Advancing deliberative reform in a parliamentary system: prospects for recursive representation

Abstract

Recent theories of democratic representation push beyond ‘minimalist’ notions that only rely on periodic elections to connect officials and constituents. For example, Jane Mansbridge (2019) calls for ‘recursive representation’, which seeks ongoing, two-way interaction between representatives and their constituents. Given the scale and complexity of modern representative democracies, how can such ambitious proposals be translated into practice? We analyze two Deliberative Town Halls (DTHs) convened with a Federal Member of Australian Parliament in 2020 to discuss a complex issue, mitochondrial donation, ahead of a parliamentary debate and conscience vote on this issue. Drawing on interviews with participants, we argue that democratic innovations such as DTHs can contribute to realizing recursive representation when three criteria are met: authenticity, inclusion, and impact. We discuss the significance of each criterion and the role of DTHs in advancing recursive representation in a parliamentary system.

Introduction

Efforts to realize the normative promise of deliberative politics face an uphill struggle when they confront the entrenched institutions of existing representative democracies. Here, deliberation can be defined broadly as mutual communication that involves ‘weighing and reflecting on preferences, values, and interests regarding matters of common concern’ (Bächtiger et al., 2018: 2) across equal participants (Gutmann and Thompson, 2004). Deliberative democracy promises to embody meaningful political participation, enhance democratic legitimacy, and yield collective decisions responsive to the concerns of relevant actors. Yet when embodied in practice, for example in structured forums, these aspirations can face various obstacles. Elites who prosper in the existing system will not willingly share more of their influence in the interests of inclusive deliberation.

How then do we get from a relatively minimal democracy in which all that matters is that choices such as votes occasionally get aggregated through mechanisms via elections to a fuller deliberative democracy? Democratic minimalists believe we should not even try; we should instead be content with competitive elections and the peaceful transfer of (elite) power (Schumpeter, 1942; Riker and Weingast, 1988; Przeworski, 1999). In contrast, participatory democrats think that deliberative reform can be too timid and want to shift power away from (or replace) electoral institutions of representation with more participatory democratic practices (Pateman, 2012). Some deliberative democrats seek to replace the legacy institutions of representative democracy with assemblies constituted by random selection from the citizenry (Landemore, 2020).

But before moving quickly to either minimalist resignation or advocacy of radical change, we believe it is advisable to explore the prospects for making the representative aspects of existing institutions more deliberatively democratic. Neblo et al. (2018) developed the theory of ‘Directly Representative Democracy’ to test the value of deliberative reforms of representative institutions in the American case and applied it using Deliberative Town Halls (DTHs) that engaged members of Congress with their constituents. DTHs were designed to reverse the growing disconnection between elected officials and their constituents (Foa et al., 2020), which has been a pressing issue in contemporary representative democracies (Hendriks et al., 2020). In DTHs, representatives discuss pressing policy issues in a structured environment with randomly selected constituents who have been provided with high-quality background materials and information about the issues at stake. Research with sitting members of Congress in the USA reveals that DTHs are very effective at re-engaging demobilized portions of the public, restoring their trust in representatives and political institutions, and revealing reflective public opinion to legislators on issues for which they otherwise would have had to guess at their constituents’ considered views (Neblo et al., 2010, 2018).

These promising results appeared to advance what Mansbridge (2019) calls recursive representation – ongoing deliberative engagement between representatives and their constituents, very different from seeing representation in terms of supposed mandates emerging from campaigns and elections (see also Saward, 2014). For Mansbridge, such engagement should be shaped by deliberative ideals such as mutual respect, inclusion, attentive listening, and the absence of manipulation.

The idea of recursive representation suggests a relationship between specific legislators and specific constituents, where legislators are relatively unconstrained in the positions they can take. As such, it fits the USA context. It is an open question as to whether innovations such as DTHs can advance recursivity in the more challenging context of disciplined parties in a parliamentary system.

We ask: Under favorable conditions, can recursive representation be advanced in a parliamentary system through the practice of DTHs? Here, favorable conditions are those in which party discipline is temporarily relaxed – a minimum threshold to prove Directly Representative Democracy can apply before moving to more challenging conditions involving party leaders and policy development.

