A Multilevel Approach to Network Effectiveness

This section first discusses a multilevel approach to network effectiveness— the evaluation of network performance at the organizational, network, and community levels (Provan & Milward, 2001). Then, it introduces the use of social network analysis for evaluating network performance.

Evaluating Nebvork Performance at Different Levels

Networks are examined at multiple levels such as: dyads, actors, groups and whole networks and network can be analyzed at the micro, meso, and macro levels (Borgatti, Everett, & Johnson, 2013; Kilduff & Brass, 2010; Mandell & Keast, 2007, p. 585). As Provan and Milward noted in their seminal work on network effectives, networks need to be evaluated at different levels: community, network, and organizetion/participaut levels (2001, p. 414). Three components are involved with these levels: principal organizations/administratiou; agent and clients. Interactions between three levels are required for an effective network, but network effectiveness at one level does not guarantee effectiveness at the other two levels (Provan & Milward, 2001).

Effectiveness at the Organization/Participant Level

The effectiveness of network at the individual level can be evaluated using four criteria: “client outcomes, legitimacy, resource acquisition, and cost” (Provan & Milward, 2001, p. 420). For small and less-established organizations, joining a publicly funded network may provide more legitimacy. In addition, organizations can get more resources by engaging in a network. Organizations, even larger ones, can reduce their costs and enhance client outcomes through network involvement and tapping into network resources and learning. Service networks can integrate resources from individual organizations and help network members to provide a broad range of needed services more efficiently and effectively (Provan & Milward, 2001). When evaluating network performance, one should ask the following questions: Does your organization gain more resources after joining the service network? Does your organization reduce the cost of service delivery after joining the service network? Does your organization gain more legitimacy? Does your organization improve efficiency of service delivery? Does your organization improve quality of services provided to clients?

Effectiveness at the Network Level

The first way to evaluate network effectiveness is to examine "the ebb and flow of agencies to and from the network” by measuring the numbers of network members that join and stay engaged (Provan & Milward, 2001, p. 418). A stable number of members can reflect the stability and function of the network. The second way of evaluating network-level effectiveness is to examine the extent to which the network fulfills the actual needs of clients. To evaluate the strength of network members' relationships is the third way to assess network effectiveness. In an effective network, many organizations have multiplex relationships, meaning that they are connected to one another through different types of programs or client services or through general information sharing and friendship. The last way of assessing network-level effectiveness is to judge whether its administrative structure can obtain and distribute resources in ways that maximize resources utilization within a network (Provan & Mil- ward, 2001). Questions can be asked: How many organizations join the network and stay engaged in the network? Does the service network improve sendee integration across organizational boundaries? What types of relations do organizations have with one another? Does the governance structure support the function of the service network?

Network Effectiveness at the Community Level

The satisfaction of clients and other community-interest groups provide important legitimacy and external support for networks. To evaluate the effectiveness of a network at the community level, the main measures are the impact on the clients' well-being, the total cost of sendee, public satisfaction about the sendee, and growth of social capital (Provan & Milward, 2001). These questions should be asked: Does community social capital change after the network is formed? Does the service network improve clients’ well-being? How much does it cost for the network to deliver the service? Another way to evaluate network effectiveness is to link the community-level network effectiveness to collective impact. Instead of making isolated individual impact on communities, collective action of network actors can make collective impact (Kania & Kramer,

Collective Impact

The positive effect generated by the network at the community level.

2011). For example, in Cincinnati, to address student performance crisis in greater Cincinnati and northern Kentucky, Strive, a nonprofit organization has brought together more than 300 community leaders, government officials, educators, school district representatives, heads of foundations, and advocacy groups. These network actors focus on the entire educational community and set goals to be obtained by the entire community (Kania & Kramer, 2011). These goals cannot be achieved by any individual organization's isolated efforts but requires collective effort. Therefore, the extent to which these goals are accomplished can be used as measures of community-level network effectiveness.

Network effectiveness at the three levels is interdependent, as the outcome at one level may influence the outcome at other levels. Various stakeholder groups may have different priority levels of network effectiveness, making leading, integrating, and balancing network effectiveness at three levels an uneasy task (Provan & Milward, 2001).

Nebvork Analysis to Assess Network Performance

Traditional performance measures focus on individual organizations; therefore, it is imperative to find new ways to measure performance, structural patterns and processes of interorganizational networks (Kapucu & Demiroz, 2011; Mandell & Keast, 2007). Developing appropriate performance measures for interorganizational networks is a challenging undertaking (Walker et al., 2012). Networks have their own unique attributes, and thus additional tools are needed to measure their performance. Social network analysis allows researchers to evaluate network performance at individual, dyadic, group, and network levels.

Interorganizational relations are examined to evaluate the effectiveness of a collaborative network. Network members can use their relations more effectively if they understand their existing relations and if they are able to build new relations with other members when needed. Since members are interdependent, they must assist their partners. Otherwise, a member may hamper the coordination of the whole network (Kapucu & Demiroz, 2011). Organizations can better engage with their stakeholders by understanding the relations among organizations, and central and peripheral roles played by organizational stakeholders (Prell, Klaus, & Reed, 2009). Furthermore, social network analysis can help enhance interorganizational effectiveness of a network by identifying high and low performers in a network and the reasons behind performance discrepancies.

Social network analysis focuses on how different actors are connected to one another through relations. A detailed introduction of the network measures has been provided in Chapter 2. Many network concepts are meaningful for studying network performance, such as strong ties, weak ties, homophily, centrality (both degree and betweenness,) and centralization.

Centrality measures (degree, closeness, betweenness, and eigenvector) that are used to identify important actors in a network are indicators of the amount of collaboration and cooperation, connectivity, and communication. Those concepts are also the objectives of creating network strucnire. Centrality measures "reveals the organizations that are able to interact with others not only in a small group but also in the network as a whole” (Kapucu & Demi- roz, 2011, p. 560). The performance of September 11 and Hurricane Katrina response networks have been examined using network analysis techniques. They compared the levels of the collaboration and networking identified in the plans- Federal Response Plan (FRP), National Response Plan (NRP)— and the implementation of these plans in response to disasters. The FRP and NRP illustrate every aspect of how the federal government should respond to a disaster by identifying the functional responsibilities of each agency and department of the federal government. The UCINET program was used to evaluate the relationships among the agencies that responded to these catastrophic disasters.

The study results stated that “the September 11 response network performance differed from the FRP network and the Hurricane Katrina response network outcomes showed different structures from those of the NPR. The structural differences between these formal versus informal and planned versus acmal networks demonstrate the utility of providing measures of network outcomes” (Kapucu & Demiroz, 2011, p. 573). On the other hand, two factors are necessary for a healthy network performance evaluation in this approach. "First, there should be an identifiable planned network structure that can be compared against the acmal network, as is the case for disaster response networks. Second, there should be reports or other reliable data sources that identify network actors, and their collaborative actions” (Kapucu & Demiroz, 2011, p. 573).

< Prev   CONTENTS   Source   Next >