Literature review

Review of the academic literature on democracy and deliberation

In a self-satisfied post-war USA dominating not just global politics but global political science, democracy was conventionally conceived according to the ’pluralist’ model (Dahl, 1967). Voting was central, understood as the aggregation of pre-existing interests. These interest groups could then be balanced politically in a ‘polyarchy’ such that none dominated—contrary to the Marxist claim that there was a US ‘power elite’ (Mills, 1956). Amid the American boom decades, there was an associated ‘economic’ conception of democracy, which somewhat reductively argued that voters chose parties based on their individual ‘utility income’ from government activity (Downs, 1957). Yet, as the (Canadian) political scientist Simone Chambers (2001) put it, democracy is, or should be, more about talking than voting—essential though that is—thereby placing the quality of democratic discussion in the spotlight.

 This was recognised, according to John Dryzek (2000: 1), in a ‘deliberative turn’ around 1990 in the theory of democracy, premised on the notion that ‘deliberators are amenable to changing their judgements, preferences, and views during the course of their interactions, which involve persuasion rather than coercion, manipulation, or deception’. The polarisation of American politics in the aftermath of the Vietnam debacle and through to the Clinton presidency suggested such a concern was at a premium. If preferences are to be transformed rather than merely aggregated, of course, this implies that arguments justifying particular claims have to be made in universal terms which others can reasonably accept (Besson and Martí, 2006: xvi) and communication must be as much about ‘listening and reflecting’ as talking (Ercan and Dryzek, 2015: 242). Deliberation is thus the antithesis of the remarkable affirmation by the former president George W Bush (Jost et al, 2003: 353), ‘I know what I believe, and I believe what I believe is right.’

 This notion of ‘deliberative democracy, or decision making by discussion among free and equal citizens’, as Jon Elster (1998: 1) put it, can be philosophically grounded in the work of Jürgen Habermas, standing in the very different post-war shadow of Nazi Germany—although Americans could also find some roots in John Dewey, for example (Elstub and McLaverty, 2014: 4). The first sentence of the 1949 Basic Law, the de facto constitution of the Bundesrepublik, says: ‘Human dignity shall be inviolate.’ For Habermas (2012: 81), ‘human dignity’ (rather than homo economicus) is foundational of a democratic order in which individuals, thereby constituted as citizens, ‘respect one another as members of a voluntary association of free and equal persons’.

 Based on his earlier work elucidating the notion of the ‘public sphere’, Habermas’ (1998: 299) discursive model of democracy thus also recognises that ‘processes of reaching understanding’ will ‘take place through democratic procedures or in the communicative network of public spheres’. For Habermas (1998: 382), informal communication in a ‘liberal public sphere’ is an antidote to the mere ‘plebiscitary legitimation’ of ‘populist leaders’. The deliberative perspective thereby not only enlarges the substance of democracy (talking as well as voting) but also extends its purview beyond parliamentary assemblies to include the public sphere, or what has come to be known as civil society (Edwards, 2004). This inclusiveness (Setälä, 2014: 149) meets the criticism that deliberation could only ever be an elite activity and indeed incorporates and transcends the alternative 1960s idea of ‘participatory democracy’ (Pateman, 1970)—which in its own terms faced the intractable difficulty of a zero-sum relationship to representative democracy and associated legitimacy challenges.

 Habermas’ connection can be seen even earlier in an otherwise cryptic equation, ‘state = political society + civil society’, by the political thinker and practitioner who had observed fascism destroy the Italian social fabric and been indefinitely imprisoned for his pains, Antonio Gramsci (2011: 75). In Gramsci’s vision, civil or ‘regulated’ society should increasingly take on elements of ‘political society’, since ‘all men [sic] are really equal and hence equally rational and moral’, in line with his fundamental political goal of breaking down the division in society between rulers and ruled.