Answering this question requires theorization because no one has worked out the operational criteria for Directly Representative Democracy in non-USA cases. Thus, before diving into our empirical case, we first develop the theoretical connections between recursive representation and Directly Representative Democracy. Elaborating on these connections allows us to step back from the USA case and explore the conditions required for recursive representation in practice. We do so through an in-depth study of the two first-ever DTHs held in a parliamentary political system (Australia). In these DTHs, the participating Member of Parliament (MP), Hon. Dr Andrew Leigh MP, met with constituents to discuss pending legislation concerning a reproductive technology: mitochondrial donation. ¹

Our empirical analysis relies primarily on the qualitative analysis of interviews conducted with participants in the DTHs, as we are interested in understanding whether and how constituents experienced these forums, and the increased connections with their MP (the most basic criterion for DTHs ‘working’). Qualitative interviews are particularly well suited for studying the democratic innovations, such as the DTHs (Talpin, 2019). Our methodological approach is interpretive and abductive, moving iteratively between theoretical ideas (about recursive representation) and empirical insights (how participants view and experience DTHs). An abductive approach is neither entirely theory-driven (deductive) nor entirely practice-driven (inductive). It begins with an observation of ‘real-life’ events and draws on the pre-existing theories or concepts to make sense of them (Hendriks et al., 2020: 32).

Drawing on interviews with thirty-five participants in the DTHs, we argue that DTHs enabled a highly effective, two-way communication channel between constituents and representatives about the issue at stake. Our analysis shows that the ability of DTHs to advance recursive representation depends on their deliberative quality. We elaborate on the key criteria required for deliberative quality drawing on the interviews we conducted with the participants of the two town halls. More broadly, our proof of concept establishes that the idea of Directly Representative Democracy, and the DTHs, can successfully travel into political systems different from the USA – which offers a promising prospect for deliberative reform in other representative democracies.

Reforming representation through deliberation

Minimalist accounts of political representation posit that democratic legitimacy and accountability emerge from procedures like universal suffrage and free and fair elections, along with freedom of press, access to information, and maximum terms of office. Formal voting power, in this view, provides citizens with the legal authority to elect representatives. Between elections, citizens fade into the legislative background (Rehfeld, 2006; Urbinati, 2011).

Historically, debates about the scope of political representation have oscillated between trustee and delegate interpretations of the legal and ethical purview of elected officials: elected representatives act more or less on orders as delegates or as trustees exercising their own judgment about promoting the interests of their constituents. A nuanced account of how public officials work recognizes that elected representatives balance multiple goals: working (and seeking re-election) in their own constituencies, promoting their own policies, and also coordinating as members of political parties seeking to implement programs. This means that elected officials neither should rely only on their own personal judgments, nor take strict orders from the public or party leaders. Moreover, the balance and dynamics of such considerations are constrained by the system in question. For example, elected officials in strong party systems (like our case here, Australia) simply must, unless they are independents, give more weight to party positions than to their own policy positions.

The emerging crisis of representative democracy across the globe makes it difficult to uncritically accept the standard account. Therefore, we shift the emphasis to consider deliberative consultation as a type of non-electoral interaction between citizens and representatives. While reformers increasingly experiment with innovative deliberative practices within the organization of political parties (Junius et al., 2023), we are just beginning to understand the precise mechanisms and norms that might guide a more inclusive and deliberative representative-constituent relationship. So here the focus turns to the understudied question of how constituents might channel input through their elected officials to party leaders (and legislatures as a whole) between elections (Leston-Bandeira, 2016). More specifically, we study how constituents regard a particular democratic innovation, DTHs, in the context of a strong party system.

On a recursive account, quality representation requires ongoing, dynamic interaction, rather than relatively infrequent elections warranting a fixed mandate until the next one. Some scholars suggest that such recursivity requires alternating between politicians and citizens for the authorization and legitimation of inputs and outputs (Sørensen, 2020). Others emphasize recursive communication, where representatives and constituents take turns listening to and responding to what the other is saying (Mansbridge, 2019).

Some democratic innovations have sought to create conditions that strengthen the connection between citizens and representatives. These take the form of citizens’ assemblies (Farrell and Suiter, 2019), deliberative minipublics adjacent to legislative committees (Hendriks, 2016), and a permanent citizens’ assembly to parallel the parliament (Caluwaerts and Reuchamps, 2018). Other proposals seek to ‘bring the citizen back in’ to the formal arenas of decision-making as part of participatory policymaking and interactive governance (Michels, 2011), co-governance (Geissel, 2012), collaborative governance (Torfing and Ansell, 2017), and plebiscitary and cooperative governance (Hendriks, 2019). However, empirical research on how democratic innovations specifically enable recursive representation by improving the representative-constituency relationship remains quite limited (cf. Hendriks et al., 2020; Prior, 2018; Hendriks and Dzur, 2015).

