Which conception of political equality do deliberative mini-publics promote?

Abstract

In democratic political systems, political equality is often defined as an equality of opportunity for influence. But inequalities in resources and status affect the capacity of disadvantaged citizens to achieve an effective political equality. One common thread running through recent democratic innovations is the belief that appropriate institutional devices and procedures can alleviate the impact of background inequalities on the presence and voice of the disadvantaged within those designs. My objective is to achieve a clearer understanding of the conception of political equality that informs a specific subset of these designs: deliberative mini-publics. I focus firstly on the methods of participant selection advocated to secure equal presence. According to what principle is participation distributed? If it is according to the ‘equal probability’ principle, rather than ‘equal opportunity’, what difference does this make in terms of political equality? Secondly, achieving equality of voice is usually conceived in terms of equalising opportunities for influence among participants. How is this objective understood and what does this say about the underlying conception of political equality?

Keywords

Political equality mini-publics deliberation deliberative polls equality of opportunity random selection

Political equality is at the heart of democratic citizenship. Contemporary democracies affirm the equal rights of citizens to vote, join political parties and stand for elections. The conception of political equality that informs these rights is standardly described as an equality of opportunity for influence (Christiano, 2004: 275; Cohen and Fung, 2004: 171; Knight and Johnson, 1997: 280; Swift, 2006: 189). Though citizens may choose not to participate in the political life of their society, political rights formally secure their equal opportunity to participate in processes of opinion and will formation. This conception of political equality includes agency as one of its key elements: equal rights enable citizens to perform political activities individually as well as collectively. Indeed, the capacity to act with others is a central component of democratic politics.

Yet contemporary democracies are also societies in which vast inequalities in resources and status exist between citizens, and these inequalities significantly affect opportunities for political influence. Empirical enquiries consistently show that background inequalities in socio-economic resources and education, as well as inequality associated with gender and ethnicity, translate into unequal ‘presence’ and ‘voice’ within the institutions and practices of contemporary democracies.¹ Citizens who are economically and socially disadvantaged find themselves in a situation of ‘political poverty’ − defined by James Bohman (1997: 333) as the ‘inability of groups of citizens to participate effectively in the democratic process’, which is related to a deficit in the ‘capability for effective social agency’ (1997: 343). Political poverty means public exclusion, since politically poor citizens are incapable of successfully initiating the joint activity of public deliberation. It also means a form of coercive political inclusion since the politically marginalised are the legal addressees of decisions over which they have next to no control or influence.

One common thread running through the different participatory and deliberative innovations that have developed over the last 30 years is the belief that appropriate institutional devices and procedures can alleviate the impact of background inequalities on the presence and voice of the disadvantaged within those designs (Chambers, 2009; Cohen and Fung, 2004; O’Flynn and Sood, 2014; Smith and Wales, 2000; Thompson, 2008a). Though inequalities in resources and status persist, democratic designs can be structured to function as ‘more perfect public spheres’ (Fung, 2003: 338). Neither erasing background inequalities nor vying to replace existing political institutions, one of their goals is to provide arenas in which disadvantaged citizens may exercise their citizenship in ways that are too often denied to them in the real world of democratic societies.

My objective in this article is to achieve a clearer understanding of the conception of political equality that informs at least some of these democratic designs in relation to equality of opportunity, but also in relation to agency, both individual and collective.

To do so, I will focus, in the first section, on the methods of participant selection advocated to secure equal presence. According to what principle is participation distributed? If it is according to the equal chance or equal probability principle, rather than equal opportunity, what difference does this make in terms of the underlying conception of political equality? Is ‘equal presence’ conceived strictly in individualist terms or is it related to groups? And, if so, how?

In the second section, I consider the issue of voice. Achieving equality in this context is conceived in terms of equalising opportunities for influence among participants (Fishkin, 2009: 100–101; Fung, 2003: 348; Smith, 2009: 21–22). I intend to clarify the conditions the designs establish to achieve this objective despite the existence of background inequalities. How is the political agency of participants understood and facilitated in this respect? And what does this say about the underlying conception of political equality?

Given the proliferation of innovative democratic designs, I cannot aim to cover them all in one article. I will focus my enquiry on one family of designs, often referred to as deliberative mini-publics (Ryan and Smith, 2014: 20–21).² These designs bring together ordinary citizens to deliberate over public policy issues, such as health care, local budgets, environmental issues, etc. They are usually convened by a public authority or by independent, non-profit organisations. In most cases, mini-publics have no authority to make legally binding decisions; their aims are either to make recommendations to government decision makers or to empower participants and stimulate public discourse over important public policy issues. The quality of deliberations between participants is a key concern.

Designs belonging to this family use sortition in selecting participants. As we will see, it is important to distinguish between designs that are based on random sampling (deliberative polls, planning cells) from designs that use some form of random selection without aiming to satisfy the requirements of random sampling. This second category covers designs such as citizens juries and consensus conferences. In respect to deliberation, deliberative mini-publics share a reliance on procedural devices like the use of facilitators and small group discussions while enforcing a strict separation between participating citizens and stakeholders.

Concerning presence, I will argue that, apart from designs like planning cells and, to a certain extent, deliberative polls, that are premised on the principle of equal chance or probability, most designs are better understood as attempting to instantiate the idea that all citizens share a basic capacity to participate. This view relates political equality to groups, since to represent publicly the principle of ‘basic equal capacity’, the designs are described as having to be broadly inclusive, securing, in particular, the presence of marginalised groups. But what the requirement of inclusiveness entails remains sketchy.

