This is often the underlying question when we map out networks: Participants are eager to know if they have the correct, the best network structure. But giving a normative answer to this question (as in: This is how your network should be) isn’t as easy or straightforward as they might hope. Because before you can judge whether or not this is the best structure, you have to ask yourself: For what?
I have set myself the task of reading the whole Wasserman and Faust (which is one of THE social network analysis standards works and heavy enough to kill a healthy rat) and they say: Networks (as in relatively open systems with no or overlapping hierarchies and diverse participants) are ideal for solving non-standard problems. That menas on the other hand, if you know the solution to your problem already and you just have to implement it effectively and efficiently, maybe a more clearly hierarchical structure with defined membership and sanctions might make more sense.
I started thinking about this when brooding over the data collected in the Pro-Poor Risk Reduction – Avian Influenza Project and maybe the answer to “What is the best network in this case?” is: You need two different ones for two different steps of the solution. First you have a non-standard problem: The threat of avian flu to your country is a complex new problem and you don’t know yet, who has to react how to make sure your measures stomp out the disease without stomping out the livelihoods of your poor population. To find solutions, rules, regulations, project goals and structures you want to involve a lot of different stakeholders in an open manner where everyone can contribute without being stiffled by formal hierarchies and power play. A time consuming but potentially sustainable process.
But then, after you have gone through this and developed your rules and structures, you are hit by an outbreak on a farm or on a life bird market. ACTION! NOW! Start stomping! If you are prepared and have put appropriate rules and processes in place, this should now be a standard situation, where actors have to follow some simple, pre-defined rules to put movement restrictions in place, cull birds within infected zones, enforce necessary action fast. This is a military style concerted action that needs clear coordination and hierarchy structures to make sure that things happen without delay and that it is possible to enforce those steps that farmers or traders might be reluctant to take.
From our discussions when drawing network maps about disease response in Ghana and Ethiopia, I got the impression that the involvement of too many actors in unclear hierarchical relations, where everyone and their brother have to sign specific forms to make sure that action is taken, was not seen as the best model for avian influenza response, while clear directives and a lean structure were apprechiated.
(This is one of these posts that help me develop ideas that are just growing while I write. I would really apprechiate comments and discussion to see if this makes sense and how you have experienced the need for different network structures in different phases of processes.)