Citizen Participation: are Human-Machine Networks the solution?

Photo: JLogan – Own work, Public Domain,

It is a common cry, particularly around the time of elections and referenda, that ordinary citizens don’t vote and they don’t engage directly with their elected representatives. Many observers now turn to the power of social media hoping to solve what is seen as a problem of democracy.

They want to encourage ordinary on-line citizens to get more involved in the political scene surrounding government and policy that affects them. Although this is the hope, it is more difficult to initiate than at first appears. People talk to each other over social media, they broadcast their opinions to the world over twitter and they track current affairs as reported by news websites, but they don’t engage directly in the democratic process. If Twitter and other social networks have the power of influence  [1], then by its agency citizen participation can potentially move to another level. And we are not just thinking here about the human participants in the network, but also about the role played by machines and their potential intervention in network-based exchanges [2-4].

In this context, we can validly ask the question of where we think that citizen participation HMNs are headed. We need to avoid generating suspicion around unexpected government use of online engagement [5-8], and go back to those most directly involved in citizen participation, asking questions like: Who is involved? Who needs to be involved? And what do they expect from the network?

With that in mind, we approached attendees from a series of related workshops. In response to the questions we raised via anonymised survey, they told us:

  • Online engagement is common: digital literacy is not an issue;
  • Lack of governmental involvement is a frustration: citizens will talk, but they expect to be listened to;
  • It’s the citizens themselves, and not just government organisations or their representatives that are regarded as key players: what citizens are talking about is important.

Some of the responses we got back were as we might have expected, but some were not. Trust is obviously the number one issue in everyone’s mind. The public don’t trust politicians to listen to them. And the politicians recognise that trust is hard to gain when the whole business of politics is the conflict between differing proposals and ideologies. But what was interesting is that the highest ranked motivation to improve participation from all parties was to create a culture of engagement amongst the public, so that they don’t feel that such engagement with politicians is pointless or ineffective. This understanding tells us where the network needs to be. We can therefore begin to map out the citizen participation HMN, but significantly, starting from a target destination for the roadmap, and not where it’s starting from.

If we know where we are trying to get to with the roadmap, then we need to know how we intend to get there. With that in mind, the HUMANE methodology provides valuable insights. The dimensions, as described by [9], are the levers that can be adjusted to effect change in the network. The levels of human and machine agency, the tie strength between actors, the network extent and the network structure can all be developed and adjusted to create different network outcomes such as the meta-dimensions of trust and motivation – clearly the most important identified drivers for success. You cannot tell people to have more trust, but by changing the levels of agency and tie strength so communication and self-efficacy can be improved, potentially leading to greater trust and confidence. Similarly engaging others through changing the extent of the network and the structure, allowing participants more freedom for communication paths (bottom up) rather than dictating behaviour (top down) has the potential to improve motivation. These strategies – which include descriptions of when and how they may be of use – provide on the one hand indications of the types of challenges which the HMN may face in reaching the goals identified by stakeholders in our survey, and on the other, ways to circumvent or handle those challenges.

Bringing together our survey results with HUMANE allows us to approach roadmapping for citizen participation HMNs in a rather different way. We are less concerned with where we are now. Instead, using HUMANE outcomes, we can form an understanding of where actors in the network want to be and offer them guidance and strategies to achieve their goals and overcome potential hurdles along the way.


[1] W.L. Bennett, “The personalization of politics political identity, social media, and changing patterns of participation,” The ANNALS of the American Academy of Political and Social Science, vol. 644, no. 1, 2012, pp. 20-39; DOI 10.1177/0002716212451428.

[2] D. Murthy, et al., “Automation, algorithms, and politics| bots and political influence: a sociotechnical investigation of social network capital,” International Journal of Communication, vol. 10, 2016, pp. 20.

[3] A. Bessi and E. Ferrara, “Social bots distort the 2016 US Presidential election online discussion,” First Monday, vol. 21, no. 11, 2016; DOI 10.5210/fm.v21i11.7090

[4] P.N. Howard and B. Kollanyi, “Bots,# StrongerIn, and# Brexit: computational propaganda during the UK-EU Referendum,” Browser Download This Paper, 2016.

[5] S. Colombo, “The GCC and the Arab Spring: A tale of double standards,” The International Spectator, vol. 47, no. 4, 2012, pp. 110-126; DOI 10.1080/03932729.2012.733199.

[6] P.N. Howard, et al., “Opening closed regimes: what was the role of social media during the Arab Spring?,” Available at SSRN 2595096, 2011.

[7] C. Stott and S. Reicher, “Mad Mobs and Englishmen?: Myths and realities of the 2011 riots,” Book Mad Mobs and Englishmen?: Myths and realities of the 2011 riots, Series Mad Mobs and Englishmen?: Myths and realities of the 2011 riots, ed., Editor ed.^eds., Robinson, 2011, pp.

[8] K. Glasgow and C. Fink, “Hashtag lifespan and social networks during the London riots,” Social Computing, Behavioral-Cultural Modeling and Prediction, Springer, 2013, pp. 311-320.

[9] A. Følstad, et al., D2.2: Typology and method v2, 2016.