We focus on the application of recursive representation in the model developed by Neblo et al. (2018). This model seeks to hone the relationship between a constituent and their elected official, notably by soliciting input from constituents on issues and through accountability via explanation to constituents of the representative’s actions. These processes foster legitimacy and trust (ibid, 14–16; 28–29), cutting across the traditional trustee-delegate dichotomy.

Recursive representation alters the trustee-delegate model in three important ways. First, it disaggregates who has primacy vis-à-vis ends and means: if we think that constituents are communicating well-informed, individually considered, and publicly defensible opinions, then we move toward a stronger warrant for more of a delegate model with respect to the value trade-offs and goals (the ‘ends’ as part of ‘consultation’) but we retain more of trustee relationships with respect to the particular policy instruments (the ‘means’ as part of ‘accountability’) to achieve those ends. Second, consultation and accountability come with rationales rather than just choices. This contrasts the trustee-delegate dichotomy, which only countenances what to do rather than why we should do it. Finally, recursive representation rejects the single direction of influence posited in the trustee-delegate dichotomy.

To generate recursive dialogue between government officials and constituents, Neblo et al. (2018) revised the traditional town hall to produce a ‘Deliberative Town Hall’. In collaboration with dozens of Members of Congress, they have hosted online DTHs with randomly selected, statistically representative samples of constituents, discussing controversial issues in the USA such as immigration, the future of work, and Covid-19 policies. Randomized controlled trials assigned participants to different groups (treatment, information only, and true control) to identify the causal effects between design choices and participant engagement (Neblo and Wallace, 2021; Minozzi et al., 2015; Esterling et al., 2011; Neblo et al., 2010).

Findings from these studies suggest that legislators’ engagement in substantive communication with their constituents is linked to citizens’ motivation to be informed about politics. Participants showed a capacity to become informed and gained knowledge by increasing attention to policy outside the context of the experiment, irrespective of prior knowledge. Moreover, they find that even participating in a single DTH increases general political trust (Minozzi et al., 2015: 110).

In strong party systems, DTHs can still create an entry point for citizens into the legislative dynamic. Parliaments might suspend party discipline to enable ‘conscience votes’ on legislation, as we outline in the next section, allowing MPs to use citizen-led decision-making processes to influence their votes in parliamentary debates.

Recursive representation in a strong party system

Here we present the first DTHs convened in a strong party system in a project called Connecting to Parliament (CTP). Like the USA, the Australian system has a direct constituency component (unlike many proportional representation systems) that facilitates isolating potential reasons for any divergent results from the success observed in the weak party system of the USA. Australia also has a history of extensive deliberative innovations, mostly in the form of citizens’ juries in state and local government. One noteworthy exception at the federal level was the pioneering Australian Citizens’ Parliament, which was held in 2009 and brought together 150 randomly selected citizens to discuss Australian democracy and how it can be improved (Carson et al., 2013).

Background: key design features and the issue

We designed and hosted the DTHs in three stages. First, the MP identified the issue, which was the legalization of mitochondrial donation in Australia. In mid−2020, party leaders announced they would put the legalization of mitochondrial donation to a conscience vote.

The MP took this opportunity to engage his constituency on the issue through a deliberative democratic process. Next, the research team designed two DTHs as engagement interventions, which would also generate data allowing us to assess the research questions set for the project. Recruitment is the third and final stage ahead of the town hall events. We recruited for the DTHs by mailing invitations to every household in the constituency. Residents interested in participating were directed to register online and were then assigned to one of two forums (Appendix A). The MP’s invitation indicated that the informed opinion emerging out of the DTHs would influence his vote.

Based on Neblo et al. (2018), key design features specific to DTHs include:

1. 1. Representation by assigning a stratified, randomized sample of constituents to each town hall.

2. 2. Informed discussions by providing participants with background material and information on the issue at stake. In our case, we provided the selected participants with an 8-minute video produced by the Australian Government National Health and Medical Research Council (NHMRC), and the associated Issues Paper which described the ethical, legal, and scientific dimensions of mitochondrial donation (see NHMRC, 2020; 2019).

