FROM POLITICAL TURBULENCE TO CHAOTIC PLURALISM
We started this book with a question: how does the changing use of social media affect politics? This line of enquiry was suggested by the dramatic rise in the use of social media from the mid-2000s, built on the previous decade of expansion of Internet-enabled activities and fuelled by the widespread adoption of mobile telephones connected to the Internet. Such technologies have changed the way most people work, shop, and travel and how they manage their financial affairs, educate and entertain themselves, keep in touch with friends and family, and get involved in politics and interact with government. Of course, the medium itself could be entirely neutral, with electronic communications simply replacing activities that were once done on paper, by telephone, or in face-to-face contact. Even with this rapid adoption of new technologies, the same institutions and behaviours might carry on as they did before, just with the added convenience afforded by electronic forms of communication. But this has not happened. Instead, the Internet and social media are now inextricably intertwined with the political behaviour of ordinary citizens. So what can we say about the changing nature of politics?
The kind of politics we have observed and analysed is characterized by rapidly shifting flows of attention and activity. Social media extend the range of political activities that citizens can undertake, lowering the costs to an extent whereby people are offered the opportunity to make micro-donations of time and effort to political causes throughout their daily activities. In every decision over whether to undertake these tiny acts of participation, people are exposed to social influences: the knowledge of what other people are doing, or the knowledge that what they do will be visible to other people, so their own actions are interdependent with those of large numbers of other people, causing chain reactions that can scale up to large mobilizations—but usually do not. This is a turbulent politics, which is unstable, unpredictable, and often unsustainable.
In 2013 two countries started to experience the waves of protest and demonstration that typify this kind of political mobilization. In Turkey, demonstrations against plans to build a shopping mall on the Gezi Park in Istanbul grew into a sustained protest against the increasing authoritarianism of Prime Minister Recep Tayyip Erdogan. Two weeks of protests climaxed in a drastic response by riot police in which five people died; demonstrators sustained a silent vigil, with the ‘standing man’ Erdem Gunduz standing silently for eight hours in Taksim Square, a protest copied many times over across the country. In Brazil, as we noted at the start of Chapter 1, rallies that started in Sao Paulo in mid-June, triggered by a rise in bus and underground fares, swept across the country to around a hundred cities. Although the government reversed its original policy on transport pricing, the protests continued, based on a more generalized objection to a whole range of issues such as rising prices, continuing inequalities in society, and irrational spending on bringing the 2014 World Cup and 2016 Olympics to Brazil. Social media were implicated in the spread of both these movements. In Turkey, the Facebook page ‘Diren Gezi Parki’, set up to coordinate events, had 622,000 ‘likes’ by mid-June, with 707,000 ‘talking about this’, while an app to allow private messaging on a virtual network gained 120,000 users in a week. Even from Gezi Park itself, over 15,000 users sent at least one tweet during the first five days of the protest.1 In Brazil, demonstrators carried banners proclaiming ‘We are the Social Network!’ Members of both movements described themselves as leaderless and resisted any institutional involvement, from political parties for example. Both started with relatively minor issues, but quickly grew in support, reflecting a more generalized dissatisfaction and greater distrust of the government and political leaders. Both appeared to melt away from global news in the year that followed, but could reemerge at any time. Both cases illustrate strongly all the trends and patterns we have identified in this book.
In this, the final chapter of the book, we outline our account of contemporary collective action, using the experimental and large-scale data analysis presented in the previous chapters to shed light on the dynamics of political turbulence. In so doing, we hope to provide the reader with a way of understanding mobilization in the twenty-first century. We investigate pluralism as a democratic model, identified in Chapter 1 as the best bet for the theoretical framework to encapsulate the changing form of politics. From the turbulent political world we have explored here, we seem to be able to discern a new kind of pluralism, highly decentred and chaotic in its operation, that might revive strands of thought from the earliest pluralist thinkers of the previous century, such as Dahl and Lindblom, while being based on a transformed notion of many of its principles, particularly the conception of interest groups as the basic building block of society and the state. We believe that the turbulence of contemporary politics is best captured by the term ‘chaotic pluralism’. We explore the policy implications of this democratic model for the future, looking at possible state responses to the challenge of turbulent politics and the design options for groups seeking to maximize civic engagement and political participation. Finally, we assess the contribution this book makes to the research base and to the methods currently used in social science.
WHAT WE HAVE FOUND OUT
What have we found out? We have identified the capacity of social media to enable contribution of micro-donations of time and money that can scale up to mass mobilizations. These tiny acts of political participation may seem insignificant on their own, and they have been dismissed by sceptical commentators as mere ‘slacktivism’—the idle and aimless tapping of keys and pressing of buttons with only vague political aims. But they represent a growing form of political participation, which in some countries and contexts is overtaking voting as the political act that people are most likely to undertake. They alter the costs and benefits of political action, reducing the transaction costs when compared to the costs of getting involved. When set against the possible benefits of political action, such a reduction of costs can tip an individual decision over into participation rather than apathy. When carried out by millions, as in the revolutions of the Arab Spring or the mass demonstrations in response to the 2008 financial crash, these tiny acts of participation are bringing about social and political change.
Social media platforms—by providing what can be termed zero-touch coordination for these micro-donations of time, effort, and money—are replacing organizations and institutions in some areas of political life. Indeed, organizations increasingly resemble social media platforms in the way they present themselves to the public, with facilities for commenting and encouraging the sharing of content on social media. These new forms of political participation are individually based, unmediated by organizations or institutions in at least the early stages, although the rules that private companies impose are a hidden form of mediation of course. We see this as an individualization of collective action in that individuals can quickly and easily make some small contribution to a political campaign or mobilization without interacting with the personnel of organizations and their representatives, such as politicians and interest group leaders. In this political landscape, citizens may express multiple allegiances to issues and support a heterogeneous range of causes without formal membership of any group.
By far the majority of these mobilizations will fail to attain any kind of measure of success, disappearing with little trace within hours of their initiation. But those few that succeed do so extremely rapidly, attaining critical mass and tipping over into large-scale participation. As we showed in Chapter 3, they seem to defy the expectations of formal political science theory, by showing no gradual buildup in a classic S-shaped curve, with an increase, maximum, and decrease of acceleration in the curve of changes over time. Rather, they shoot up and peak in the earliest stages, rendering them less predictable than any smooth buildup of support.