Concerning voice, my contention is that most deliberative mini-publics rely on a cursory assessment of the impact of background inequalities on political agency. As such, they address political poverty in ways that will scarcely be effective for those who experience severe and compounded forms of marginalisation. In particular, some design features, aimed at securing the overall quality of deliberative exchanges, may have mixed effects on the capacity for voice of the disadvantaged, though they serve important democratic goods. In many cases, deliberative mini-publics will be unsuccessful in alleviating the effects of political poverty within the forum itself, though this is an essential condition of fair deliberative exchanges (Fung, 2003: 344). This suggests the need to confront more explicitly the difficult choices that designing mini-publics involve in the context of democratic societies beset by persisting inequalities.

Securing equal presence through sortition

As opposed to elections, sortition does not require individuals to display any particular set of resources, whether moral, intellectual or material. An individual’s selection does not depend on others’ judgement of her abilities, however defined. There is no need to finance an electoral campaign or to build a supporting social and political network. Considered from this perspective, sortition looks like an egalitarian method of participant selection that can circumvent the impact of background inequalities on the participation of the politically disadvantaged.

But sortition cannot be described as a method that achieves equal opportunity, formally or substantively.³ As Rawls (1971: 72) puts it, formal equality of opportunity requires that ‘all have at least the same legal rights of access’ to the desired positions. If equality of opportunity is substantive, then all those who have the willingness to secure these positions should have ‘the same prospects of success, regardless of their initial place in the social system’, given ‘the same level of talent and ability’ (Rawls, 1971: 73). In short, equality of opportunity supposes that everyone is provided with the opportunity to secure the desired position. Sortition does not aim to realise that form of equality: it is not the case that all those who have the willingness to achieve a certain position or to participate in a given process will have the opportunity to do so, formally or substantively. Access depends on the luck of the draw. Unlucky individuals have no possibility to participate even if they would have wanted to, given the opportunity (Brown, 2006: 212–213). Participation, to use Lynn Sanders’ (2010) words, is ‘by invitation only’.

What sortition – under specific conditions – can do is treat all citizens equally by securing their ‘equal chance’ or ‘equal probability’ of being selected. If opportunities to participate are limited, then sortition appears as a mechanism that is fair in respect to the distribution of the scarce good of participation since, as Graham Smith puts it, ‘no citizen or social group from the given population is systematically excluded’ (2009: 80).

Notice that, in the case of deliberative mini-publics, this holds only for the first stage of the selection process at which random selection is used to draw the initial sample of potential participants. Since participation in any mini-public is voluntary, background inequalities are likely to have an impact at the second stage, when an invitation is extended to those randomly selected, affecting who accepts and who declines the invitation to attend the forum.⁴

To choose sortition as the method to select participants is also to manifest publicly the belief that all citizens are politically equal in the sense that they share a basic capacity to participate in discussions over complex policy issues (Goodwin, 2005: 92–93). This belief was also fundamental to the experience of democracy in Athens: all citizens were recognised as having the capacity to hold office and to speak their mind in the Assembly if they so desired.⁵ The conjunction of sortition with rotation in office secured to each citizen ‘a reasonable expectation of serving his country’ (Montesquieu, 1989: 13). In the contemporary context, however, public recognition of citizens’ equal capacity to participate remains symbolic since few citizens receive the actual opportunity to exercise that capacity: Mini-publics represent the basic capacity that all members of the demos are deemed to share.

Political equality and representativeness

In their defence of deliberative polls, Fishkin and Luskin (2005: 40) contend that the equal probability to participate is only one of two components of the political equality secured by random sampling, the other being statistical representativeness: ‘political equality stems from random sampling. In theory, every citizen has an equal chance of being chosen to participate, and on average, over infinitely repeated sampling from the same population, the sample would resemble the population exactly’. In presenting the rationale behind deliberative polls, Fishkin (2009: 50) notes that citizens who participate regularly in the institutions of representative democracy are generally:

more white, more prosperous, and more educated by far than those who do not. In that sense, effective political equality has been achieved far less than the breakdown of formal barriers to participation (in terms of voting rights) would suggest. A strong claim of political equality would attempt to minimize participatory distortion, making those who choose to participate as much like the entire electorate as possible.

Imbalances in gender, ethnic or class representation in parliaments can be understood as pointing to failings in how representative institutions redeem the promise of political equality contained in the affirmation of equal political rights. But what does this say about the relation between political equality and representativeness?

The (lack of) representativeness of an elected assembly is often used as an indicator of the extent to which equality of opportunity between citizens is effectively realised in the electoral process. If women or ethnic minorities are systematically underrepresented, few argue that this imbalance is the result of either a lack of interest or a lack of capacity in participating. The inference is that there must be barriers to participation that explain this systematic underrepresentation.⁶ Conversely, one could infer from the absence of any significant imbalance in representation that citizens enjoy an effective equality of opportunity. In short, when political equality is understood in terms of equality of opportunity, one assumes some causal relationship between the realisation of equality for individual citizens and the elected assembly’s representativeness.

What about sortition? If its use is justified in reference to the equal chance of being selected principle, what is the relation between this principle and representativeness? There is no necessary relation since the principle can be fully realised independently of any claim to produce a representative body. For each member of a population to have an equal chance of being chosen, it is both necessary and sufficient that all members be included in the draw. But this in itself has no implication on the representativeness of those selected. It is only if one chooses a specific type of random selection, based on modern techniques of random sampling, that one may claim that ‘on average, over infinitely repeated sampling from the same population, the sample would resemble the population exactly’. Thus, the decision to choose a specific method of random selection to achieve statistical representativeness cannot be justified simply by referring to the equal chance of being selected principle. It must rest on distinct reasons. What are these reasons and how are they related, if at all, to political equality?

A first reason that features prominently in the deliberative polls’ methodology is the ability to extrapolate from the results of the discussions held in the deliberative forum back to the population in general. Fishkin’s (2009: 82–83) claim is that if the initial sample is statistically representative of the population, and if the sample of participants (e.g. those who end up participating in the deliberative forum itself) is also statistically representative, then it should be possible to claim that the results of the discussions held in the deliberative forum would be what the demos itself would think if it had the opportunity to deliberate in the kind of facilitated environment that deliberative polls provide.⁷ This is an interesting, if controversial, claim to make. Its importance, for Fishkin, is easy to understand: representativeness increases the legitimacy and the normative authority of the forum’s deliberations’ results. But this still does not explain how representativeness is related to equality.

