Mini-publics express a variety of different decision rules. At one end of the spectrum are citizens’ juries and consensus conferences, where participants collectively craft recommendations in response to a charge; at the other end, deliberative polling, where participants offer individual responses to a survey instrument that is delivered pre- and post-deliberation. Somewhere in-between is the Citizens’ Assembly model where participants develop solutions in response to a charge, but vote in private at the end of the process to decide the final recommendation. The relative impact of these decision rules on the preceding deliberation lacks systematic attention within the theoretical literature.
The structure of deliberative polling limits the potential for creativity amongst participants. While participants hear evidence from and are able to question competing experts, they are not in a position to develop alternative policy solutions: the preset questionnaire structures the output of the design. This entails the unlikely assumption that the survey instrument is able to capture the various potential directions that deliberations might take between participants and the different understandings of a policy issue that might emerge during the process. This is indicative of a significant difference in scope between deliberative polls and other designs: the function of the former is to provide a more informed sense of public opinion based on preordained questions; other mini-publics provide more freedom for citizens to develop recommendations for how to solve a current policy problem.
But the more open structure of other mini-publics may have its own unintended effects. First, there may be an implicit (or even explicit) pressure to achieve consensus between participants that suppresses conflict. The very name ‘consensus conference’ implies a search for areas of consensus on controversial issues of scientific and technology policy. As a project manager from the Danish Board of Technology stresses: ‘every effort is made to attain the greatest consensus between the lay-panel members on the actions to be recommended. Minority opinions should be allowed only when the process reveals very wide differences of opinion’ (Grundahl 1995: 37). There is a fine line between the search for consensus and the suppression of disagreement and much rests on the skills and approach of the facilitator - another reason why that particular role requires more theoretical attention.
We find growing concerns that the judgements of citizens in mini-publics may be shaped by irrational group dynamics, rather than reasoned argument (Merkle 1996: 607). Cass Sunstein offers an overview of relevant social psychology literature, arguing that deliberation in small groups may lead to group polarization - movement towards and adoption of more extreme positions (Sunstein 2000). He argues that there are two broad mechanisms at work: a reputational effect where participants aim to maintain their self-conception in relation to the group; and the effect of limited argument pools where participants tend to hear only arguments that reinforce their own point of view. However, his findings suggest that these tend to be properties of socially homogenous groups with a shared identity, rather than the more diverse groups that constitute mini-publics. Sunstein contrasts his findings with evidence from deliberative polls, which indicates that the nature of opinion change within the small discussion groups is not consistent with polarization (Sunstein 2000: 116; Luskin et al. 2002: 477-8; Ackerman and
Fishkin 2004: 61-5). The design of deliberative polls is a crucial factor in enabling depolarization: participants are highly diverse in their social perspectives; the process is facilitated to ensure openness; balanced information is provided; and citizens are not required to make decisions as a group - their opinions are sought in private. As Sunstein notes: ‘Fishkin’s experiments suggest that group polarization can be heightened, diminished, or possibly even eliminated by seemingly small alterations in institutional arrangements’ (Sunstein 2000: 117; see also Fishkin in this volume).
But, as we have already noted, one of the characteristics isolated by Sunstein - citizens are polled individually - is not common to all mini-publics. The design of the BCCA shares this characteristic: while citizens worked together to evaluate different electoral systems, their final decisions were made through secret ballot. The extent to which the freedom to craft recommendations in citizens’ juries and consensus conferences heightens the possibility of group polarization is a topic that deserves more attention and, again, arguably relates to the style of facilitation.