Gramsci also recognised (2011: 163) that western Europe’s developed civil societies ensured that the Russian revolution would not be exported, as Lenin had envisaged: ‘The events of 1917 were the last instance of this kind.’ And the deliberative turn in democratic theory also coincided with the collapse of the husk of the Soviet system Lenin bequeathed and its eastern-European satellites, leaving no alternative political model in play—other than the pre-pluralist populism, resurgent in recent years, which the Hungarian leader Viktor Orbán has celebrated as ‘illiberal democracy’. A notable feature of such right-wing populist forces is that their communication is always strategic (that is to say, instrumental and manipulative), rather than deliberative (Wodak, 2015).

 In a blog on the ‘Brexit’ referendum in the UK in June 2016, Alan Renwick and Meg Russell of the Constitution Unit at University College London identified this struggle between populism and deliberative democracy at its heart. They contend that ‘the essence of democracy lies not just in voting, but in discussion: a decision is only as democratic as the deliberation that produces it. And quality deliberation has two essential features. First, it is thoughtful: people have accurate information about the issues; they reflect carefully upon the validity of the various arguments; they consider how to add up the different points into an overall decision. Second, it is inclusive: everyone is listened to with respect and everyone listens respectfully.’ They suggest that the referendum hardly measured up as a deliberative exercise in these regards. A challenge facing the incoming US president, Donald Trump, will be whether his populist instincts will jar with more thoughtful and inclusive figures on the Republican centre-right, on Capitol Hill and in the media, whose support he may require.

 It is against this real-world backdrop that a 2014 volume on deliberative democracy could be opened with the words (Elstub and McLaverty, 2014: 1): ‘Liberal democracy around the world is in crisis … and deliberative democracy is increasingly seen as the solution.’ And the introduction by two experts in the field to a 2015 journal special issue began thus (Ercan and Dryzek, 2015: 241): ‘A deliberative approach now dominates the theory—though not yet the practice—of democracy.’

 There has, it is true, been a competitor within the political-science discipline to the hegemony within it (as Gramsci would have said) of deliberative approaches. If the ‘pluralist’ (Dahl) and ‘rational-choice’ (Downs) approaches have had their day in crisis times, Chantal Mouffe’s ‘agonistic’ alternative (which she rehearsed at a lecture in Belfast, at the invitation of the think tank Democratic Dialogue) has had to be reckoned with (Besson and Martí, 2006: xvi-xvii). Mouffe (2005) challenged the consensual aspiration of deliberation by claiming that politics was about the clash of incommensurate world views. She warned, with foresight, that if democratic publics were not presented with alternative Weltanschauungen populists could successfully rail against such elite convergence. But her own position was vulnerable in its association with the thinking of the pro-Nazi intellectual Carl Schmitt, for whom the friend-foe distinction was irremediable in politics.

 In testament to critics such as Mouffe, however, the concept of deliberative democracy has itself evolved, in a deliberative fashion, taking reasoned criticisms on board (Dryzek, 2016: 210). To those who have claimed that the theory cannot accommodate conflict in society, it has been argued, in a study of opposition to fracking in New York state, that conflict is a necessary stimulus to the ‘reflexivity’ on which deliberation depends (Dodge, 2015). Every individual deliberation, after all, starts off with a disagreement about a claim—to empirical truth or moral rightness (Knops, 2016: 309)—so ‘contestation and contentious politics are essential to deliberation’ (Ercan and Dryzek, 2015a: 360). And such recent conflictual developments as the social protests in Brazil in advance of the 2014 football World Cup and the Gezi Park protests in Istanbul in 2013 demonstrated precisely deliberation active in the public sphere, including via the vehicle of ‘social media’ (Mendonça and Ercan, 2015)—such events cannot be understood as if they were old-fashioned demonstrations where hierarchical organisations directed their supporters on to the streets to chant slogans. This is in no way to contradict the argument of Georgina Blakeley (2014: 30) that the ‘15M’ movement which emerged in 2011 ‘engaged in and fostered deliberation in the public sphere in Spain but it also engaged in mobilisation and collective action’. No one is attempting to make deliberation a totalising theory which would reduce social complexity to such a simplifying essence.