The root notion of political equality, according to Fishkin et al. (2010: 4), is the ‘equal consideration of the views of everyone’ in the relevant demos. Equal consideration can be achieved in either one of two ways: either through ‘an equal counting of everyone, when everyone or virtually everyone participates’ or through ‘an equal chance of being selected, via random sampling, combined with an equal counting of the views of those selected’. The two alternatives are described as equivalent because the deliberative poll’s methodology ensures that the sample of participants in the deliberative forum is statistically representative and that the views of those who participate in the forum are ‘counted equally’. This assumes that if random sampling is properly applied, and if there is no distortion between the initial sample and the deliberative forum, then this process is equivalent to counting the views of everyone equally, where ‘everyone’ includes all members of the relevant population because of the sample’s representativeness. This is a strong claim to make in support of the capacity of random sampling to mirror the demographic and attitudinal makeup of a large population.

More importantly, it also raises difficult issues on the normative front. Firstly, it shows little respect for the ‘plurality and distinctness of individuals’, to use Rawls’ words in his critique of utilitarianism (1971: 26–29). Though a deliberative process that proceeds on the basis of a representative sample of citizens registers an important diversity of views and is, in this sense, much more respectful of plurality than the utilitarian’s ideal spectator, to claim that such a process is equivalent to counting everyone’s views equally is to deny the basic separateness of persons.

Secondly, notice that the representativeness of the assembly is produced by the sampling procedures. It is not related to the agency of individual citizens. In fact, it seems that agency is not considered by Fishkin to be an essential element of political equality since equality can be achieved though a significant number of citizens – indeed a majority of them – may not exercise their political agency in any way. In this sense, political equality appears as a feature of institutions, not an attribute of individual citizens understood as agents.

From random sampling to quasi random sampling

If we examine more closely the different designs that use sortition, we soon realise that only some can be described as aiming to instantiate the equal chance principle. Designs like planning cells and deliberative polls, in which either simple random sampling or proportionate stratified random sampling (SRS)⁸ are used, may legitimately claim ‘equal chance’ as their goal, though this ideal remains difficult to reach in practice (Griffin et al., 2015; Ryan and Smith, 2014: 17). However, in all cases where the individuals included in the draw form only a subset of the relevant demos, one cannot justifiably claim to be relying on the equal chance principle. In consensus conferences, for instance, participants are often randomly selected from the list of those individuals from the relevant population who have answered an advertisement (Carson and Martin, 1999: 59–60; Hendriks, 2011: 47). Neither is the equal chance principle operative in designs where disproportionate SRS is used. Though in the latter case, all members of the demos may be included in the draw, the use of a weighted lottery means that those individuals belonging to a smaller stratum will have more chances of being selected than those who are part of a larger stratum (Hendriks, 2005: 82 and 99, note 12).

The reasons motivating the use of these techniques vary. Certainly, the substantial costs attached to organising planning cells or a deliberative poll, as well as the complexity of the tasks involved are significant factors. These considerations aside, citizens juries and consensus conferences are small deliberative forums involving in between 10 and 25 people. With such small numbers, it is not possible to realise both the equal chance principle and representativeness. When confronted with the choice between securing some form of diversity and ensuring that each individual of the relevant population have an equal chance of being selected, organisers often prefer to sacrifice equal chance to secure some form of representativeness, though statistical representativeness remains necessarily out of reach.⁹ Since most proponents of mini-publics do not aim for the kind of robust counterfactual claim that Fishkin is looking for, achieving statistical representativeness is not an objective of paramount importance, but diversity or inclusiveness often is.¹⁰

The conception of political equality that underlies these designs cannot be interpreted in terms of either equal opportunity or equal chance. But the use of sortition manifests the fundamental belief in the basic equal capacity of all citizens to participate in discussions about complex policy issues, no matter whether they are old or young, male or female, white or black and rich or poor. This public recognition is understood as requiring the presence of a broad cross-section of the population, including citizens who belong to groups that are too often underrepresented or disadvantaged in their societies’ institutions.¹¹ To achieve inclusiveness in a small forum, it may well be necessary to over-represent certain minority groups either through oversampling or the use of quotas. Here, it is the groups’ exclusion from mainstream institutions that justifies the use of specific measures to secure their presence. As Ryan and Smith (2014: 20) put it:

the primary aim is clearly to engage an inclusive group of participants from the affected population, from which no social group or perspective, particularly those who are traditionally politically marginalised, is excluded. (…) Inclusiveness may be better served by overrepresenting those social groups and perspectives not normally present or heard.

The importance given to inclusiveness, so understood, indicates that political equality is also related to groups, not strictly to individuals.

Fishkin and his colleagues reject any move away from statistical representativeness on the ground that it constitutes a breach of political equality: over-representing any group means that the views of the individuals belonging to that group will count for more than the views of others. Though representativeness is an essential element of deliberative polls, their underlying conception of political equality remains strictly individualistic. The beauty of random sampling is that it enables its supporters to remain faithful to the individualist basis of their conception of political equality while ensuring the descriptive representation of the population.¹² But Fishkin’s model is not strictly impartial in its consideration of the sociological or group composition of the society as it does not usually rely on simple random sampling, but rather uses stratification.