3. 3. Facilitated interactions by recruiting an independent facilitator, not a part of the research team or the MP’s staff, who ensured that norms of mutual respect, equal turn-taking, and reciprocal reason-giving were observed.

4. 4. Impact of citizen deliberation in this DTH was achieved in a guarantee by the MP to be guided in his vote in Parliament by the process.

Party leaders loosening control for a conscience vote was important for this experiment. In Australia, the fusion of executive and legislative powers allows the majority party (or coalition) almost always to determine the timing and content of legislation. Party discipline makes it difficult for MPs to vote against policies of their party in Parliament, and the domination of the legislature by political parties enables the government to use the Parliament’s rules and procedures to its own advantage through gag motions, guillotines, and bypassing or condensing committee work (Summers, 2014: 34, 36–38).

MPs representing major parties are rarely granted the opportunity to vote on motions or policies according to their own conscience. Conscience votes are mostly used on moral and ethical issues such as gender discrimination, euthanasia, access to abortion, embryo research, and same-sex marriage (Ross et al., 2009; Balint and Moir, 2013; Plumb, 2015; Lausberg, 2016). Such votes occur, on average, about once every electoral term. We chose the ‘favorable condition’ of a conscience vote for the present study to begin with a proof of concept.

During each DTH, a facilitator started the conversation by introducing CTP and its purpose and then the procedure and agenda of the town hall. Explanation of the procedure helped establish what makes these town halls deliberative (in the terms we introduced at the outset of this paper, stressing mutual communication and reflection about matters of common concern). Deliberation was promoted by the provision of nonpartizan, in-depth background information about the issue; diversity and representativeness of the participants; turn-taking in asking questions, respectful expression of agreements and disagreements; and respectful listening. One of the authors presented a brief about the policy issue, followed by opening remarks from the MP who reiterated that discussions from the town halls would heavily influence his vote. For the online DTH, questions were submitted via the platform and then read by the facilitator. For the in-person DTH, the facilitator invited participants to come to a microphone to ask their questions.

The design of CTP differed from previous DTHs in two ways. First, the MP explicitly stated on the invitations that the process will inform his conscience vote. In previous DTHs, legislators have made similar, though typically less emphatic, statements in the forum, rather than in the invitation. This commitment provided constituents with a heightened incentive to engage in the process. Secondly, the Australian DTHs were hosted in-person and online, both held in September 2020. Other than the mode, their format was kept as similar as possible. Both DTHs were 90 minutes long and moderated by a facilitator where participants raised their concerns and asked questions to the MP about mitochondrial donation law. One notable difference between the two was that the participants of the in-person DTH conversed with one another at the end of the meeting, while the online DTH was solely a conversation between the MP and individual participants.

Before the conscience vote, mitochondrial donation was illegal in Australia under two laws: the Research Involving Human Embryos Act 2002 and the Prohibition of Human Cloning for Reproduction Act 2002. Both prohibited the creation of embryos with genetic material from more than two people and heritable alterations to the genome of human embryos for reproductive purposes. Mitochondrial DNA disease is an inherited condition with an extremely high fatality rate, where many children do not live to their fifth birthday. The only potential to eliminate the risk of inheriting mitochondrial disease is through a new assisted reproductive technology, mitochondrial donation. Yet the technology directly contradicts the two Acts as it involves creating an embryo using the nuclear DNA from both parents, plus mitochondrial DNA donated by another woman. A potential mother who carries mitochondrial DNA mutations, then, can have a genetically related child with almost no risk of experiencing the disease (NHMRC, 2020).

And thus, these DTHs addressed a weighty topic. Parliamentary attempts for reform began with a Senate inquiry in June 2018, which was reported in February 2019. The Federal Government followed the report with submissions from experts and the public on key questions concerning the disease and the reproductive technology (NHMRC, 2020). An expert working committee reported to the NHMRC in 2020, and the NHMRC hosted a call for public submissions, webinars, and a citizens’ panel. Both consultations illuminated the diversity of ethical and social considerations should mitochondrial donation be introduced into Australian clinical practice.

In March 2021, months after our DTHs had concluded, the Mitochondrial Donation Law Reform (Maeve’s Law) Bill 2021 was introduced to Parliament. The conscience vote took place in the House of Representatives on December 1, 2021, and the Bill passed with 92 votes in favor and 29 against. The Australian Senate passed the Bill on March 30, 2022, with 37 Senators voting in favor, and 17 against, becoming law on April 1, 2022. The participating MP delivered a speech in Parliament explaining the outcomes of the DTHs and how they informed his vote to legalize mitochondrial donation. ²

When do deliberative town halls advance recursive representation?