This dynamic generates a pattern of mobilization that has been shown in other areas of political science research, in particular the punctuated equilibrium model of Baumgartner and Jones. They show a leptokurtic or extreme distribution of changes to the policy agenda and other arenas, which implies that otherwise similar mobilizations can differ hugely in their level of success, an almost random allocation based on their falling just over or inside the tipping line. We see this pattern for mobilizations initiated on social media, suggesting that they contribute to punctuations in the policy agenda. That is, where political mobilizations using social media fail, they do not interrupt the period of stasis, and on the few occasions where they take off, they initiate or contribute to punctuations. This is a new ecology of mobilization, which promotes an even more intense process of competition, ensuring survival only of the fittest, with a large number of failures and a small number of unpredictable, extreme events. If this pattern is sustained across other areas of political life playing out on social media—and we have every reason to suppose it would be—then this kind of collective action will continue to act as an important influence on policy change. It will inject turbulence into every area of politics, acting as an unruly, unpredictable influence on political life.
In this book we have revealed the dynamics of these new kinds of mobilizations. We have tested the effects of two key forms of social influence that an individual experiences when deciding to participate using social media by simulating a social media environment in an experimental setting. First, we have analysed the effects of social information, real-time information about the participation of others. When levels of social information are high, they provide potential participants of a collective action with crucial signals of its likely viability that can encourage them to also participate. Second, we have looked at the effect of making individuals’ actions visible to others when they make their participatory decision. Both these forms of social influence can encourage individuals to participate. At the individual level, visibility has the greater effect. But when we look at the collective, social information proves to be the more efficient. Under social information, rational individuals are able to behave more strategically. They can target their resources where they are needed most, where they will not be wasted on goods that will never be provided or will be provided anyway regardless of their contribution.
These forms of social influence do not act in a uniform way but differently affect different types of people. In mainstream social science research, the traditional way to distinguish between people in terms of their likelihood of participating politically has been to look at demographics, which have, for example, been a consistently good predictor of voter turnout. Where costs of participation are very small, demographics will be less useful as a determinant of whether people participate or not. With these factors less salient, other factors, which were always present, can come to the fore. We have been inspired by recent political science research that uses personality as a useful predictor of political behaviour. We have identified the personality types that are resistant to these forms of influence and those that are particularly susceptible. We have shown that those with emotionally stable personalities consistently give more, while agreeable and conscientious people consistently contribute less—it appears that some disagreeableness is a precondition of political action. Behaviours determined by some personality types are relatively resistant to social information. But when it comes to visibility, we find large differences at the individual level. Agreeable people contribute more under visibility, while extravert people contribute less. In particular, those individuals who have a pro-self, or individualist, social value orientation contribute significantly more when they are visible, whereas pro-social or cooperative types contribute less. Individualists it seems are more shameable than we might have thought, whereas cooperative people can be put off participation by the same shaming mechanism. Such a finding can inform the design of online initiatives seeking to encourage civic engagement, matching design to constituency, for example, by using visibility when dealing with individualist groups, and social information for more cooperatively oriented communities.
We find that collective action is dependent upon the distribution of individual thresholds for action, elaborating and extending Schelling and Granovetter’s earlier work on threshold models. That is, we have taken some steps towards identifying the distinct thresholds people have for joining mobilizations. For critical mass to form, some people with low thresholds obviously have to join at the earliest stages, when there are no signals of viability, and their participation will send a signal to those with slightly higher thresholds. The personality type most associated with this willingness to join early (or to start) is extraversion, although a high internal locus of control can also play a part. A mobilization needs a minimum number of extraverts in the population to get off the ground. In contrast, agreeable people tend to have high thresholds and will join only in the later stages. Most people are likely to join somewhere in the middle. This distribution of thresholds will be crucial in determining whether a mobilization can assemble a chain reaction of starters, ensuring a succession of followers with gradually increasing thresholds that allows a campaign to become viable.
Such a pattern means that mobilizations can launch without leaders, needing only sufficient people with low thresholds who are willing to start when there are few signals of viability. That poses a real challenge for their sustainability. Many of the most prominent protest movements of the past decade have faded away. In many of the countries of the Arab Spring, even where the rebellions of 2011 were seemingly successful in forming critical mass and achieving revolutionary change (as in Egypt), they provided no leaders in the wings, and none of the organizational trappings of pre-Internet revolutions. The post-revolution period was dominated by the one long-standing organization that was present on the political scene of the region, the Muslim Brotherhood, even though their participation in the revolutions had been limited and their commitment to democracy questionable. In some countries, mobilizations have now started to engage with the mainstream, as in Spain where some elements of the Indignados protest movement formed the political party Podemos (We Can), now a viable electoral entity, with 8 percent of the vote in the 2014 European elections. Podemos gained 600,000 supporters on Facebook between May and July 2014 and by the end of October had 200,000 members, who could sign up on their website for free, challenging and posing a credible alternative to all the established political parties in Spain in an increasingly turbulent 2015 following the election of the left-wing Syriza party in Greece in January. Podemos candidates were extremely successful in mayoral elections in May 2015, particularly in Barcelona and Madrid: ‘the agendas of Spain’s two largest cities will be driven by the priorities of anti-establishment parties with roots in Spain’s indignado movement’.2 But even now their sustainability must be in question if comparisons are made with other antiestablishment parties in Europe, as in the case of the Five Star Movement led by Beppe Grillo in Italy, which gained 25 percent of the electoral support in national elections in 2013 but remained incapable of forming a parliamentary bloc in the Chamber of Deputies, the lower house of the Italian Parliament.
FROM POLITICAL TURBULENCE TO A CHAOTIC SYSTEM?
Our core claim is that social media, as an indelible part of the way people interact in the twenty-first century, are creating a new kind of politics, characterized by turbulence, meaning ‘full of commotion’ (in Latin) or a ‘state of conflict or confusion’. In fluid dynamics, turbulence is a flow regime characterized by chaotic property changes, including ‘rapid variation of pressure and velocity in space and time’, which seems to provide an adequate metaphor for what we have described. But there is a sense in which the idea of political turbulence could go beyond being merely a metaphor. Turbulence, or turbulent flow, is one kind of a chaotic system, a dynamic system with unpredictable behaviour despite the fact that its governing rules are known and deterministic. The reason for this is the high sensitivity of such systems to the initial conditions and slight perturbations while they evolve in time, and since measurements of the initial conditions of a system by the observer have errors (albeit very tiny), predicting the long-term behaviour of the system is impossible.