In all cases of mini-publics where some form of stratification is used (proportionate or disproportionate), it is legitimate to ask what justifies the decision to choose some specific group or category for stratification and not others. In other words, in stratifying we abandon one of the chief advantages of random selection, that is the impartial allocation of the valuable good of participation, which relieves organisers from the need to justify selection criteria. As Peter Stone (2009: 376) argues, ‘a lottery contributes to decision making by preventing the decision from being made on the basis of reasons’. Of all the designs considered in this article, only Dienel’s planning cells, based on simple random sampling, can claim to be impartial in this way. Wherever stratification or quotas are used, the need for good reasons arises again, especially since minority groups who are not selected for stratification may well turn out to be underrepresented in the forum relative to their demographic importance in the population.

To illustrate, the decision made by the organisers of the British Columbia Citizens’ Assembly (BCCA) to secure the proportional participation of men and women, old and young, and, to a lesser extent, Aboriginal people, but not the less educated or ethnic minorities, raised important issues related to the justification of these choices.¹³ The decision not to select education and visible minorities for stratification translated into the underrepresentation of those with less education and those belonging to visible minorities in the deliberative forum.¹⁴ This consequence becomes problematic if one accepts the claim that, for instance, ‘visible’ ethnic minorities have a specific perspective to defend over the reform of the electoral system such that their underrepresentation may have an impact on the result of the deliberations (James, 2008). Pointing out the practical difficulties associated with the multiplication of strata does not constitute a satisfactory answer (pace Thompson, 2008b: 43). If convenors insist on the proportionate or significant presence of some, but not all groups composing the population in the deliberative forum, then it is essential to show that the categories selected for stratification reflect a considered judgement over which groups should be present. This judgement may be based on the evaluation of the situation of certain groups within the society and/or on the nature of the issue under discussion (James, 2008: 114). In the absence of adequate justification, the decision to secure the representation of specific groups seems to say more about what the convenors consider necessary for the plausible symbolic representation of the society than it signals a substantive interest in securing the inclusion of the marginalised in the deliberative forum.

Securing equal voice for the disadvantaged

All supporters of innovative democratic designs agree that background inequalities (in status and in resources) often translate into an inequality of voice understood as the capacity to influence the process of opinion formation. Even if we assume that the chosen method of participant selection secures the adequate presence of members of disadvantaged groups, one still needs to ensure that the inequalities in status and resources that exist between selected participants will not have a significant adverse impact on their effective capacity to enjoy equal voice within the design (Fishkin, 2013: 497). In this section, I attempt to answer two questions: firstly, what are the conditions that mini-publics see themselves as having to satisfy in order to offset (as much as possible) the impact of background inequalities? Secondly, what do these conditions say about the understanding of political equality?

Deliberative mini-publics share three basic features that equalise opportunities for influence. Firstly, they insist on the importance of providing all participants with adequate and balanced information to limit informational inequalities. Differences between designs over this issue concern whether or not the information is provided in advance of the event, whether it is transmitted mainly through oral presentation by invited experts and/or in the form of written briefing materials. There are also differences in the way the information itself is put together: in some instances, stakeholders have a role to play, while in other instances not.¹⁵

The second basic feature is that deliberation takes place within relatively small groups and that it is usually facilitated by trained personnel, to make sure that participation is not monopolised by anyone, that those who are less secure (women, the less educated, members of minority groups) get a fair chance to speak and that norms of mutual respect are enforced. This means that in devices involving a large number of people, participants are randomly divided into smaller groups (ranging from 10 to 20 participants) to provide a space in which deliberation may actually occur. The third basic feature is the provision for plenary meetings where participants can ask questions to a balanced panel of experts or stakeholders in order to clarify specific issues. The idea is to provide every participant with the chance to ask questions and hear the same answers.¹⁶

In some designs, notably deliberative polls and ChoiceDialogues, there is no expectation that discussions should lead to a common result like a consensus report or a set of recommendations. The individual opinions of participants are recorded at the end of the deliberative event in confidential questionnaires and aggregated. This is to avoid any artificial ‘push for consensus’ and to limit the ‘social comparison effect’ (Fishkin, 2013: 498).

These different procedures are meant to achieve deliberative exchanges in which differences in social and economic status are neutralised to a significant extent. In other words, the intention is to create a ‘safe space’ in which ordinary citizens, including members of disadvantaged groups, feel encouraged to express their views and participate fully in the discussions.

A prior condition

To achieve a safe space and ensure the effectiveness of the features presented above, a prior condition is set: special interests and partisans cannot enter the discussions as participants, though they may be included in the process as experts or contribute to the material that is put together to inform participants. This explains the required use of sortition as a method of participant selection and the rejection of self-selection, even if an appropriate incentive structure could secure the presence of the disadvantaged.¹⁷ On this view, the problem with voluntary self-selection is not simply that it leads to the overrepresentation of the more affluent or the better educated, but that it increases the risk that the process will be overwhelmed by partisans, activists, mobilised citizens, that is all those who are concerned enough about an issue to want to participate without first having been invited to do so (Goodin and Dryzek, 2006; Hendriks, 2011: 50; Ryan and Smith, 2014).¹⁸

Consider the notion of an ordinary citizen.¹⁹ The term is often used as the opposite of expert or specialist: ordinary citizens are those who have no specialised knowledge about the issue at stake. This contrast is not distinctive of deliberative mini-publics: processes associated with ‘empowered participatory governance’²⁰ are also intended to act as contexts in which non-experts can participate meaningfully in the process of opinion formation. Both families of design make the implicit claim that lay citizens are capable of making worthwhile contributions to political debate on complex topics.

Contrasting ordinary citizens to partisans and special interests calls to mind the distinction often made in the literature on participation between ‘citizens’ and ‘stakeholders’ (Kahane et al., 2013). ‘Stakeholders’ refer to the representatives of advocacy groups and organised interests while the citizens refer to unaffiliated individuals, citizens who do not act on behalf of a formally organised group. This distinction is not without difficulties as the term stakeholders conflates very different realities: groups with varying degrees of structure, levels of organisation and resources (Kahane et al., 2013: 6). But it does capture a great deal of what supporters of deliberative mini-publics mean when they contrast ordinary citizens to partisans and special interests: individuals who act on behalf of organised groups are attached to a particular perspective and are less open to changing their preferences (Hendriks et al., 2007).