 A related challenge is that in a capitalist society members of dominant social classes will have their voices heard more readily than those from subaltern social groups—not just because of their privileged access to the public square megaphone, so to speak, but also because of the greater esteem likely to be accorded to what they have to say, regardless of its merits (Chambers, 2009: 339). This is compounded by how such voices will tend to be amplified by the media, particularly where corporately owned and largely unregulated. It is therefore reasonable to argue that, whatever its normative appeal, ‘substantial moves towards greater deliberative democracy will need to go hand in hand with moves to greater social equality’ (McLaverty, 2014: 49). Having said that, the normative claim of free and equal citizenship does have a pertinent effect, as Ian O’Flynn and Nicole Curato (2015: 300) argue, as if a deliberative framework is accepted ‘it will not do for the strongest members of society simply to discount the views of the weaker members or to ignore them altogether’.

Shawn Rosenberg (2014) raises a more fundamental concern, with his claim that social-psychological studies have demonstrated the limited capacity of human beings for deliberation. Individuals may gravitate towards pinning blame for a social problem on to visible agents rather than perceiving underlying structures, even if the latter are really determinant, and be guided by personal experience, even where this needs to be contextualised and qualified by (say) statistical evidence. ‘Cognitive shortcuts’, such as drawing on ethnic stereotypes when faced with a stranger, and ‘motivated cognition’, such as highlighting any data which might support denial of man-made climate change in the face of scientific consensus, can also inhibit considered deliberation. 

Yet while these are serious worries, in a way this is to put the cart before the horse. Post-war social psychology emerged precisely to answer the question as to why so many ‘ordinary’ Germans had come to embrace Nazism (Chryssochoou, 2004). Its lessons point to the need to embed individual human dignity in the body politic, as Habermas recognised. Moreover, Rosenberg concedes that some individuals will develop to a higher psychological capacity for deliberation than others. And so Gramsci (2011: 338) insists on the need for an ‘ethical and cultural’ state which acts as ‘educator’, bringing ‘the great mass of the population to a certain cultural and moral level’—which can come via effective citizenship education in the school, for instance. Rosenberg does point importantly to the psychological underpinnings of populism—the night-time tweets of the newly-elected US president, Trump, could be cited as a textbook case—but the manner of its victories and defeats over the last century has highlighted the contingency of such outcomes, dependent as they are on wider social and political currents, irreducible to individual psychology.

A related concern, exacerbated in the age of the internet, is the risk of ‘enclave deliberation’ (Setälä, 2014), where individuals gravitate only towards those of like mind, magnifying their own ‘cognitive errors’ through a process of ‘group polarisation’. And there is very real unease that the enclosure of the internet ‘commons’ by a very small number of very large Californian ‘social media’ corporations is driving such ‘balkanisation’ through the use of algorithms to filter what users see, so as to maximise corporate revenue from online advertising. Yet one experiment using a mini-public (see below) did at least find that, even where like-minded immigration sceptics were put together in a group, when exposed to a process of deliberation with impartial materials they did change attitudes in a more tolerant direction. And the internet is a two-edged sword. Highlighting the autonomous, ‘leaderless’, character, of latter-day social movements such as Occupy Wall Street, Manuel Castells (2012: 224, 228) affirms that they channel outrage into hope ‘by deliberation in the sphere of autonomy’ and he describes the internet as ‘a privileged platform for the social construction of autonomy’. Ariadne Vromen et al (2016) found that Facebook was particularly valued in this regard by young people who were politically engaged but came from less socially privileged backgrounds.

 Painting this wider intellectual and political canvas helps us contextualise this study of Northern Ireland. The formation and early years of the polity were by no means out of synch (Prince, 2007: 52) with the way the wider Europe’s ‘maddening mosaic’ of ethnic (including religious and linguistic) diversity was subjected from 1914 to three decades of ‘total war, brutal occupations, the fall of empires, the emergence of new states, the constant redrawing of borders, the forced relocation of entire peoples, and murder on an industrial scale’. Northern Ireland did, however, largely escape the final phase, in which Hitler and Stalin found ‘inhuman solutions to human problems’ associated with stigmatised members of minority communities to which conspiracy theories were attached. Instead, these persisted into the region’s post-war years, until their articulation by the civil-rights movement—part itself of wider European and US radical trends—was to precipitate, albeit unwittingly, decades more violence.

