An Ode to the Network Periphery
Social network research often focuses on the core of a network instead of the periphery. There are practical and theoretical reasons for this. The practical reason is that it is often difficult to measure the periphery of the network, for example peripheral actors are less likely to answer a survey or be mentioned by survey respondents. The theoretical reason is that many people think all of the “action” is in the core. For example, in policy networks, the core actors might have the most political resources and therefore have control over how policy decisions are made.
But perhaps we really should think more about the periphery. The periphery does exist, even if it is hard to measure, and social actors are in the periphery of the network for some reason. More importantly, there are probably important relationships between the core and periphery that we have not fully explored. I’ve had several ideas when I think about this issue in the past, but I think a conceptual approach that may integrate these ideas is to think about the periphery of a network as having some type of disruptive potential that can reshape the core, leading to institutional change. Then the question is when is this disruptive potential realized?
The disruptive potential could be thought about in different ways. The periphery actors could be excluded from the core because they don’t agree with the values or policy preferences of the core actors. The core actors might spend resources to erect barriers to entry in order to protect their established institutions and distribution of power. In environmental policy, it is often the case that purportedly “inclusive” collaborative partnerships actively discourage the participation of more radical environmental or conservative groups that cannot “play nice with others”. Excluded peripheral actors may also create environmental justice issues; even if disadvantaged communities are not actively excluded from the core, their lack of capacity makes it hard for them to engage in the networks. Such excluded peripheral actors may fight to get into the core, or what is probably more likely they will seek out other venues to pursue their interests and threaten established institutions of the original core actors. Environmental justice programs often seek to increase the capacity of the peripheral actors to become involved in the core network.
In the case of cooperation problems where group formation is dynamic, the periphery actors might be the defectors. The work of T.K. Ahn is instructive here. In his experiments, when actors are allowed to choose their partners in cooperation games, all of the cooperators cluster together while the defectors are in the periphery of the network. T.K. calls this periphery “Nashville” because the periphery actors are playing the one-shot Nash equilibrium strategy of defection. In evolutionary simulations, such Nashville defectors can invade cooperative groups where random drift has allowed the development of altruist strategies that do not punish defection—neutral mutants. Given the right mix of conditional cooperation strategies and neutral mutants, the overall group can be invaded by defectors and experience a punctuated equilibrium to low levels of cooperation.
A more positive viewpoint is to think about the peripheral actors as a source of innovative ideas and policy learning. Core actors often have high levels of network closure (transitivity), which can lead to redundant ideas and group think. Periphery actors might be less subject to redundant ideas, and they could be engaging in lots of policy or other types of experiments that would be beneficial for the entire system. Think about states and cities experimenting with policy innovations, which are then picked up by the national government and translated into some type of national law. Most of our national environmental laws follows this pattern, and climate policy is the most current example. In this case, the core actors would be wise to expend effort to actively engage the periphery in order to tap in to these new ideas. Integrating the new ideas into the activities of the core network would in theory improve the performance of the overall system.
Regardless of whether the more negative or positive views of the periphery actors is correct, when a stronger linkage between periphery and core is established, disruption is likely. In addition, the distribution of disruption events will not follow a normal curve. It is likely to have more of a punctuated nature, with lots of failures and rare successes where some aspect of the periphery leads to a significant change in the core. There are many processes in the core of network that provide fundamental stability and change only incrementally. It is only the rare occasion where a disadvantaged actor manages to change how power is distributed, or a new idea fundamentally changes the perceptions of the core actors and hopefully improves the performance of the system. Network researchers should spend some more time thinking about core-periphery dynamics, from both a theoretical and empirical standpoint.