The classic example of a chaotic system is weather. In essence long-term weather forecasting is impossible, not because of unknown mechanisms but because of the sensitivity of the governing rules and dynamical equations to the initial conditions and parameters, such as wind speed and humidity, which cannot be measured with perfect accuracy. Even if precise measurement were possible, any future tiny perturbation, invisible to detection tools, could alter and totally transform the predicted scenarios. The famous example is known as the butterfly effect, a theory coined by Philip Merilees as ‘Does the flap of a butterfly’s wings in Brazil set off a tornado in Texas?’ The answer is possibly yes, although it is not too likely. This example indicates that a tiny perturbation in one system parameter, as small as a flap of a butterfly’s wings, could usher in a huge difference in the long-term trajectory of a nonlinear chaotic system like the global weather system. The important point here is that although the system looks random, it is not. As mentioned above, a chaotic system is deterministic in terms of mechanisms, but lack of exact knowledge of the initial conditions and tiny undetectable noises make the future behaviour of the system unpredictable to the observer. Chaotic systems are different to stochastic systems and quantum systems, in both of which the uncertainty comes as an intrinsic feature of the system, whereas the unpredictability of chaotic systems originates from the lack of precise measurements of the initial conditions magnified by nonlinear sensitive mechanisms.
The essential features of a chaotic system are non-linearity and a high degree of interconnectivity, which creates positive and negative feedback, causing changes to take place exponentially faster. Chaos theory has been successfully applied to explain a wide range of phenomena in physics, biology, and economics, including turbulent flow but also population fluctuations and financial crises. Simple chaotic models of social phenomena to explain many of the near-random and difficult-to-explain empirical patterns observed in the social world have proved elusive.3 Although one of the key conditions for chaos often applies, the complexity, high noise, and lack of long-running datasets have worked against their application in the social sciences. However, recent commentary points to a development that offers renewed hope, in that ‘new models inspired by complex systems build social systems from the bottom up; behaviour is simulated for individual agents, then taken to the aggregate level either through analytic methods or through explicit computer simulations’, providing a model that is less mechanical and that uses simpler, better-defined assumptions about small-scale behaviour, which may well become important in the future, even outside the ‘low-noise world of experimental physics’.4
The transition of our daily lives into a super-connected world embedded in and provided by online social media, with the very low (tending to zero) cost of participation in collective movements, could have brought political systems into a chaotic state, where small perturbations or micro-contributions could eventually lead to large social phenomena. The same micro-contributions would have been dampened and neglected in a pre-Internet society with high barriers to free communication and high transaction costs. At the same time social media and digital life records could also be of help in order to measure the real current state of a system more precisely to avoid being surprised by its long-term behaviour, unpredictable due to our lack of knowledge of its current state, to develop the kind of models that LeBaron proposes.5 In his book The Signal and the Noise, the political forecaster Nate Silver argued that just as meteorologists have become better at predicting the weather with a combination of long-running, large-scale data sources, and calibration through analysis of previous forecasts and actual meteorological patterns, we ought to be able to predict economic and social phenomena such as financial crises and elections, as he himself did with great success in the 2012 US elections.6 As these kinds of data sources become available through social media, such methods might be applied to other forms of collective action.
DEMOCRACY AND TURBULENT POLITICS
We need to understand not only the current form of political turbulence, but also its implications for democracy, representation, and governance. What model of democracy and representation helps us to understand what has happened to politics as a result of the greater use of social media—the empirical reality of turbulent politics and the emergence of a chaotic system? In Chapter 1, we expressed great caution about the idea of the Internet as a new public sphere, which has been particularly associated with Habermas’s idea of a network for communicating information and points of view, where citizens can express their opinions, deliberate, and formulate some kind of common view or will. The main ideas from his approach are based around the notion that the Internet provides a kind of public sphere for public discussion and deliberation, in the same way that the coffeehouses of Vienna provided the forum for Habermas’s original conception of the public sphere.7
We believe that the political world on social media that we have analysed here is too heterogeneous, too individualized, too chaotic, and too ill-suited to deliberation to represent the revival of the public sphere beloved of communication theorists.8 This is not to say that we accept Sunstein’s gloomy prediction of a fragmented public discourse of echo chambers in which individual citizens experience the political world as an extension of their own narrow pre-existing interests and concerns.9 Sunstein believes that the ability to control the content they see and follow would mean that people would chose to follow only organizations, information, and friends that reflected their own viewpoint, thereby reinforcing it. Rather than individuals being subjected to a diversity of views—one of the key elements of a democratic polity—they would not be challenged, simply existing in a small bubble of the ‘Daily Me’.10 But neither Sunstein nor anyone else has demonstrated any evidence for this self-exclusion applying more online than offline, while others have argued persuasively that the Internet is too vast, too dynamic, too interconnected for such a state of affairs to exist in practice, particularly in comparison with conventional media channels.11 As we discussed in Chapter 2, social media allow individuals to conduct their lives in a ‘time-based world stream’, pumping out and receiving information and social influence, in which they are exposed to many contradictory and overlapping currents of information, views, influences, causes, campaigns, and concerns that widen rather than narrow their political experience. Rather, our claim is that what has emerged is a new form of pluralism: what we call chaotic pluralism. In the following sections, we discuss why this is the case, and its potential for conceptualizing turbulent politics.
WHAT IS PLURALISM?
To make the argument for a more pluralistic politics, we need to set out the context and offer a definition. For that we need first to return to the intellectual debate of the 1950s and 1960s when pluralism became a dominant account of the operation of democratic political systems promoted by some of the key figures in post-1945 political science, such as Robert Dahl, David Truman, Charles Lindblom, and Nelson Polsby.12 The key idea of pluralism is that society is made up of many competing and varied elements, which form into groups and associations. This diversity penetrates and structures the operation of political institutions and the making of public policy, which limits the exercise of power by a small group or ruling organization and ensures that power is spread throughout the political system. The pluralism of the political system reflects the diversity of societies as a whole with their different elements and groups. In politics these interests tend to balance out in the struggle for influence. If one group is strong, then another will counter-mobilize so as to undermine the monopolistic exercise of power. Forms of pluralism can exist even in authoritarian states, as in what was called the state pluralism of the former Soviet Union,13 characterized by factionalism and the delegation of power to different elements within the bureaucracy. Pluralism, however, is best promoted by differentiated institutions and democratic mechanisms of liberal democracy, what Dahl calls polyarchy, whereby elections ensure that leaders will seek to balance out interests and aim for a winning coalition.14 There are multiple access points for interest groups to influence public policy, enabled by fragmentation of institutions, such as that caused by the separation of powers and federalism. In liberal democracies, policy emerges as a result of balance and the mediation of interests. No one interest is dominant.
