Deliberative Polls and Constitutional Juries

12.3.1 Deliberative Polls and the Community Values Approach

The previous sub-section explored some theoretical considerations pertaining to whether judicial deference to informed majority opinion could promote democratic or general legitimacy.

Fishkin noted that George Gallup touted polling as an instrument of democratic reform, calling it the “sampling referendum” (Fishkin 2005, 286-287). However, Fishkin mentioned that the “typical respondent answering the typical political atti­tude item has barely ever thought about the question before being interviewed and can call on precious little information in answering it” (Fishkin 2005, 287). The deliberative poll seeks to elicit, by contrast, deliberative public opinion. Fishkin’s work fits within the dominant contemporary democratic theory of deliberative democracy (Fishkin and Luskin 2000; Lafont 2015, 44). This has involved close examination of the ideal of deliberation. Fishkin’s deliberative public opinion is substantially similar, though, to the informed preferences that were mentioned in discussing Dahl (1991, 229-230).

Deliberative polls commence with a pre-deliberation survey of a statistically rep­resentative sample of the population (Luskin et al. 2002, 463). The sample is then invited to participate in a weekend discussion a few weeks later concerned with the same issues as those raised in the survey. Inducements offered include free accom­modation, meals and travel, and an honorarium.

Typically, participants in the weekend discussions have attitudes to the issues in question which are similar to the attitudes of those initially surveyed. They also have sociodemographic attributes similar to the population as a whole. They are generally a bit older, better educated and more interested and knowledgeable about those issues than non-participants, but not by much.^[317] The polls typically result in significant shifts in opinion. Two examples will suffice. In a poll concerned with which party to vote for in the 1997 British election, support for the Conservatives and for Labour decreased by 7 and 8 percent respectively, and increased for the Liberal Democrats by 22 percent.^[318] In a 2001 poll concerned with reconciliation between indigenous and non-indigenous Australians, there was over a 20 % increase in support for certain positions sympathetic to indigenous Australians, including an apology to the “stolen generation”, ie, an apology to indigenous people affected by the policy of forcibly removing children from their families and placing them in institutions or with non-indigenous families.^[319] The polls also typically produce gains in knowledge (Fishkin 2005, 290). Indeed, the changes in opinion seem infor­mation-driven: those who gain the most knowledge are most likely to change their opinions (Fishkin 2005, 291).

These polls demonstrate that empirical evidence of deliberative public opinion is obtainable. That evidence is not, of course, conclusive. Deliberative polls involve compromises. Their voluntary character affects their representativeness, but may well enhance their deliberative quality. Sampling error must be recognised, as with any poll, and also contingencies specific to deliberative polls, such as the selection of experts and the persuasiveness of particular experts in plenary sessions.

Furthermore, deliberative polls on issues relating to minority protection have tended to elicit opinions more sympathetic to these groups. The poll on indigenous Australians is one of several examples (Fishkin 2010, 69).^[320] The number of polls dealing with minority protection has, admittedly, been small, but it nevertheless provides some support for Waluchow’s assumption that reflective opinion may be more sympathetic to minority groups than the public opinion that influences legislation.

On the other hand, Waluchow’s hope that deliberative opinion will involve greater consensus than non-deliberative opinion is not supported. Significant minor­ities continued to oppose the majority positions in many polls. More significantly, the existence of an empirical approach to determining reflective public opinion throws into sharp relief the inadequacy of relying on assertions by judges as to what such opinion would endorse. It raises the question of whether this empirical method could assist in constitutional cases. If it can, the idea of grounding constitutional review in deliberative public opinion, which is found within the community values approach, could perhaps be vindicated.

12.3.2 Constitutional Juries

This sub-section mentions some proposals for employing deliberative public opin­ion in constitutional review. US scholar Ethan Leib recommended in 2002 that if judges wish to make claims about the community’s deliberative values, they could be obliged to consult a deliberative poll (Leib 2002, 369, 409).

By contrast, Argentinian scholar Horacio Spector recommends vesting a discre­tion in complainants rather than judges. He proposed in 2009 that complainants in constitutional cases might choose between a judge or jury trial (Spector 2009,117­118).^[321] Vesting the discretion in complainants rather than judges removes the pos­sibility of the slight or sham deference referred to in Sect. 12.2.3, where judges, for instance, only defer to informed public opinion on matters about which they are indifferent. Nevertheless, from a probabilistic perspective, it promotes less liberty as self-government than when such trials always involve juries. From a power- centred perspective, the fact that the arbitrary discretion not to employ juries is dispersed amongst complainants, rather than concentrated in judges, alleviates some of the evils of arbitrary power, such as the deference it can engender towards those who enjoy power. Nevertheless, it leaves such jury trials at the arbitrary dis­cretion of others and it thereby fails to directly promote liberty as self-government. Cases can involve trade-offs between rights. A complainant successfully asserting a right before a judge can lead, say, to the diminution of a different right, a diminu­tion that may not be endorsed by majority informed opinion. Rights cases implicate matters of collective concern.