However ‘normal’ the Unionist state might have seemed in a Europe of the 1920s, by 1968 it clearly did not conform to the pluralist expectation of post-war democracy. Unsurprisingly, following the introduction of Westminster ‘direct rule’ in 1972, the principle of ‘power-sharing’ was immediately advanced as an alternative and became a for-the-most-part consistent basis of government policy, in London and Dublin, thereafter (Wilson, 2010). Critically, though, pluralism did not specify relationships, other than of negotiation and bargaining, between aggregated interests.

In the 1970s, pluralist theory was to be embellished by the Dutch political scientist Arend Lijphart (1977). In a nod to Dahl’s Pluralist Democracy in the United States, Lijphart’s Democracy in Plural Societies advanced his notion of ‘consociational democracy’ for societies marked by cleavages of ethnic segmentation rather than social interest. This not only advocated relationships of mutual veto between communalist parties in societies such as Belgium, but also recommended ‘group autonomy’ between members of those (in that case Walloon and Fleming) communities in civil society, supposing that high fences made good neighbours. The obvious risk with applying such an approach to an ethnically divided society, as shown by the experience of Lebanon after the Ta’if accord of 1989, Bosnia-Hercegovina after the Dayton accords of 1995 and Macedonia after the Ohrid agreement of 2002 (Wilson, 2010), is that the very ethnic identities framing division become perversely embedded at the expense of any notion of individual citizenship. Since the individual is the unit of every democratic constitution (Bobbio, 1996: 90), the polity then becomes chronically polarised and dysfunctional.


The Good Friday agreement arrived too early for these lessons to be evident and for the emerging deliberative paradigm to have an influence—although after the paramilitary ceasefires Democratic Dialogue did bring Anthony Giddens to Belfast to put his similar notion of ‘dialogic democracy’ (Giddens, 1994) into the public domain. The agreement thus did not meet the requirement of reciprocity or the recognition of civil society as a space for deliberation to which O’Flynn (2006) attends—hence the many suspensions of the power-sharing executive, particularly from 2002 to 2007, and the eventual demise of the Civic Forum signalled in the agreement. A case study of environmental regulation showed in detail how, when devolution was renewed, a well-deliberated argument for an independent environmental protection agency was successfully resisted by the Democratic Unionist Party (DUP), in deference to the agricultural industry, using the mutual-veto provision of the agreement (Turner and Brennan, 2012). More widely, as O’Flynn (2006) points out, while the end of the cold war precipitated a new focus on the challenges of ethnically divided societies (particularly in east and central Europe) alongside the deliberative turn in democratic theory, academic discussions of deliberative democracy and divided societies have since proceeded along parallel yet unconnected tracks (though see Dryzek, 2005).


Experiments in deliberative democracy, including in Northern Ireland

The process of theoretical elaboration of deliberative democracy in recent times has gone hand in hand with its experimental exploration. A range of methods have been deployed: citizens’ juries, consensus conferences, planning cells, deliberative polls and citizens’ assemblies (Elstub, 2014). These have been grouped under the term ‘mini-publics’. The idea has been to bring together random or stratified samples of citizens to deliberate together on a topic of public concern.

Contrary to the excessive pessimism of Rosenberg (2014), these experiments have demonstrated that the experience of public deliberation can meaningfully change individual attitudes through rational argumentation (Grönlund et al, 2015: 998): ‘Groups de-polarize rather than polarize, people learn during deliberation and their misperceptions are corrected.’ This has been particularly apparent with deliberative polling, where tangible differences have been shown when participants have been surveyed before and after the deliberative experience (Elstub, 2014). Specifically, on the topic of immigration in Finland, rendered neuralgic by the (True) Finns party, Grönlund et al (2015)found not only that those who responded xenophobically in the initial survey moved towards a more xenophilic stance as a result of deliberation in a mixed group, but even in a group confined to those who were initially hostile to immigration there was a softening of attitudes as a result of the experience—suggesting worries about ‘enclave deliberation’ could be overdrawn.