Almost as soon as the ink was dry on these classic works, critics targeted pluralism as an unrealistic description of the operation of politics, most notably in its account of power.15 Even what appear to be pluralist interactions may disguise power relations that operate even when interest groups are negotiating with each other in the political process. The pervasive impact of economic and social inequality determines the allocation of resources and the nature of authoritative decisions that emerge from political systems. In fact, Dahl and Lindblom spent most of their careers worrying about the impact of economic inequality on political representation, and concluded that pluralism was itself one-sided as business tended to win or to be advantaged, even if societies could still benefit from the underlying diversity of interest groups and points of view.16 As E. E. Schattschneider put it much earlier, ‘the flaw in the pluralist heaven is that the heavenly chorus sings with a strong upper-class accent’.17 In response to these concerns, pluralism morphed into neo-pluralism, with more recognition of how representation worked imperfectly, favouring business and the wealthy.18
In spite of generations of political scientists poring scorn on pluralism, it has proved surprisingly resilient as a political concept, with writers often returning to it after decades spent in the opposite camp.19 In part, most thinkers and empirical researchers realize that pluralism is not just a description of the operation of political systems, but also an aspiration that could be promoted by reforms that give the right incentives and offer an appropriate institutional framework. Such an interest reflects a perception that the age of dominance of a few ideologies grounded in basic ethical principles is at an end: the world is simply too diverse to create much more than an operating consensus on the rules of political interactions, and such agreement on the rules might be all that can be achieved. Societies can agree on certain matters such as free elections and free media, but not much else. The question then becomes, in all this diversity and differentiation, what kind of political system is most appropriate for it? This search encourages political scientists to focus on incipient trends as an indication of a return to a more fluent form of politics, such as the pluralism of interest intermediation in the European Union.20
Pluralism has been rediscovered as the property of thinkers from the late nineteenth century and early twentieth who promoted older values of association and pre-state forms of mobilization. These pluralist traditions had been shut out in the age of ideological conflicts based on left and right, which had created powerful political parties seeking a monopoly of political action and control of the state. Such was the project of the English pluralists, William Morris, G.D.H. Cole, and Harold Laski, who argued that a more community-based mobilization could overcome some of the limitations of state intervention. In his later work, Paul Hirst picked up on these concerns and argued that the world emerging in the 1980s was becoming more diverse again, in a way that gives these older pluralist arguments more traction. He wrote, ‘Associationalism makes accountable representative democracy possible again by limiting the scope of state administration, without diminishing social provision. It enables market based societies to deliver social goals desired by citizens, by embedding the market system in a social network of coordinative and regulatory institutions’.21 The model of pluralism we see emerging is not based on such gradualism, but what these associational arguments show is that social and political change is organic, resting on developments that are outside the state. Like us, Hirst believed that associationalism would counter the logic of collective action, though in his case by rediscovering the social basis to political action rather than the transaction cost argument we pursue here.
More recently, the political theorist William Connolly has offered an alternative revisionist perspective on earlier pluralist work, developing a kind of postmodern pluralism reflecting the proliferations of ‘minorities of many types’, ‘a world of interlocked minorities’ of immigrants, alternative lifestyle movements, religious groups, feminist movements, and ethnic minorities, in which the national majority is a ‘symbolic centre consisting of fewer people than the sum of the minorities’.22 His view of minority groups appreciating the fallibility of their own beliefs and thereby coexisting in a system of partisan mutual adjustment is reminiscent of the early pluralists and also Hirst’s associationalism, but has further parallels with the political world we have described where ‘new political demands often unsettle existing configurations of identity’23 in unstable systems ‘marked by an element of internal unpredictability’.24 It is not necessary to adopt a postmodern perspective to be able to use these insights and to recognise them in the political world unfolding today where the drivers for change are embedded in society, outside the state. Although some have argued that Connolly’s pluralism represents a break with the pluralism of postwar American political thought,25 others have argued that in contrast, ‘Connolly’s work is best understood as the resumption and enhancement of a distinct canon of pluralism in American political thought’.26
WHY TURBULENT POLITICS IS PLURALIST
Even in the relatively short time since Paul Hirst or William Connolly wrote, the political world has continued to become more diverse and fragmented. The use of social media is magnifying this diversity and empowering different elements within it, helping to foster different kinds of movement that have their origins outside existing forms of representation and institutional structures.
The Internet and social media may embody some pluralist ideas in their ability to rebalance power relationships, blur organizational boundaries, enable bottom-up dynamism, and reclaim politics by society. The theories and models we have presented—through their emphasis on peer-to-peer social information, the influence provided by social media platforms, and the stressing of individual-level difference—represent a theory of society rather than the state, as indeed did pluralism. What is different about chaotic pluralism is the absence of groups as we know them: for every interest there will be some kind of mobilization—maybe even a whole constellation of mobilizations—but not necessarily an organized group, at least not one organized in conventional ways.