But the two contrasts do not entirely overlap since, according to the common understanding of the stakeholder/citizen distinction, while ‘an individual who acts on behalf of an organized group focused on an issue or objective is a “stakeholder”’, an ‘individual passionate about a particular issue or political goal is categorized as a “citizen”’ (Kahane et al., 2013: 6). In other words, the stakeholder/citizen distinction does not exclude from the category of ordinary citizen the passionate or mobilised citizen, who may well be a member of an association. This mobilised citizen – who would be inclined to participate voluntarily in a process of public consultation on an issue close to her heart – may not be ordinary enough for deliberative mini-publics. Random selection is needed to ensure that passionate and engaged citizens – though not forbidden to participate – be kept in smaller numbers than would likely be the case in any process relying on voluntary self-selection.

Why should partisans or mobilised citizens be kept at bay? Two reasons are given: firstly, they will tend to take up all the room and leave little space for ordinary citizens to express their views. And their presence, it is thought, will have more impact on those with less status and less resources: women, members of minority groups, those with less education, etc., who may well be unable to find their voice amid all the noise. Secondly, their presence will prevent the magic of deliberation from operating. And here the argument is similar to Rousseau’s reflections on how the presence of ‘partial societies’ (sociétés partielles) can prevent the general will from emerging as opposed to differences between individual opinions (1964: 372). If deliberation is tracked essentially by the transformation of preferences, the presence of partisans should be avoided since they are less open to changing them.

While it is difficult to object to the desire of protecting ordinary citizens from the pressure of strong partisans and special interests, we must also realise that doing so by using sortition turns the deliberative forum into an ‘invited deliberative space’, disconnected from the thicker social context of ordinary civic life (Von Lieres and Kahane, 2007: 142). As Sintomer (2011a: 265; my translation) writes, ‘designs that rely on sortition (…) gather together citizens who are disembedded from their pre-existing social relations and “artificially” put in the presence of each other’. And this may make it more difficult for individuals coming from disadvantaged backgrounds, and with little prior experience in associational life, to voice their distinct needs and points of view.

Voice, agency and political equality

At issue is not the notion of interest itself. Fishkin, for instance, recognises that those who are less well off, less educated or members of minority groups, and who tend to participate less in political life (i.e. those he calls the non-voters), have interests that may well differ from the interests of most voters. If their ‘raw’ opinion – as recorded by opinion polls – is often similar to that of voters, his hope is that by engaging them in the political process, that is by providing them with information about the facts on the ground and having them participate in facilitated small group discussions, the deliberative poll gives them the means to develop a considered judgement that differs from their top of the head opinion and, presumably, expresses their specific interest (Fishkin, 2009: 52).

But to work in this way, as we saw above, the deliberative forum must provide a space insulated from the pressure and intensity of partisans and special interests. The assumption is that discovering and expressing one’s interest is essentially an individual, not a collective, process, that is informed by one’s participation in the deliberative forum.

That the underlying conception of interest formation is individualist can be seen in two features shared by deliberative mini-publics. Firstly, though the requirement of representativeness or inclusiveness means that the presence of individuals belonging to specific groups is secured through the selection method, these individuals are usually asked to participate as individuals and told that they are not there to represent what they take to be the interests or views of those groups. Secondly, the forum is not organised as to facilitate individuals who belong to specific groups to come together and articulate a shared view in the deliberations. On the contrary, the random assignment of individuals to small discussion groups makes it more difficult for individuals sharing specific identities to collaborate and articulate a common view.

Note that each of these two features also serves important democratic goods. The first feature speaks to the complexity of individual identity and the recognition that participants should be free to decide which aspect of their complex selves they present within deliberative exchanges. To ask participants to represent or express ‘their’ group’s perspective and interests is to ascribe to them a burden that they may not want or feel able to shoulder.

The second feature – the random assignment of individuals to small discussion groups – is justified by the need to secure the groups’ heterogeneity. It is presented as a key condition in getting participants to consider different or opposing arguments and points of view and to discuss them with individuals who come from very different backgrounds (Fishkin and Luskin, 2005: 289). The requirement reduces the risk of groupthink and the kind of polarisation effect that Cass Sunstein (2002) associates with deliberation among the like-minded. In short, this feature is an important condition in securing the quality of deliberation within the small discussion groups that play a key role in deliberative mini-publics.

While these features serve important democratic goods, they may also weaken the capacity for effective participation by members of disadvantaged groups. This point is well made by Von Lieres and Kahane (2007) in their account of the National Citizens Dialogue convened by the Canadian government to deliberate on the future of health care. The Citizens’ Dialogues included 12 one-day sessions across the country, each involving about 40 citizens randomly selected. It followed the ChoiceWorks methodology – in many ways similar to deliberative polls.²¹ In their analysis, Von Lieres and Kahane highlight the difficulties encountered in securing Aboriginal participation in the process. The Aboriginals who were selected often did not show up to dialogue sessions and, when present, typically remained ‘very quiet’ (Von Lieres and Kahane, 2007: 142).

Why was this so? Von Lieres and Kahane highlight two factors: firstly, the forums were structured in a way that ‘focused participants’ attention on what they shared as generic, individual citizens’. This followed the assumption discussed above that deliberation is more successful when individuals participate as unaffiliated citizens, rather than stakeholders. Secondly, since the ChoiceWorks methodology treats participants as individuals, ‘issues of group-based marginalization and inclusion can be dealt with, at most, as issues of protocol, equal speaking time, and imprecations to take all views seriously’ (Von Lieres and Kahane, 2007: 142). There is little possibility to question basic mainstream assumptions about what counts as a good reason in a particular case or what is the appropriate framing of the issues in discussion. In short, the design was ill equipped to address the kind of group-based marginalisation experienced by Aboriginal people.

