Today I exchanged emails with a colleague with whom I am involved in an evaluation project about knowledge flows and we found that in the same country and among very similar actors the information in one domain flows very much in a hierarchical, hub-and-spoke kind of fashion, the responsible Ministry sits in the middle and informs everyone else, mainly in one way communication. In the next domain the network is much more interconnected, while there are some more and some less connected actors, basically everyone has more than one source of information and there is much more exchange between actors on similar hierarchy levels.
Our intuitive initial reaction was: One structure has to be bad, the other one has to be good. And because of where we are comming from (our views of the world), we know that the hierarchical information distribution structure is bad and the interconnected web of information exchange is good. Now if we look beyond what we like or prefer for the sake of empowerment and just ask: How well suited are these structures to get the information where it should go, the picture is more mixed and I have to give an answer that you will hear from me whenever you ask me about what the best network structure is: That depends. Both structures have pros and cons.
The hierarchical structure where one central node controls all the information flow are great for distributing clear and undisputed information in an efficient and effective manner, everybody knows: I have to go to the Ministry to get the right information and the Ministry has total control over the content of the message. On the flip side however, this puts a large burden on the Ministry (or whoever the hub is), because if they don’t perform, no one can take their role and the system will collapse. And there are many ways in which they might fail: their capacity might be overwhelmed with the sheer number of requests, they might have outdated information, they might focus on another issue, etc. Also, innovative ideas, learning from field experience, experimentation and alternative solutions are not encouraged in a system that has pre-defined who owns and controls all relevant information.
A less centralized structure with stronger inter-connectivity and lateral flows is much less vulnerable to one actor’s lack of performance, allows for more cross-pollination and the integration of alternative approaches. On the other hand, a lot of the actors on the ground, who just want to quickly get the relevant information and then get the job done, are easily confused by multiple, contradictory messages and might not always be experts enough to judge which one is the most valid one for their work. Also, less centralized networks tend to take much more time to mobilize, there is no one actor who can take on the responsibility for training everyone. Also it is more difficult to monitor, evaluate and compare the effects of the different interventions that abound. So while there might be a lot of experimentation, that doesn’t automatically lead to learning and innovation – it might just be a lot of reinventions of the wheel. Often low centralization networks do not survive and thrive for a long time, as there is no central driving force (though, sometimes they do…).
So, as far as ideal network and recommendations go, I’d say, neither is ideal. They both have strengths and weaknesses. If you work in a highly centralized network you have the benefit of knowing who to interact with to get your message out. And your role as NGO could be to make sure they have cutting-edge information to start with, to encourage this central actor to allow for more two-way information flow and to expand the core of the network, inviting more actors to share their burden. However, this has to be done delicately, as the hub might fear loosing control and power. On the other hand, if working in a dispersed, low centralization network, you want to see if this is really better for the front line implementers in terms of enabeling them to get their job done. Especially if there is a lot of confusion around contradicting messages, your role as an NGO coming in might be to help the different actors coordinate and consolidate and develop more predictable ways of defining messages, delivering information and facilitating comparable monitoring and evaluation.