Why policy networks don’t work (the way we think they do)

Published on 25 September 2019

Image of James Georgalakis
James Georgalakis

Director of Communications and Impact

The UK policy network that responded to the Ebola crisis which struck West Africa between 2013-16 far from being one cohesive community of experts and government officials may have been three competing groups. Despite successfully increasing political commitment to the social response to the epidemic, in areas such as overcoming community distrust, social scientists remained relatively weakly connected to the government embedded biomedical community. At the 8th PhD Conference on international development at the Institute of Development Studies I shared the initial findings of a study that applies a critical lens to what was widely reported as a positive impact story for UK researchers and policymakers. How transformative and trans-disciplinary was the policy network that responded to the epidemic?


Link to interactive research policy network map


Is it who you know or what you know?

The literature on evidence uptake and the role of communities of experts mobilised at times of crisis convinced me that a useful approach would be to map the social network that emerged around the UK-led mission to Sierra Leone so it could be quantitatively analysed. Despite the well-deserved plaudits for my colleagues at IDS and their partners in the London School of Hygiene and Tropical Medicine, the UK Department for International Development (DFID), the Wellcome Trust and elsewhere, I was curious to know why they had still met real resistance to some of their policy advice. This included the provision of home care kits for victims of the virus who could not access government or NGO run Ebola Treatment Units (ETUs).

It seemed unlikely these challenges were related to poor communications. The timely provision of accessible research knowledge by the Ebola Response Anthropology Platform has been one of the most celebrated aspects of the mobilisation of anthropological expertise. This approach is now being replicated in the current Ebola response in the Democratic Republic of Congo (DRC).  Perhaps the answer was in the network itself. This was certainly indicated by some of the accounts of the crisis by those directly involved.

Social network analysis

I started by identifying the most important looking policy interactions that took place between March 2014, prior to the UK assuming leadership of the Sierra Leone international response and mid-2016, when West Africa was finally declared Ebola free. They had to be central to the efforts to coordinate the UK response and harness the use of evidence. I then looked for documents related to these events, a mixture of committee minutes, reports and correspondence , that could confirm who was an active participant in each. This analysis of secondary sources related to eight separate policy processes and produced a list of 129 individuals. However, I later removed a large UK conference that took place in early 2016 at which learning from the crisis was shared.  It appeared that most delegates had no significant involvement in giving policy advice during the crisis. This reduced the network to 77.

I produced a social network using software that treated the co-occurrence of two individuals in the same policy interaction as a connection between them. Their mutual participation in say the UK Government’s  Scientific Advisory Group for Emergencies (SAGE) was used as a proxy for a social tie. This produced a matrix which I could then run further analysis on. This included calculating centrality values for each node   – essentially a popularity algorithm that measures how well connected to other well-connected nodes everyone is. I also ran community identification analysis that works out the minimum number of clusters or sub-groups in a network. The results both validated some of the accounts of the crisis and produced some surprises (at least for me).

Three networks not one

It was easy to confirm who the key people were who bridged social scientists to biomedical experts and to government officials and humanitarians. What struck me was that just four of those in the entire network were based in Africa. That includes DFID Sierra Leone staff. However, what was really interesting was that the social network analysis (SNA) produced three distinct sub- networks. Each of these enjoys very high levels of homophily (their members are really well connected to one another) but each group is weakly connected to the other two. Just six people bind the whole network together. There is a group to which almost every anthropologist and medical anthropologist belongs (all but one of whom are based in a university), although it does include some other individuals from DFID, humanitarian agencies and a few epidemiologists.

This is the only part of the network that includes social scientists that you could truly call transdisciplinary (working across disciplines and other stakeholders). Another group is dominated by the biomedical community who are well embedded in government agencies like Public Health England. The third is similar to the second but is more diverse, including humanitarians, multilaterals and international policy actors.

Personal and professional relationships

My sampling of nodal policy events fails to capture the many informal meetings that took place – some of which must have been really important. It could be argued that using membership of these interactions is a weak proxy for real relationships. I am not suggesting the network represents the actual flow of knowledge. At best it infers it may have taken place in some instances. However, when you study the network membership carefully you’ll see how most of the connected nodes were probably already connected before the crisis occurred by similar academics careers and fields of research or common organisational or sector membership. Some are even related to one another. What we are seeing are traces of a very real social network.

Nerds not nodes

When anthropologists, along with their small number of allies from other disciplines or government agencies, tried to make the case for home care they came up against a far more embedded biomedical epistemic community. They did really well to operate in this space and they have had a longer-term impact on attitudes towards the social response to infectious disease outbreaks, particularly in the WHO and other multilateral agencies. However, their weak connection into the more established advisory groups limited their influence. This is where network theory comes in. Some fascinating studies have suggested that network homophily actually augments power asymmetries in policy networks.

To explain this simply think of those American High school movies when a nerd makes friends with a jock. The other nerds may become excited and feel empowered by this and the jock may even gain some kudos by showing their softer side. But ultimately this relationship won’t get the nerds all invited to the school prom (sorry in this analogy the nerds are the social scientists) or change the attitudes of the jocks towards the nerds. You create the illusion of connectivity when it’s limited to just a handful of connectors. This results in sub-groups remaining even more tightly and exclusively tied together.

The dichotomy of networks

What everyone came up against, including many clinicians and virologists who were also deeply concerned about the need for a more transdisciplinary response, was the natural tendency of like to attract like in social interactions. As your most popular nodes connect with other popular nodes the rest are left to remain connected to their own kind. The key bridgers, who worked wonders to bring these communities together at all, could not overcome this overall network dynamic. This is the potential dichotomy of networks. How can they both conform with Peter Hass’s theory that they mobilise outsiders to get a seat at the table whilst also protecting normative beliefs and dominant policy frames?

Disrupting policy networks

For the UK Ebola network you can isolate some decisions that may have directly increased the homophily of the three separate sub-groups, thereby weakening the network as a whole. Just consider the decision to have just one social scientist on the prestigious Scientific Advisory Group for Emergencies (SAGE), which reported directly to Number 10 (which at the time seemed like a real success) and then set up a sub-group for the social scientists to meet separately. The SNA analysis would have returned significantly different results if more social scientists could have joined SAGE.

There is no easy solution and I am not suggesting everyone goes to every meeting and decides everything by committee (we know how that ends). Ebola killed over 11,000 people in the West African epidemic and it is smouldering on right now in the Democratic Republic of Congo. With predictions of ever more frequent and dangerous infectious disease outbreaks  we should at least be talking about ways to disrupt conventional policy networks.  Left unchallenged they can limit the impact of knowledge which runs counter to dominant normative beliefs. Once you accept the game is fixed you are in a much stronger position to try and change the rules.


About this opinion

Related content