The organisers themselves seem to have realised this. They were disappointed by how little the process was able to engage Aboriginal participants and they attempted – as the Citizens’ Dialogues proceeded across Canada – to alter the design in order to strengthen Aboriginal participation: Firstly, Aboriginal participants were clustered in small group discussions in order to ‘build confidence and voice’. Secondly, an attempt was made to include in the process more Aboriginal people than required by the methodology, using community channels. In the event, neither of these initiatives seems to have had much success.

Von Lieres and Kahane suggest that to succeed, the attempt to build Aboriginal voice in the Citizens’ Dialogues needed to go beyond lessening the isolation of Aboriginal participants. They propose instead the creation of distinct deliberative spaces through mechanisms such as ‘group caucusing’ within the larger forum or the setting up of ‘separate deliberative streams’. These mechanisms would enable participants from marginalised groups ‘to develop strategies for naming and countering dynamics of exclusion; build confidence and capacities; reflect upon and consolidate common goals and identities; and make space for differences within the group’ (Von Lieres and Kahane, 2007: 143).

The suggestion is that in the case of deep marginalisation, a modest amendment of the deliberative forum’s procedures is not enough: members of marginalised groups need to get together and deliberate among themselves in a distinct forum in order to build their capacity for voice (recognising that this voice may itself be diverse). In other words, if silence is to be overcome, the process of individual interest and opinion formation needs to be conceptualised as part of a collective process that may have to take place prior to deliberative exchanges within the wider forum. In this kind of circumstances, the assumption that the process of interest and opinion formation takes place at the same time as one participates in the wider deliberative forum cannot be maintained. As Fung (2003: 348) suggests, when background inequalities between participants are such that they have highly unequal ‘capabilities to propose ideas and make claims’, then ‘discourse aimed mainly toward will-formation (…) may be a necessary precursor to a fuller deliberation’.

In such cases, moreover, convenors may also consider organising the distinct deliberative space as a hybrid forum, in which stakeholders and activists participate alongside unaffiliated citizens (recruited either through some form of random selection or through community channels).²² Embedding the deliberations within a thicker civic and social context gives ‘ordinary’ citizens the opportunity to collaborate with people who already have a significant political and social experience, and this may have an enabling effect.

But choosing these avenues also involves costs. As Von Lieres and Kahane (2007: 144) readily acknowledge, the creation of distinct spaces ‘enacts presumptions about individual and collective identities – both in treating Aboriginal peoples as a bounded (if internally complex) group, and by supporting narratives of separateness or incommensurability between Aboriginal and non-Aboriginal beliefs and interests’. They also note that creating a separate forum does not eliminate the need to bring Aboriginal and non-Aboriginal people into dialogue together, given the close interrelations between policy decisions affecting Aboriginal health and the health of Canadians. Finally, the thicker civic spaces that hybrid or embedded forums provide, while facilitating the expression of an effective collective voice, may also expose participants to the unequal power relations that run between and across the different groups involved.²³

How we assess these costs depends on the case at hand. The situation of Aboriginal peoples in Canada presents an extreme case of marginalisation in which the persisting effects of colonialism, of cultural, political, legal and economic domination are compounded and impose a heavy burden on individuals and communities. It is unrealistic to presume that the impact of severe marginalisation can be successfully offset, and opportunities for influence effectively equalised, through features like providing all participants with the same basic information and facilitated discussions. But where inequalities are less extreme, or uncompounded, then these features may be enough to mitigate the effects of background inequalities and empower individual participants.

This suggests the need to carefully analyse each situation, including the issue to be discussed and whether it affects significantly a group experiencing serious forms of marginalisation and inequality. If the answer is: ‘yes’, then considering specific provisions to address the group’s situation is warranted. If, upon reflection, it is decided against such provisions, then justification (or an alternative proposal) should be offered.

What is clear is that each alternative involves costs in terms of the different democratic goods that the deliberative forum may aim to realise (effective and authentic voice for all participants, dialogue and understanding between participants from different backgrounds, etc.) and, in most instances, no single design will be able to achieve all these goods in one place at the same time.

Conclusion

Which conception of political equality do deliberative mini-publics promote in the face of persisting background inequalities? I have explored this question by examining the procedures and devices put forward to secure the equal presence and the equal voice of citizens within those designs.

In using some form of random selection to select participants, deliberative mini-publics do not aim to provide every citizen of the relevant demos with an equal opportunity to participate. But they are often described as giving citizens an equal chance of being selected to participate. We have seen that few designs can actually claim that achieving this principle is their objective since they rarely work as pure lotteries. In many cases, the principle of equal chance is sacrificed to achieve some form of ‘diversity’ or ‘inclusiveness’, selecting participants that reflect a broad cross-section of the population. The notion of diversity refers to the sociological makeup of the society, to the existence of different groups to which individual citizens identify: women and men, young and old, etc., while part of the desire to achieve inclusiveness can be related to a more specific concern to secure the participation of those who are often marginalised in the political institutions of their society. Here political equality appears to be also related to groups, not strictly to individuals. But the relation remains sketchy and there rarely seems to be a systematic effort to clarify which groups ought to be included in the deliberative forum – and in what numbers – given the nature of the issue under discussion or the experience of marginalisation encountered by some within the society.

In the absence of such a reflection, the requirement of diversity or inclusiveness, as it relates to equality, appears to have a primarily symbolic meaning: that of publicly affirming the basic equal capacity of all citizens, no matter whether they are male or female, white or non white, rich or poor, to participate in discussions over complex policy issues.

