a series of propositions that flow from our reconsideration of network management