Given the clear potential of the Internet to facilitate the formation of networked groups, matching people with similar beliefs and preferences, lowering coordination and organization costs, and allowing new forms of mass mobilization based on weaktie associations, it is inevitable that it has been associated with a revival of pluralist thought. As early (in Internet time) as 1998, Bruce Bimber was writing of accelerated pluralism through the Internet,27 in which the Internet would contribute to ‘the on-going fragmentation of the present system of interest-based group politics and a shift toward a more fluid, issue-based group politics with less institutional coherence’.28 In his 2003 work, Information and American Democracy, Bimber developed the idea of post-bureaucratic pluralism, where ‘the structure of collective action is less tightly coupled than in the past to a marketplace of formal political organizations’,29 but one that would be constrained by the market dynamics of the traditional mass media, and the structure of the state apparatus itself: ‘the need for collective action organizations to orient themselves to the structures and processes of largely unchanging institutions of the state creates limits on the advantages of postbureaucratic forms’.30
What is more surprising is that there has been little attention paid to a revival of Internet-fuelled pluralism as a model of democracy since then, far less than there has been for republicanism, even as social media have emerged with the potential to challenge traditional political institutions and overcome some of the earlier constraints on postbureaucratic pluralism. Bimber’s most recent book,31 with its plea for bringing organizations back into collective action, mentions neither pluralism nor Dahl, in spite of his early and prescient predictions noted above. A 2014 collection of articles on Online Collective Action, many using some of the methodological approaches that we have explored here, makes no mention of pluralism or pluralists.32 It seems as though pluralism in the age of social media has diverged too far from the original pluralist dream, perhaps because it was seen as too disorganized, disordered, and chaotic to be viewed as a coherent model of politics at all. The early notions of pluralism, such as those of Dahl and Lindblom, were ordered and organized; pluralists had a vision of a system of interest groups applying continual pressure on the state, but each fulfilling a distinctive need and engaging in a process of mutual accommodation. Bimber’s theory of ‘accelerated pluralism’ was based on the idea that ‘the processes of group-oriented politics will show less coherence and less correspondence with established private and public institutional structures’, but that this process (1) would still be based on the idea of issue group formation and action and (2) would not alter the overall interest of individuals in public affairs or their ‘ability to assimilate and act on political information’.33 He analysed the history of assimilation of other technologies (such as television into society), to argue that there was no link between increased political information and political participation. Dismissing populist and communitarian arguments about the influence of the Internet on political life, he argued that the basic logic of pluralism would remain unchanged: ‘No less now than when David Truman wrote about American pluralism, Americans will associate themselves in groups, and structure their political participation and engagement through those associations’.34 This still sounds like old-school pluralism.
In contrast, we emphasize the breakdown of existing structures and organizations, such as interest groups. We focus on the speed of interaction and rapid growth (but also decay) of mobilizations. This changed environment puts government on the back foot, disrupting traditional institutional strategies for engaging with society. The disordered and unpredictable nature of this chaotic pluralism, which makes some contemporary societies harder to govern, is one of the key weapons of those people who have mobilized with social media. Just as an earlier generation of political activists thought non-violent tactics would disarm those in power, such as the flower in the rifle butt as a symbol of non-violent protest in the 1960s, so tweeting and liking have become the instruments of the new turbulent politics.
CHAOTIC PLURALISM
We use the term ‘chaotic pluralism’ to reflect a key divergence between the changes we have observed, and either pre-Internet pluralism or Bimber’s accelerated or postbureaucratic pluralism. Classic notions of pluralism were group-based and surprisingly ordered (particularly the English version, where neat and tidy guilds and unions provided a structured environment within which political activism could take place). Even Paul Hirst’s revision for the post-1989 era, associationalism, is characterized by a commitment to gradual reform. Corporatism, which is an account of the tripartite institutions that help govern relations between state actors and representatives of producer groups, saw politics as an orderly process of mutual accommodation among state, unions, and business. This practice is very much in retreat, accounting for the formal relationships only in some European states. This aspect of pluralism in no way represents the vision we have presented here, where mobilizations spring from the ground, facilitated by technological platforms rather than organizations, and where organizational boundaries are blurred and sometimes disappear altogether. We have in mind a variant of pluralism that reflects the degree of disorganization and speed of change in political mobilization that has taken place, more aligned to Connolly’s postmodern pluralism or the radical pluralism of William James, based on the idea that ‘the overlapping forces propelling the world are themselves messy. Pluralism is the philosophy of a messy universe’.35
A pluralist pattern is emerging, but rather than being based on stable, ordered forms of association, it is characterized by mobilizations that spring from the bottom up, highly reactive to events. We call this model chaotic pluralism to build on the idea of political turbulence as a chaotic system discussed above. Such a model conceptualizes politics as a natural system and uses scientific models of chaos theory in natural systems to understand its operation, employing the kind of big data or social data science methods we have used in this book. In this model, the tiny acts of political participation that take place via social media are the units of analysis, the equivalent of particles and atoms in a natural system, manifesting themselves in political turbulence. The laws that guide them, however, derive from social influences exerted by social media and the heterogeneity of people who undertake these tiny acts. A model of chaotic pluralism will require also a social science understanding of human behaviour, including personality, social influence, and all the other things that shape political preferences and willingness to act upon them. Such understanding cannot come from big data alone, as such sources tend to be stripped of any information about the people involved and without benchmarks or control groups for comparison. To understand this piece of the pluralist puzzle, we need the second key method we used in this book: experiments, which hold out the possibility of causal inference. Experiments can allow us to understand how different people behave differently when they use social media, how series of chain reactions can form, and why some mobilizations succeed while most do not. We hope that the experiments used here have begun to show how this method can reveal the mechanisms at work in contemporary, chaotic pluralism.
There are pointers to this melding of natural science and social science perspectives in the recent revival of pluralist thought in political theory, outlined above. Connolly argues that the latest developments in evolutionary biology, neuroscience, and complexity theory show how ‘the emergence of new formations is irreducible to patterns of efficient causality, purposive time, simple probability, or long cycles of recurrence’, and therefore work against conventional methods of understanding politics;36 that ‘we find models of evolution where open systems possess powers of self-organization, and seemingly minor variations can initiate distinct new trajectories so that change cannot be fully predicted’; and that ‘we need to come to terms with the idea that chaotic and unpredictable behaviours represent essential features of the evolution of political systems, and develop models of multi-linear, complex and emergent causality, and correspondingly intricate modes of inference’.37 However, in his postmodern pluralism, operating at the intersection of neuroscience and cultural theory, Connolly does not discuss, as we have here, the mechanisms by which these developments play out in political systems, the relationship between pluralism and societal use of technology, or the empirical approaches that might be used to research them. In this book, by showing how social media play a role in political change and by highlighting some methodological approaches and modelling techniques from the natural sciences, we offer a more concrete framework for researching this turbulent and unpredictable political world.
WHAT FOR THE FUTURE? THE LIMITATIONS OF POLITICAL TURBULENCE AND CHAOTIC PLURALISM
To conceptualize political turbulence within a democratic model of chaotic pluralism poses a number of challenges. We need to be alert to the problems of classic pluralism, in particular the tendency for interest intermediation to be an unequal contest, benefiting those who have more resources and marginalizing others. These are not necessarily overcome just because the form is different to whatever happened or was conceived of before. The problems may reappear in the chaotic successor to pluralism. But we think these problems are less severe and may return our account of the world closer to those claimed by the early pluralists, and from which they retreated partly in recognition of the distortions provided by big business and the state with its corporatist and powerful institutions. Here we look to the future of the three main problems of pluralism: inequality, the power of the state, and the collective action problems of large groups.