When considering the issue of voice, we have seen that equality is understood basically in terms of providing participants with equal opportunities for influence. The challenge is to achieve an effective equality of opportunity for all participants, despite the likely impact of inequalities in status, education levels, etc. Equalising opportunities for the disadvantaged is usually understood in terms of providing all individual participants with adequate and balanced information and securing equal opportunities for voice through appropriate facilitation and protocols. These provisions’ objective is a form of individual empowerment as citizens who have little or no civic experience receive the opportunity to deliberate with others over complex issues.

But this understanding of what needs to be done to achieve an effective equality of opportunities between participants seems to underestimate the effects of deep background inequalities on the capacity for voice of those belonging to marginalised groups. In such cases, the individualist conception of interest and opinion formation that informs deliberative mini-publics needs to be questioned: for individuals who experience severe forms of political poverty, finding their voice is also part of a collective endeavour. This may call for a process of capacity building that provides individuals from such groups with the opportunity to deliberate together in distinct spaces, prior to joining their fellow citizens in the wider deliberative forum. These distinct spaces may have to be constructed in specific ways, as hybrid forums not entirely disembedded from the thicker context of community groups and organisations.

In short, what is needed is a more careful reflection on background inequalities and their relation to political agency. If inclusiveness is a serious preoccupation of designers, then it cannot be divorced from a reflection on how presence can translate into effective voice, given the particular challenges faced by specific sections of the population. And this should lead convenors to confront the difficult choices that setting up a deliberative mini-public will involve between the different democratic goods that they aim to achieve, in the context of societies beset with increasing inequalities.

Footnotes

Acknowledgements

Earlier versions of this article were presented at a conference on the Participatory Turn: Mirages and Realities organised by the Centre de recherche sur les politiques et le développement social (CPDS) at the Université de Montréal (February 2015), at a conference on Political equality: Values and implications were held in May 2015 at the Université de Louvain (UCL) and at the Department of Philosophy at UQÀM (September 2015). For their helpful comments on these earlier drafts, I wish to thank the participants at these events and, in particular, Hervé Pourtois, Siba Harb, John Pitseys, Thomas Christiano, Charles Girard, and Christophe Malaterre, Laurence Bherer, Pascale Dufour and Françoise Montambeault. Many thanks also to the two anonymous reviewers of this journal for their extensive and insightful comments.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Notes

Appendix 1

References

Abers

(2000) Inventing Local Democracy. Grassroots Politics in Brazil, Boulder/London: Lynne Rienner Publishers.

Blais

(2000) To Vote or Not to Vote. The Merits and Limits of Rational Choice Theory, Pittsburgh: Pittsburgh University Press.

Bohman

(1997) Deliberative democracy and effective social freedom: Capabilities, resources and opportunities. In: Bohman

Rehg

(eds) Deliberative Democracy, Cambridge, MA: MIT Press, pp. 321–348.

Brown

(2006) Survey article: Citizen panels and the concept of representation. The Journal of Political Philosophy 14(2): 203–225.

Caluwaerts D and Reuchamps M (2014) The G1000. Facts, figures and some lessons from an experience in deliberative democracy in Belgium. In: The Malaise of Electoral Democracy and What to Do About It. Available at: www.rethinkingbelgium.eu (accessed 20 February 2016).

Carson

Martin

(1999) Random Selection in Politics, Westport, CT: Praeger.

Chambers

(2009) Rhetoric and the public sphere: Has deliberative democracy abandoned mass democracy? Political Theory 37(3): 323–350.

Christiano

(2004) The authority of democracy. The Journal of Political Philosophy 12(3): 266–290.

Cohen

Fung

(2004) Radical democracy. Swiss Political Science Review 10(4): 169–179.

10.

Coote A and Lenaghan J (1997) Citizens' Juries: Theory into Practice. London: Institute for Public Policy Research.

11.

Dienel

Renn

(1995) Planning cells: A gate to ‘fractal’ mediation. In: Webler

Wiedemann

(eds) Fairness and Competence in Citizen Participation: Evaluating Models for Environmental Discourse, Dordrecht: Kluwer, pp. 117–141.

12.

Fishkin

(2009) When the People Speak: Deliberative Democracy and Public Consultation, Oxford: Oxford University Press.

13.

Fishkin

(2013) Deliberation by the people themselves: Entry points for the public voice. Election Law Journal 12(4): 490–507.

14.

Fishkin, JS, Camarano E, Luskin RC, et al. (2010) Deliberative participatory consultation: The first Deliberative Poll in Porto Alegre. Available at: www.cdd.stanford.edu/research/page/2/ (accessed 20 February 2016).

15.

Fishkin

Farrar

(2005) Deliberative polling: From experiment to community resource. In: Gastil

Levine

(eds) The Deliberative Democracy Handbook. Strategies for Effective Civic Engagement in the 21st Century, San Francisco: Jossey-Bass, pp. 68–79.

16.

Fishkin

Luskin

(2005) Experimenting with a deliberative ideal: Deliberative polling and public opinion. Acta Politica 40: 284–298.

17.

Fishkin

Rosell

Shepherd

et al. (2004) ChoiceDialogues and deliberative polls: Two approaches to deliberative democracy. National Civic Review 93(4): 55–63.

18.

Fournier

Van Der Kolk

Carty

et al. (2011) Who were the participants? In: Fournier

Van Der Kolk

Carty

et al. (eds) When Citizens Decide: Lessons from Citizen Assemblies on Electoral Reform, Oxford: Oxford University Press, pp. 51–64.

19.

Fung

(2003) Survey article: Recipes for public spheres: Eight institutional design choices and their consequences. The Journal of Political Philosophy 11(3): 38–367.

20.

Fung

Wright

(2003) Thinking about empowered participatory governance. In: Fung

Wright

(eds) Deepening Democracy. Institutional Innovations in Empowered Participatory Governance, London: Verso, pp. 3–42.

21.