In order to identify the conditions of whether and to what extent the DTHs can advance a recursive form of representation, we conducted interviews with the participants. Within the two weeks following the town halls, two of the authors interviewed 35 participants out of 49 across the two DTHs. Interviews were semi-structured and lasted on average for 20 minutes (yielding 237 pages of transcripts).

Two of the authors conducted the interviews and initially coded the transcripts through abductive thematic analysis to identify the kind of deliberative norms that advance recursive representation. A third co-author, who did not conduct the interviews, then coded independently. As explained in the introduction, as we adopt abductive inquiry, our analysis process is non-linear. Rather, our case-based research relies on an integration of our theoretical knowledge to examine political and social phenomena, what Dubois and Gadde (2002) call systematic combining. Abduction connects findings observed in the empirical data with themes and concepts arising from literature in the field (Schwartz-Shea and Yanow, 2011).

The interview questions were designed to gather data to speak to the uniqueness of deliberative engagement compared to other modes of communication with an MP. In addition, the questions sought to elicit the perceptions and experiences of participants themselves, especially concerning what ‘connection’ in representative democracy means and how it can be realized. In turn, our interpretations of the interviews code these meanings to nuance the normative criteria for recursive representation (Ercan et al., 2017). ³

The synthesis of our interpretations aligned on three core thematic areas portraying the conditions under which DTHs can advance recursive representation: authenticity, inclusivity, and impact. These conditions resonate with the normative framework of ‘deliberative capacity’ developed by Dryzek (2010), as well as with the key normative standards for DTHs suggested by Neblo et al. (2018) (inclusion, informed justification, good reason-giving, promoting legitimacy, and scalability). Dryzek (2010) argues that these standards are likely to be met when a town hall recruits a broad audience and adopts rules and technology encouraging equal participation, focuses on a single timely and relevant issue, provides balanced background information in advance, selects a neutral moderator, and focuses on unscripted, authentic interaction (29–31, 133).

Both Neblo et al. (2018: 85) and Mansbridge (2019) recognize the need to make citizens’ perspectives more visible, yet we know little about the content and the quality of communication between representatives and constituents in direct encounters (Karpowitz and Raphael, 2014). The semi-structured interviews allowed us to engage with the participating citizens’ perspectives thoroughly and openly. We asked follow-up questions if we thought a position was implied in what a participant said, and sought clarifications where needed (Appendix B). For example, when participants note other ways they engage with their MP such as by writing letters to their MP, we would ask for clarification on how participating in the DTH differed.

One might wonder, however, how citizens’ perceptions of legitimacy can speak to legitimacy per se. Prominent theories (Habermas 1984; Walzer, 1993) suggest that one cannot have actual legitimacy without the public being able to eventually (and under sufficiently favorable conditions) perceive and assent to that legitimacy (or at least we cannot confidently infer actual legitimacy). Observers and social critics can offer up arguments and interpretations, but the ultimate test can only be assessed from the participants’ perspective. Thus, perceived legitimacy is a necessary condition for inferring actual legitimacy.

Most of our interviewees highlighted the disconnection between the public and politicians, which they clearly observe in key policy issues. Citizens explained that what looks like political apathy and disengagement is a manifestation of the persistent disconnection between politicians and ordinary citizens. In other words, as politicians continue to ignore the problems that matter to them, the citizenry falls away from the usual (and limited) channels of engagement: ‘[…] Everyone wants some change and action [on climate change], but that’s being wholly ignored by the current Parliament. I think that’s concerning. I think that’s part of why people are becoming more apathetic and disengaged from politics’ (In-person participant 8).

For many, the disconnection between citizens and their representatives in Australia reflects problems within the Westminster parliamentary model which encourages combative rather than cooperative politics. Television coverage of Parliamentary discussion tends to focus on Question Time – the most combative hour of sitting days. Several participants felt that politicians are more dedicated to political point scoring and ‘bending the truth’ to help secure re-election rather than working for the public good. Participants gave this as one reason for the lack of trust in elected officials.

You’re going to have parties which is part of politics, but so far all I see over the last number of years is one side attacking one side continually, and nobody’s prepared to work together for what’s good for Australia (Online participant 16).