Spector envisaged juries of around 36-72 members “to ensure the jurors repre­sent a great diversity of viewpoints in the whole community” (Spector 2009, 118).

In an article published in 2010, and without awareness of Spector’s article, I sketched and defended a proposal for an Australian court, called a Citizens’ Court, which employs constitutional juries (Ghosh 2010).^[322] I suggested that bill of rights matters be only decided by constitutional juries modelled on some features of delib­erative polls. The juries would be large, say, 200. Voluntary participation, with the assistance of inducements, could also be followed partly because this should enhance deliberation. Constitutional juries would, of course, be expensive. They should only be employed where it is fairly clear that legislation or an executive act may well be at odds with deliberative opinion about matters of principle and that the issue would be viewed as significant by deliberative public opinion. Juries should also play a role in determining matters that go to a hearing. With this proposal, the calling of constitutional juries would not lie in the discretion of a judge or a com­plainant, in the sense that they can chose between a judge- or jury-trial, and the juries would be large enough to enable a plausible claim of representativeness.

That article provided further details of the proposal, which will not be repeated. It suffices to mention here that if the Citizens’ Court could achieve the popular legitimacy that confers significant authority on its decisions, it could provide a check upon the other branches of government that is broadly consistent with delib­erative public opinion.

This was also attempted in the 2010 article, and that appraisal is summed-up here (Ghosh 2010, 348-352). On the process-related consideration of popular legiti­macy, I suggested that constitutional juries may eventually obtain greater popular legitimacy than courts. Being representative of the people is a strong source of legitimacy. On the outcome consideration, factors relevant to sound decision­making include diversity, analytical skills, deliberation, empathy, and capacity and willingness to act on understanding of justice reached. On diversity, the juries would do significantly better. Australian judges are appointed from an elite group within the legal profession. Furthermore, governments can sometimes decisively shape the ideological complexion of benches especially after a lengthy period in office. On analytical skills, judges would be superior, but this advantage can be reduced if juries are presented with accessible information. Furthermore, part of the complexity associated with judicial decisions is due to a formalist style of reason­ing, a consequence of which is discussion of precedent that goes beyond what is necessary to reach sound outcomes. It is true that judges can be expected to deliber­ate more than jurors, but much of the time expended by Australian judges is devoted to demonstrating that their decisions are authorised by previous cases rather than demonstrating that they are justified through substantial and open discussion of the competing policy considerations. On willingness of decision-makers to decide in accordance with conceptions of justice, concern about democratic legitimacy can lead judges to be deferential towards the elected branches of government. That is less likely to affect jurors. I suggested that juries could well make better decisions than judges. In that case, an additional process consideration is promoted. The juries provide a symbolic affirmation of citizens’ capacity to reach sound conclusions about justice when provided with appropriate deliberative opportunities.

The most detailed critique of my proposal (and indeed Spector’s) was provided by US philosopher Christopher Zurn (2011). Zurn has himself used the model of deliberative polls in recommending the democratisation of the constitutional amendment process. He has suggested juries could determine what constitutional amendment proposals would go to a referendum (2006, 336).^[323] To elaborate, the first step would be a citizen initiative, involving the collection of signatures in favour of a proposal, or a referral from a legislature. The second step would be cer­tification of a proposal, requiring agreement by three separate deliberative and rep­resentative forums spaced over a significant time span. The third step would be a deliberation day concerned with the certified proposal, involving as much of the population as possible in deliberative forums. The fourth step would be a popular referendum on the certified proposal.

Zurn’s critique of constitutional juries is launched from a stance that is support­ive of using representative samples of the population to enhance the deliberative democratic character of decision-making relating to constitutional matters. Zurn’s critique merits a full response, but I will only make four observations here. First, some of Zurn’s criticisms are based on the assumption that the Citizens’ Court would only examine the legality of actions by the executive. He suggested, for instance, that the Citizens’ Court would lack effectiveness due to its inability to review legislation and that ordinary courts would step into this vacuum (Zurn (2011, 81, 83, 85-86). In fact, I envisaged constitutional juries as primarily concerned with the validity of legislation (Ghosh 2010, 349).