Gaxie

(1978) Le cens caché. Inégalités culturelles et ségrégation politique, Paris: Seuil.

22.

Goodin

Dryzek

(2006) Deliberative impacts: The macro-political uptake of mini-publics. Politics & Society 34(2): 219–244.

23.

Goodwin

(2005) Justice by Lottery, Exeter: Imprint Academic.

24.

Griffin

Abdel-Monem

Tomkins

(2015) Understanding participants representativeness events: A case-study comparing probability and non-probability recruitment strategies. Journal of Public Deliberation 11(1): 1–29.

25.

Hansen

(1991) The Athenian Democracy in the Age of Demosthenes, Oxford: Blackwell.

26.

Hendriks

(2005) Consensus conferences and planning cells: Lay citizen deliberation. In: Gastil

Levine

(eds) The Deliberative Democracy Handbook. Strategies for Effective Civic Engagement in the 21st Century, San Francisco: Jossey-Bass, pp. 80–110.

27.

Hendriks

(2011) The Politics of Public Deliberation. Citizen Engagement and Interest Advocacy, Houndmills, Basingstoke: Palgrave Macmillan.

28.

Hendriks

Dryzek

Hunold

(2007) Turning up the heat: Partisanship in democratic innovation. Political Studies 55(2): 362–383.

29.

James

(2008) Descriptive representation in the British Columbia Citizens’ Assembly. In: Warren

Pearse

(eds) Designing Deliberative Democracy: The British Columbia Citizens’ Assembly, Cambridge, NY: Cambridge University Press, pp. 106–126.

30.

Kahane

Loptson

Herriman

(2013) Stakeholder and citizen roles in public deliberation. Journal of Public Deliberation 9(2): 1–35.

31.

Knight

Johnson

(1997) What sort of political equality does deliberative democracy requires? In: Bohman

Rehg

(eds) Deliberative Democracy, Cambridge, MA: MIT Press, pp. 280–319.

32.

Lafont C (2015) Deliberation, participation, and democratic legitimacy: Should deliberative mini-publics shape public policy? The Journal of Political Philosophy. 23(1): 40–63.

33.

Landemore

(2012) Democratic Reason: Politics, Collective Intelligence and the Rule of Many, Princeton/Oxford: Princeton University Press.

34.

Mansbridge

(1999) Should blacks represent blacks and women represent women? A contingent ‘yes’. The Journal of Politics 61(3): 628–657.

35.

Montesquieu C (1989) The Spirit of the Laws. Translated and edited by Cohler AM, Miller BC and Stone HS. Cambridge: Cambridge University Press.

36.

O’Flynn

Sood

(2014) What would Dahl say? An appraisal of the democratic credentials of deliberative polls and other mini-publics. In: Grönlund

Bächtiger

Setälä

(eds) Deliberative Mini-Publics. Involving Citizens in the Democratic Process, Colchester: ECPR Press, pp. 41–58.

37.

Phillips

(1996) Dealing with difference: A politics of ideas, or a politics of presence? In: Benhabib

(ed.) Democracy and Difference. Contesting the Boundaries of the Political, Princeton: Princeton University Press, pp. 139–152.

38.

Rawls

(1971) A Theory of Justice, Oxford: Oxford University Press.

39.

Rousseau

J-J

(1964) Oeuvres complètes III. Du contrat social. Écrits politiques, Paris: Gallimard.

40.

Ryan

Smith

(2014) Defining mini-publics. In: Grönlund

Bächtiger

Setälä

(eds) Deliberative Mini-Publics. Involving Citizens in the Democratic Process, Colchester: ECPR Press, pp. 9–27.

41.

Ryfe

Stalsburg

(2012) The participation and recruitment challenge. In: Nabatchi

Gastil

Leighninger

et al. (eds) Democracy in Motion: Evaluating the Practice and Impact of Deliberative Civic Engagement, Oxford, NY: Oxford University Press, pp. 43–56.

42.

Sanders

(2010) Making deliberation cooler. The Good Society 19(1): 41–47.

43.

Sintomer

(2011a) Délibération et participation: affinité elective ou concepts en tension? Participations 1(1): 239–276.

44.

Sintomer

(2011b) Petite histoire de l’expérimentation démocratique. Tirage au sort et politique d’Athènes à nos jours, Paris: La Découverte.

45.

Sintomer

Herzberg

Röcke

(2008) Les budgets participatifs en Europe, Paris: La Découverte.

46.

Smith

(2009) Democratic Innovations. Designing Institutions for Citizen Participation, Cambridge: Cambridge University Press.

47.

Smith

Wales

(2000) Citizens’ juries and deliberative democracy. Political Studies 48: 51–55.

48.

Stone

(2009) The logic of random selection. Political Theory 37(3): 375–397.

49.

Sunstein

(2002) The law of group polarization. The Journal of Political Philosophy 10(2): 175–195.

50.

Swift

(2006) Political Philosophy. A Beginner’s Guide for Students and Politicians, Cambridge: Polity Press.

51.

Thompson

(2008a) Deliberative democratic theory and empirical political science. Annual Review of Political Science 11: 497–520.

52.

Thompson

(2008b) Who should govern who governs? The role of citizens in reforming the electoral system. In: Warren

Pearse

(eds) Designing Deliberative Democracy: The British Columbia Citizens’ Assembly, Cambridge, NY: Cambridge University Press, pp. 20–49.

53.

Von Lieres

Kahane

(2007) Inclusion and representation in democratic deliberations: Lessons from Canada’s Romanow commission. In: Cornwall

Coelho

VSP

(eds) Spaces for Change, London: Zed Books, pp. 131–151.

54.

Williams

(1998) Voice, Trust and Memory. Marginalized Groups and the Failings of Liberal Representation, Princeton: Princeton University Press.

55.

Wolfinger

Rosenstone

(1980) Who Votes?, New Haven: Yale University Press.