Democratic Dialogue was responsible for the two deliberative-democratic experiments which have taken place in Northern Ireland. It organised a citizens’ jury, commissioned by the then Eastern Health and Social Services Board, on public involvement in health, which took place in the aftermath of the Good Friday agreement. This focused on consultation procedurally rather than any particular health issue substantively, and the commissioning body compared it favourably with its routinised experiences with public meetings (Barnes, 1999). DD also organised a deliberative poll in Omagh, in conjunction with the US expert James Fishkin (and involving O’Flynn), on the potential future of the Lisahally former army base as a shared-education project, which eventuated shortly before the renewal of devolution in 2007. The poll showed a clear before-after shift to support for integrated education and the project was very well received by the then vice-chancellor of Queen’s University and author of a government-commissioned report on the reconfiguration of the educational estate, George Bain, who described it as ‘highly professional—on the edge of social scientific technique’ (Wilson, 2007). The idea of a shared-education scheme at Lisahally did subsequently accrue cross-sectarian and cross-party support and will involve six schools in the area, coming to fruition in 2020.

In the Republic of Ireland a deliberative mini-public of sorts was established by government in 2012, called the Constitutional Convention, sitting periodically over 18 months. It comprised 100 members, two thirds of whom were a (broadly) representative sample of citizens, the remainder political representatives. It was charged with addressing eight concerns, including how to enhance women’s participation in politics and public life and what to do about the provision in the constitution valorising the assigned role of women in the home. It most significant impact was in supporting marriage equality, which led to a referendum on a change to the constitution to that effect, passed by a 62-38 majority in May 2015 and followed by consequent legislation later that year. While this report was in preparation, a successor, 100-member Citizens’ Assembly was deliberating—this time without any politician members—on a further raft of measures but particularly tasked with addressing the still controversial issue of abortion. The Citizens’ Assembly explains its purpose as ‘an exercise in deliberative democracy, placing the citizen at the heart of important legal and policy issues facing Irish society today’. 

Chambers (2009), noting how such experiments with mini-publics have proliferated, argues that these remain no substitute for work at the scale of the ‘mass public’ to promote deliberative democracy. Indeed, André Bächtiger and Alda Wegmann (2014) contend that mini-public events have had little impact on macro-political decision-making and, indeed, that if they had that would raise serious questions of legitimacy, given the very small number of individuals involved in the episode—an argument echoed by Cristina Lafont (2015: 42). And here there has been an interesting development in the theory, to meet the challenge of scaling up to the macro-level, with the notion of deliberative ‘systems’ (Elstub et al, 2016), although perhaps more strictly we should speak of ‘networks’ (Knops, 2016), operating across the whole of a society or indeed transnationally.

This could be significant because, it has been contended (Elstub and McLaverty, 2014: 7): ‘Despite the incredible advancement of deliberative democracy, it is now at a crossroads.’ Indeed, there has even been a suggestion, based on publications, that we have reached ‘peak deliberation’. A search of Google Scholar for ‘deliberative democracy’ by year by John Parkinson, co-convenor of the participatory and deliberative democracy specialist group of the UK Political Studies Association, generated a graph (reproduced below) which shows exponential growth in the 1990s, arithmetic progression in the 2000s and a plateau of just under 5,000 publications a year since 2012—although, as Parkinson stresses, that is still an awful lot of publications.