First, inequality remains, even under chaotic pluralism. The classic accounts of political participation stress the importance of wealth, education, and parental socialization,38 which reduce the access of the poor and those without cognitive resources to the political process. Such measures of socioeconomic status predict every kind of political act from voting, through membership of parties and interest groups, to complaining.39 They are less influential in the dynamics of the mobilizations we have studied here because we have concentrated on those who use the Internet and social media, that is, people who are already likely to be in more economically privileged groups. Such an emphasis is, we feel, justified here by our aim to identify trends and patterns in contemporary collective action in the context of still-rising Internet penetration, where we can expect usage to reach that of the mobile phone (that is, near universal in developed nations and rapidly growing in developing countries) at some unspecified point in the medium-term future. But of course, in spite of growing wealth in Africa and Asia, inequalities remain in the social media age and will still shape collective action behaviour. And the Internet itself introduces new forms of inequality, such as those based on speed and cost of Internet access, and different levels of Internet expertise and skills. Indeed, as discussed in the book, many Internet-based activities exhibit power-law-like distributions, where attention is focused disproportionately on a small number of points, which could also increase inequality, as happens in natural systems at critical point.40 But, we argue, the Internet will remain a force for redistribution of power because of the ease with which it assists mobilization, given that it can ensure that groups with the most resources do not always win, partly because the Internet can harness the power of large numbers of people making micro-donations. Because the costs of participation and transaction are lower, the online world increases the access of the poor to political action and reduces the relative advantage of the rich. By how much we cannot know at present, but we are sure that the equilibrium of power will be different to the one that existed during the twentieth century.
The second big drawback of pluralism is that it tends to downplay the influence of the state. Yet the state remains a powerful actor in terms of structuring the progress of group negotiation, especially when producer groups are also powerful, such as in the corporatist framework, for example, whereby the powers of the state and business and worker groups reinforce each other. The onslaught of mobilization under chaotic pluralism has the potential to weaken the influence of the state in ways that might have been attractive to the early pluralists. Governments have been slow and inefficient in taking advantage of the Internet. Large-scale bureaucracy was well suited to earlier information technologies, and in the early days of computer technology from the 1960s, governments led the way in developing huge information systems to undertake administrative operations.41 However, as information technology became increasingly networked and interactive, bureaucratic culture in large states in particular mounted considerable barriers to innovation,42 exacerbated by troubled contract relationships with global computer service providers.43 With the Internet, the first information technology that is used by significant proportions of populations, governments have found it particularly difficult to innovate, lagging behind both private and voluntary sectors in terms of interacting or engaging with citizens, and being unable to respond to the swiftness of coordination enabled by social media. In the 2011 London riots, groups of looters armed with mobile phones running BlackBerry Messenger were able to outwit the technologically superior police forces, being able to coordinate their movements in real time. Similarly in the Arab Spring, citizens outmanoeuvred the less technologically savvy security services. States can fight back in ways that expose this limitation of the state’s role in pluralist thought. They will employ various tactics in an attempt to catch up, often by piling massive resources into Internet surveillance or, in authoritarian states, blocking or censoring online access. Now that technological innovation has become domesticated in everyday life, protestors (and criminals) are likely to remain one or more steps ahead. But in formulating a response, the state faces the challenge of how to do so in a way that is commensurate with the principles of democratic legitimacy.
The third limitation is the difficulty of sustaining mobilizations that lack institutions or organization, a key theme of this book. Social media have proved time and time again to be useful in overcoming some elements of the collective action problem, through the kinds of mechanisms that we have analysed here. But some of the classic features of collective action problems will continue in the age of social media because there will remain an asymmetry between those contributing to groups, such as those who provide more content over those who consume it, with collective action failures happening as leaders get bored or fed up with providing collective goods for nothing, and followers become disillusioned by the lack of policy responses to campaigns, protests, demonstrations, and even revolutions. And of course the Internet and social media have ushered in a whole cast of corporate actors to the political stage: Google, Facebook, Twitter, and so on. The potential of handing over power to civil society may be limited by Internet giants like Google, new stakeholders in public policy, who have started to take on some of the traditional activities of the state, such as counterterrorism, but with no democratic accountability. Will the chorus of the pluralist heaven now sing with a Californian accent? Our argument is that even with these tendencies the cost-benefit equation has changed, meaning that it is harder for any larger group to dominate for long. The speed of change affects corporations too, which can find it hard to stay ahead of the game. Witness Google’s attempts to break into this market with Google+, for example, or the continual process of mergers and acquisitions among social media companies as the popularity of social networking sites peaks and falls. Companies that were dominant may become so large they cannot innovate, often because they cannot keep up with the fast-changing predilections, fads, and interests of the very groups that use their products.
Finally, in addition to these three main limitations, the greater levels of instability that chaotic pluralism brings to political systems, by amplifying perturbations that otherwise would have been stamped out, have both advantages and disadvantages. The advantage is that an unstable agenda can benefit those who challenge older established agendas by offering new opportunities and ensuring that old routines and defences do not work; but the disadvantage might be that some depend on a degree of stability to follow and to shape the agenda, and if events are moving too fast those with good access and their finger on the pulse will benefit purely for those reasons. It may be that in spite of bringing new actors into the political process chaotic pluralism might exclude those without the time and resources to invest in following the twists and turns of the social media agenda. Again this is for future empirical investigation.
Certainly, political turbulence poses major challenges to both governments and citizens. For politicians, there is a tension in learning to deal with the current state of play in the new democratic climate in ways that are consistent with the objectives of democratic pluralism. For citizens making micro-donations to large-scale mobilizations that circumvent the need for organized groups, there is the challenge that in the end, institutions may be required to achieve sustained policy or regime change. And the mechanisms of chaotic pluralism do not lend themselves to institution building. The process of mutual accommodation, with the values of compromise and learning that underpinned the early models of pluralism, is extremely difficult to develop in a turbulent and disordered environment. If there are no leaders, how do groups bargain and compromise? The change engendered by the types of mobilizations we have discussed will be vulnerable to takeover.