The second observation is that his support for representative groups of citizens being involved in democratising constitutionalism sits awkwardly with his worry that the juries would be relatively if not entirely unconstrained by an interest in achieving coherence with past juries’ decisions or to follow judicially developed substantive or methodological doctrine (Zurn 2011, 83-84). With respect to prece­dent, the extent to which consistency should be favoured even where a decision­maker believes that some previous decisions are wrong can be a difficult value-choice. It is unclear why Zurn’s faith in the rationality of deliberative polls does not extend to this choice. With respect to judicially developed doctrines, Zurn mentioned principles of statutory interpretation (Zurn 2011, 83). However, I envis­aged trials with a panel of judges presiding over their conduct and the parties enjoy­ing legal representation. It should not be assumed that juries would ignore the advice they receive. Perhaps the most important question here is not whether juries may be less constrained by precedent or legal methodology but whether juries may be in a better position to reach attractive understandings of how strong those con­straints should be in particular cases.

However, Zurn could reply that even if juries give appropriate weight to values such as predictability, which might be furthered by following precedent, they will nevertheless be engaged in what he describes as constitutional legislation, which properly belongs to the people (Zurn 2011, 85). The distinction between constitu­tional legislation and the application of constitutional norms in a particular case is a distinction between the application of constitutional norms in a way that involves significant, as opposed to modest, development of those norms. Zurn recognises that judges engage in some constitutional legislation, but says that juries would engage in much more.^[324]

While Zurn has perhaps overstated the contrast between juries and judges by assuming that juries would be entirely unconstrained by precedent, juries may well be bolder than judges not only because the former may be less inhibited by concerns about their own democratic legitimacy but also because, as Zurn indicates, I pro­pose a fairly open-ended bill of rights. This leads to my third observation. Zurn’s concern about the Citizens’ Court engaging in constitutional legislation sits oddly with his disapproval of my suggestion that parliament with a 60 % majority should be able to abrogate Citizens’ Court decisions or pre-empt review (Zurn 2011, 80). The possibility of pre-empting review could easily be dropped from my proposal, but a power of abrogation is appropriate for reasons including the fallibility and novelty of constitutional juries. Zurn, however, fails to acknowledge that the pos­sibility of parliamentary abrogation denies to Citizens’ Court determinations the fundamental status that amendments through a referendum process enjoy. Constitutional legislation that is unalterable by parliament will require a process that has a stronger claim to legitimacy than constitutional legislation that is alterable by parliament through a fairly weak supermajoritarian requirement.

My final observation is that one might wonder if Zurn has identified an appropri­ate process for the most fundamental of constitutional legislation. Zurn sets a gold standard for legitimate amendments, but in a broader sense it risks legitimacy by privileging the status quo to an extent that goes beyond the desirability of constitu­tions enjoying some stability. Zurn has not challenged my argument in 2010 that very few amendments might make it through his proposed process (Ghosh 2010, 344). The sheer expense of the process, with its use of deliberation days, might be a factor that juries consider in determining whether to certify a proposal. And that is, of course, assuming that a party is able to clear the hurdle of the initiative process, with the substantial resources that involves. Constitutional legislation is likely to remain largely with judges. By contrast, my proposal enables constitutional legisla­tion at admittedly a less fundamental level by a body that is deliberative and inclusive.

As indicated, my intention is not to provide here a full response to Zurn’s cri­tique. Instead, the principal aim of this chapter is to link my proposal to the com­munity values approach in order to illuminate both. Light on constitutional juries is thrown by recognising that there is a strand of justification found in judicial reason­ing and scholarly defences of judicial review which suggests that judicial decisions would be more legitimate if they were based on reasonable community values. It was argued that these values should be understood as informed majority opinion, and a consideration of how this can promote the democratic legitimacy of constitu­tional review leads to consideration of constitutional juries.

Understanding of the community values approach is also enhanced. First, attempting to connect the community values approach to constitutional juries involves a sympathetic look at the approach’s aim of promoting democratic legiti­macy by connecting constitutional review to community values. The previous cri­tiques of this approach do not examine this ideal and how it fits with the democratic theories referred to by its scholarly followers. They fail, therefore, to explore this provocative aspect of the community values approach. Secondly, the critiques do not consider if the community values that the approach refers to can be understood in an empirical way that is consistent with democracy.

Of course, the discussion also sheds critical light on the community values approach. Pointing to the existence of an empirical approach to deliberative public opinion more starkly reveals the inadequacy of relying on judicial intuition. An additional and final point is that it also renders more problematic a narrow focus on the outcomes of judicial review when assessing its legitimacy. I have mentioned that the deviation from political equality found in the popular branches of government can be used by defenders of judicial review to suggest that the latter does not signifi­cantly undermine political equality. However, if constitutional juries are feasible, the existence of judicial review instead indicates a lost opportunity for significantly greater political equality. Rather than judges interpreting fundamental rights, a rep­resentative group of citizens could themselves decide cases that another representa­tive group views as raising the greatest concern.

12.4