The notion of deliberative systems/networks allows of a more fine-grained representation of deliberation in society, beyond Gramsci’s ‘state = political society + civil society’ equation. That is to say, a number of sites or spaces of deliberation can be specified, which can be analysed in themselves and in their interconnections with others. The reason for thinking of this, preferably, as a network is not only that communication, particularly in the internet age, is inherently networked, as Habermas implied. More, in discussion of governance, the notion of ‘networked’ governance has been elaborated as a ‘citizen-centred’ successor to the ‘new public management’, which treated the citizen as a mere ‘consumer’ with a right of exit but none of voice (Wilson, 2002). Most problematically, to speak of a system implies a clear boundary between the system and a non-deliberative beyond, which does not capture social reality well, or a blurring of that boundary, which then dilutes the conceptual frame. Indeed, the global political maelstrom of recent years, and in particular the rise of populist forces on the one hand and spontaneous social movements on the other, is much better captured by thinking of deliberative networks, with their inevitable fractures and holes, in a dislocating state of volatile flux, rather than anything as (relatively) stable sounding as a system. Such networks will operate at different levels, from the interpersonal to the local to the regional to the national to the transnational—here the fact that we don’t have to think of networks as bounded is an asset.

Making sense of this may seem like capturing mercury with a fork but recent work in neuroscience (Almqvist and Haag, in press) offers a simple heuristic. The brain is a mass of tens of billions of neurons and we cannot capture how it functions unless we move beyond a mere aggregation of individual neuronal signals to conceive of it as a series of networks. What recent experimental work has shown is that it is the integration within and across these networks which generates the phenomenon of a conscious self (capable, among other things, of deliberation)—for example, the poor mental functioning of some former heavyweight boxers has been linked to the severing of the small strip of neural tissue which joins the two hemispheres of the brain.

In a deliberative context, one could here point in salutary fashion to the perverse demographics of the Brexit referendum: despite the evidence showing xenophobia to be significantly correlated with a ‘leave’ vote, areas of high cumulative in-migration, notably London, were significantly more favourable to ‘remain’. Thus, to the intense frustration of many Londoners and more cosmopolitan youngsters who found the referendum highly unsatisfactory as a democratic experience (and not only because they lost), there was a lack of integration of many ‘leave’ voters into their deliberative networks—a social fracture filled by anti-immigrant press stories and populist political rhetoric. In the US presidential election, similarly, some ‘rustbelt’ (ex-)workers in swing states, no longer integrated into (Democrat) politics via the trade unions, were amenable to being interpellated in populist style by Trump, as ‘Americans’ supposedly shortly to be ‘great again’. As we shall see below, indeed, the issue of ‘transmission of views’ has been identified as an important dimension of whether a deliberative system (or network) works well.

In this context, several sites have been identified as arenas for deliberation. The most obvious is the parliamentary assembly. Here, research on the National Assembly for Wales (Chaney, 2006) has shown how inclusive deliberation has been favoured: by the assembly’s statutory requirement to promote equality of opportunity, by the additional-member electoral system allowing women to be promoted via the top-up list and by the emergence of equality ‘champions’ among assembly members bringing their experience as feminist practitioners to bear. Such sites obviously include the media—although all too often these are spaces for advertising-driven clicks using ‘plebiscitary’ rather than deliberative forms of communication (Chambers, 2009: 341). Privileging the latter would require not just media pluralism but more broadly a media ecology which engendered a public forum regulated by a supportive legal framework and moderated by journalistic ethics (Girard, 2015). Perhaps less obviously, such spaces also include the workplace, recognising that ‘private’ companies have very public ‘externalities’ and that these may well be more central places for day-to-day deliberation than those requiring citizens voluntarily to make a demanding civic commitment (Felicetti, 2016). And the family, despite feminist contention to the contrary, is often walled off as an even more ‘private’ sphere, although it is again a site of ‘collective decision-making’ where whether decisions are made deliberatively or in a manner of domination is a relevant question (Chalmers, 2015: 346).

This then brings us to the empirical meat of our study in the next section, in which we treat Northern Ireland as the location of a deliberative network and assess how well it functions in that regard, in whole and in part. The Involve approach identifies seven elements of such a network, devised by Dryzek, including both hubs and interconnections. This paper is an early attempt to operationalise them, which doubtless will lead to refinements in the approach.

Next Section: Public Space