HOW CAN STATES RESPOND TO POLITICAL TURBULENCE?
As the world observed the events in Turkey and Brazil in the summer of 2013, commentators in other countries noted that the same kind of dissatisfaction and unrest was also present at home: ‘A crisis of legitimacy could strike Britain too’ bemoaned the editorial of the Observer, indicating that ‘modern politics is not fluent in the vocabulary of shifting power and empowering citizens’.44 Clearly, the first step is to understand these new forms of collective action, even without the institutional road map of earlier forms of pressure and protest.
The second step is to take what plays out on social media seriously, rather than viewing it as peripheral to the political system. Political leaders have shown a tendency to play with the language of social media, or to use it as a glossy veneer for their offline activities. In Egypt in the summer of 2013, amid mass protests and demonstrations in Cairo and other cities across the country, President Morsi tweeted to say that he was ‘with the protestors’, but his speech later to the state television channels showed him to be in a very different place.45 Likewise, after he had been ousted by the military, army leaders posted on Facebook on 5 July that they were not going to persecute the Muslim Brotherhood or other religious groups; but they then immediately started shutting down Islamist TV channels and rounding up and imprisoning Muslim Brotherhood leaders, while the president remained in prison. By these acts, both major protagonists showed that they did not take social media seriously as the venue in which the political future of the country was being decided. Rather, they saw social media as a way to feign that they were acting in the public interest and doing the right thing, not in the same league as conventional media. In contrast, the protestors in Egypt, Tunisia, Turkey, Brazil, and Hong Kong really are utilizing social media as the primary means of communication and coordination, and indeed for the general population there is early evidence to suggest that Facebook (for example) is overtaking conventional media as a trusted news source,46 particularly in countries where there are reasons to distrust mainstream media. Ironically, it seems that former President Mubarak of Egypt really understood the significance of social media, as he blocked the Internet a few days into the revolution, as did the Turkish prime minster in the events of 2013, when he described social media as ‘the worst menace to society’.47
States face a challenge in engaging with social media, in that they are bound to encounter scepticism as to their motive. In authoritarian states, engagement can be viewed as the development of an instrument to repel and quash demonstration and protest, and indeed this is often the case. Recent research has demonstrated this approach to be the key pillar in the highly sophisticated censorship regime developed by the Chinese government, whereby criticisms of the state, its leaders, and it policies are permitted on social media, which indeed are even used by the authorities to inform themselves about instances of corruption or incompetence within the state apparatus, but posts about ‘real-world events with collective action potential’ are censored.48 In democratic states, any government collection of data pertaining to citizens can reveal major weaknesses and lack of trust in government-citizen relationships. Take the furore in 2013 surrounding the whistleblower Edward Snowden’s revelations that the security agencies of the US and UK governments were collecting a significant proportion of all Internet traffic entering their respective countries. The response highlighted the public’s distrust of government with respect to their data, a general view that nothing good would be done with them. Such views are exacerbated by a tendency of political leaders to use big data for private rather than public goods; the UK Prime Minister David Cameron operates an alert system that uses social media to predict which policy plans are accumulating so much opposition that they will be politically damaging to implement, rather than indicating illegitimacy or underlying societal problems. Only by developing an ongoing programme of working with this kind of data can states hope to institutionalize this kind of activity.
One way that states might develop such a programme would be to use some of the data sources and approaches that we have used in this book. Data generated from social media can allow policy makers to monitor and understand undercurrents of public opinion and dissatisfaction. They can be used to identify weaknesses in government services, such as failing schools or hospitals with deep-rooted problems with cleanliness or management style. They can be used to work out when policies are illegitimate, or impossible to implement. They can also be used to conduct reviews of government agencies, and to inform programmes of self-improvement. For example, in 2014 the UK Department of Work and Pensions commissioned a feasibility study on the use of social media data for social research and analysis at the department, focussing on recent changes to the UK benefits system, carried out by some of the authors of this book.49 Likewise, the kind of randomized controlled trials (RCTs) that we have used here could be implemented to understand which policy and service designs work, and which do not. They can be used to illuminate the kinds of information environments that encourage citizens to engage with public policy and those that do not. In this way, experiments could be used to capitalize on citizens’ evident willingness to make micro-donations of time and money to improve public goods. There is new enthusiasm among policy makers to conduct experiments, particularly in the United Kingdom, where the Behavioural Insights Team first operated from the Cabinet Office during the coalition government of 2010 to 2015. The team carried out a number of RCTs to nudge citizens towards socially efficient behaviour, such as timely payment of taxes and fines. These experiments were partly inspired by Thaler and Sunstein, who in their 2008 book Nudge presented a range of ways in which the choice architecture of policy initiatives could be designed to nudge citizens to undertake a certain kind of behaviour, in a rather paternalistic way. There might be other, more empowering ways that allow citizens to undertake a more proactive role in policy design, using feedback, reviews, and recommender systems for example, and experiments would be the way to design such systems.
If governments were to conduct such a programme of using data and insight generated by social media to understand the impact of non-linearities and chaos in social systems, the payoffs could be huge. As LeBaron observes, ‘The control of nonlinear systems can actually be easier than the control of linear ones, because it might take only a small push to engender a big change in the system. In other words, small low-cost policy changes could have a large impact on overall social welfare’.50 The key to such payoffs could be using the same methodological toolkit for understanding turbulent politics as we have used in this book: the analysis and modelling of large-scale data, and greater use of experiments. Data generated by social media could be deployed by governments to understand trends and patterns in citizens’ needs, preferences, concerns, behaviour, and complaints, as a barometer of their own legitimacy (or illegitimacy), and to identify the warning signals of critical transitions. Chaotic pluralism could allow them to work with rather than against the grain of citizens’ willingness to act in pursuit of collective goods.
We hope that states do not react too rapidly or strongly to political turbulence, nor do we think they have the capacity to do so, outmanoeuvred as they often are. Citizens have much to gain in the world of chaotic pluralism through their new capacity to set the political agenda from outside the political system. In this sense, social media have unleashed a more citizen-based politics, which for all its turbulence can open up new kinds of transparency, freedom of expression, and representation, at least to a certain extent. Turbulence may dash the hopes of some progressive challengers to the system, who find the disordered and unruly nature of politics too complex and confusing to join in. But it does seem to bring swathes of people into politics who have not been involved before. It would probably be an impossible task to systematically compare levels of political engagement before and after the advent of social media, particularly given our inclusion of new, previously unavailable tiny acts at the lower end of the ladder of political participation, and it is not one that we have undertaken here. But we do believe that the waves of online mobilization that we have described and analysed offer political possibilities to people who would, prior to 2005, have eschewed politics altogether.
