From their extensive work with networks of mental health providers, Provan and Milward developed a framework for assessing the effectiveness of public sector service delivery networks (1995, 2001) The framework provides for three levels of analysis: community (the population to be served), network (the organizational manifestation of the working arrangements among agencies), and organization (the individual agencies that participate in the network). This conceptualization serves as a point of departure for building a framework for evaluating the success of public sector knowledge networks.

A key difference between service delivery networks and knowledge networks requires adjustments to the Provan and Milward framework as a basis for evaluating knowledge networks. First, the primary community served by a knowledge network is not external to the participating organizations as it is with service delivery networks. The organizational and individual participants in knowledge networks are their own “customers” or “clients.” The knowledge resources and other benefits that the network provides serve directly the knowledge and information needs of the participating organizations.

Second, individual professionals play an important role in knowledge networking that should be considered apart from the interests of the organizations that employ them. Individual participants negotiate and manage organizational participation and individually use the network’s knowledge and information resources. This perspective follows Ring and Van de Ven who posit that cooperative IORs develop due to the actions, decisions, and interpersonal relationships among individuals within the participating organizations rather than from organizational factors alone (Ring and Van de Ven, 1994).

Taken as a whole, the foregoing literature suggests several salient categories of success measures: structural measures that address network and organizational forms and viability, performance measures that address the stated objectives of the network and its participants, and process and relationship measures that address the quality of the interactions that constitute network dynamics. These levels of analysis (network, organization, and individual) and categories (structure, performance, and interaction) are used to summarize measures of PSKN success as shown in Table 1.

Different sets of measures are appropriate for evaluating PSKN success in terms of the network, participating organizations, and individual professionals. For instance, institutionalization and data sharing mechanisms have meaning only at the network level. By contrast, some measures pertain to more than one level of analysis, but carry different meanings and therefore require different metrics at each level. For example, knowledge acquisition and learning“ pertain to both the organizational and individual levels of analysis. However, the kind of knowledge that an individual might gain from network participation likely to enhance his or her expertise in a particular profession (an urban planner, for example, might learn how to use geographic analysis tools). At the organizational level, new knowledge might be embodied in enhanced planning processes that take into account new information repositories, different stakeholders, or special legal requirements.

	Table 1. Measures of Success for Public Sector Knowledge Networks
	Level of Analysis
	Network	Participating Organizations	Participating Individuals
Structural measures	legitimacy creation and maintenance of a network administrative structure institutionalization growth in network membership network stability and resilience in the face of environmental threats resource growth survival beyond the tenure of key individual participants	organizational survival enhanced organizational legitimacy
Performance measures	achievement of specific long-term & interim substantive goals joint product development range of services provided to network members integration and coordination of services provided to network members growth in knowledge content and use high quality data efficient data distribution and sharing mechanisms	organization-specific performance goals are met efficiency improvements operational improvements resource acquisition contribution to core organizational competencies knowledge acquisition and learning	individual professional goals are met contribution to core professional competencies knowledge acquisition and learning
Interaction measures	achievement of process goals equity or “fair dealing” among participants relationship strength (multiplexity)	enhanced reputation or legitimacy equity or “fair dealing” among participants minimum conflict across membership in multiple networks overall satisfaction with the network	enhanced reputation or legitimacy minimum role conflict across multiple networks building of social capital overall satisfaction with the network