RESEARCHING THE FUTURE
To understand the turbulent world of chaotic pluralism requires new approaches, which could have profound implications for social science research. We have only embarked upon them here. Our aim for this book was not to be comprehensive, but to start to understand the causal mechanisms behind online mobilizations, generic to whatever applications are being used, which could be corroborated or refuted in the next wave of research. As promised in Chapter 1, we analysed large-scale, fine-grained data related to petition signing: a dataset of real-time transactional data of a kind that was rarely available to social scientists in the pre-Internet era. Researchers in the growing field of computational social science or social data science have seized the opportunities afforded by these kind of data. But mainstream social science has been rather slower to capitalize on the new possibilities, particularly outside the United States. We believe that this kind of social data science analysis should and will grow common as researchers realize its potential. To make this happen, social scientists will need to work more closely with researchers in the mathematical, physical, and life sciences in future years. This will present a challenge to conventional social science departments, where faculty may lack the technical and multidisciplinary expertise required to generate, analyse, and model big data. The authors of this book comprise two political scientists, one physicist and a computer scientist, but such a research team would be difficult to form in a traditional social science department.
The payoffs of collaboration between the social sciences and mathematical, physical, and life sciences may be great, in terms of understanding social systems as chaotic systems. Other research from the scientific world has shown that in various non-linear, highly interconnected complex systems (most of them with chaotic characteristics) there can be tipping points at which a sudden shift to a different dynamical regime may occur.51 Although traditionally the prediction of such critical points has been very difficult, research is starting to suggest that there may be generic indicators of the approach of a tipping point or critical transition, with obvious relevance for political turbulence. These early warning signals could come as particular temporal and spatial patterns such as slowing down, symmetry breaking, scaling of fluctuations, or the emergence of non-regular geometrical patterns such as fractals and scale-free distributions. Data science approaches to understand the world of collective action would need to develop far further than hitherto, but it might be that an understanding of patterns of communication or interaction on data generated from social media could provide such early warning signals in social systems in general, and in political turbulence in particular.
Of course, we should acknowledge some challenges to developing these approaches. We have used two key methodologies in this book: social data science (or computational social science) approaches and experiments or RCTs. Both these methods face technical, ethical, and logistical barriers that have restricted the range of sources that we could use within the scope of one book. Our large and comprehensive data on petition signing in two countries, including Twitter data relating to petitions, involved a three-year programme of hourly data collection, which was technically complex and time-consuming to both generate and analyse. The Google analytics data we presented in Chapter 3 resulted from a long-term relationship with the Government Digital Service; such datasharing arrangements take time to establish, and we were able to do so in only one government department. Historical Twitter data, as presented in Chapter 2, are most easily obtained through the use of commercial data providers, as we did here, which are usually costly and subject to technical and legal considerations. Although we were able to obtain a limited amount of Facebook data, it is not possible to access data on private posts that users have shared only with their social networks, and we had to extrapolate from data on public posts. Other social media platforms that delete posts after they are shared, such as Snapchat, pose far greater challenges to data capture, and we did not attempt to generate data from any of these platforms. Field experiments on online political behaviour, as we presented in Chapter 4, offer far greater external validity than laboratory experiments but usually rely on collaborations with social media platforms to implement the intervention and measure outcomes. We developed such a collaboration with the social enterprise mySociety, but we did not do so with any of the major social media platforms, as James Fowler did with Facebook in his 2012 voter turnout experiment.52 In fact, Kramer et al.’s experiment with Facebook on emotional contagion on social networks, which received a great deal of negative publicity for Facebook amid accusations that the experimental team had manipulated Facebook users’ emotions, has soured the company’s enthusiasm for such experiments, probably for many years to come.53 We could have appraised the importance of Google’s suite of applications and the way in which they use social information, in search algorithms, for example, were it not that these strategies are closely held secrets and those data are unavailable for research. There are numerous research projects underway that do expand the range of data science approaches and experimental methodologies beyond those used here, and the authors of this book are engaged in several of them.
We stress the urgency of pursing this research agenda at this particular time. At the World Economic Forum in 2015, the CEO of Google, Eric Schmidt, argued that ‘the Internet will soon be so pervasive in every facet of our lives that it will effectively “disappear” into the background’. Before this happens, and the Internet and social media dissolve into the ether, we need to understand the generic influences on digital behaviour, like social information and visibility, while we can still disentangle their effects. Furthermore, from a data perspective, we are at a turning point. As Schmidt observed, ‘there will be so many sensors, so many devices, that you won’t even sense it, it will be all around you’,54 but sensors and devices operate across multiple cross-cutting networks, presenting a data universe that is opaque, fragmented, and often proprietary, owned and jealously guarded by the corporations that market these devices. It may be difficult to derive data from some social media platforms—such as Facebook, as discussed above—but it is not beyond the ingenuity of Internet researchers to work out ways of doing so.55 It will be a very different task to obtain data from drones, from wearable technology, from heavily commercialized gaming environments using immersive technologies, from mobile medical devices, from driverless cars, or from all the devices that make up the Internet of Things. Before the Internet becomes too ubiquitous, researchers and policy makers need to seize the current moment to understand the profound political changes already underway.
We believe that the empirical and methodological paths we have started to take in this book reflect important developments for social science research. They must be accompanied by theoretical development, based on social scientific understanding and modes of enquiry and asking wide-ranging questions about social behaviour and societal development. Social scientists will need the collaboration of researchers from the mathematical, physical, and life sciences. There is no doubt of the value of observations and empirical laws of Johannes Kepler in the astronomy of the solar system and understanding of planetary motion, but it took nearly one century before Isaac Newton discovered his universal gravitation theory based on them. Only then did we become able to make accurate predictions about the behaviour of planets in a general framework and beyond the specific case of the solar system. Theoretical development of chaotic pluralism must be inspired by and validated against empirical observations and experiments. But to reveal the ‘universal laws’ governing political turbulence, we need to go beyond observations limited to a specific platform or phenomenon, and generate theories that address more fundamental aspects of collective human behaviour. We hope with this book to have taken a step in that direction.