DTHs can mitigate the shortcomings of partisan politics because they are explicitly designed to induce constructive communication between citizens and elected officials (Neblo et al., 2018: 43). Yet the ability of these DTHs to enable recursive representation and bridge the growing disconnection between MPs and their constituencies depends on their deliberative capacity to embody authentic, inclusive, and impactful deliberation. Here we refine the significance of these desiderata in relation to participant insights.

The requirement of authenticity

Based on our analysis, the first norm that the DTHs should meet to enable recursive representation is authenticity. In the context of DTHs with elected representatives, authenticity involves communicating reasons that derive from the representative’s own thinking and are not driven merely by pandering to an audience or party discipline. Our interviews with the participants of DTHs suggest that authenticity in practice manifests itself through the following attributes:

1. i) Honesty and humility of the elected representative

Honesty along with non-coercive and non-manipulative communication means that representatives engage in dialogue relatively unbound by party discipline. In this respect, constituents ought to expect representatives to communicate their personal understandings, thoughts, and reflections on a given issue rather than a ‘script’ sanctioned by the party, as well as admit to their own limitations (though in some cases, the MP’s position and reasoning may track the party’s relatively well).

Some interviewees expressed a concern that party discipline and politics suppressed the voices of individual representatives and hence undermined the representatives’ connections with their constituents. Participating in the DTH allowed citizens to know their representatives better. The ‘undisciplined’ views of the representatives played an important role in connecting with the community they represent. As one participant put it, ‘I do not want to hear the party line, I want to hear what the person is like’ (Online participant 5).

Acting more authentically and making citizens feel they are connecting with a member of their community may have a performative aspect. Yet it is far easier to perform this role if it reflects the MP’s personality. In this case, many interviewees perceived the MP to be authentic: the MP answered from his own perspective and did not indicate or follow any party line (Online participant 4). Also in the case of this MP, his political image is of an ‘academic’ – balanced and trustworthy (Online participant 12). Throughout the town halls, citizens seemed to ‘quiz’ their representative to test whether he could uphold an authentic conversation without ‘referring to notes’, as well as maintain his political image (Online participant 3). Although many representatives cannot draw on an academic persona similar to this particular MP, other trustworthy personas are available, and can be adapted to the MP’s biography and presentation of self.

Citizens knew, however, that there were limited opportunities for their representatives to connect with them on issues subject to parliamentary debate. Because the MP did not take an initial stance, it encouraged participants to hear from and learn about their local politician on the subject, which would create more trust in the conscience vote (Online participant 5). Trust specifically arose from thinking through answers and communicating reasons for his responses, which is the opposite of merely holding strongly polarized views based on personal beliefs or assumptions (In-person participant 11).

Moreover, the democratic innovation allowed citizens to experience deliberation themselves and have a reference point to how decisions are made in Parliament:

It [the DTH] gives the people who engage with it a better idea of what it is that politics is actually about. That politics is not something that goes on over there. It is something that is how they behave and how their attitudes work out in their own community. So, one of the things I noticed there was that of the people, when we had the table discussion, were not aware how the deliberation process in Parliament goes, particularly the committee stage, so that Parliament is actually itself an open deliberative process is not something that comes across to people, I think. So, anything that improves that is a really good thing (In-person participant 3).

This put citizens in a position to contrast how they go about their own deliberations with what they observe in parliamentary deliberations. Specifically, we heard many concerns about the state of insincerity and ‘circus’-like atmosphere in question time in Parliament (Online participant 7). Party discipline limits the range of considerations that can be presented and discussed (Kam, 2009). As a result, members of the public view question time as combative and inauthentic because it comes across as being about ‘point scoring’ rather than collaboration to get things done (Online participant 3).

While the Australian Parliament is perceived as the main arena for combative party politics, other spaces for citizen engagement also seem overly partisan and combative. In rallies, for example, elected representatives often propagate their parties’ platform. By contrast, the design of an authentic space, like these DTHs, allows citizens and representatives to communicate beyond the binary of support/oppose in a rally and focus more on the questions and discussion (Online participant 2).

1. ii) Reciprocity

For DTHs to advance recursive representation, it is crucial that the representative and the constituents reciprocate reasons, arguments, information, and considerations on the issues. This feature is closely related to the requirement of communication unconstrained by party discipline. Reciprocity of reasons and considerations allows constituents to communicate their motives for participating and affective aspects to be considered when deciding on contentious issues. Such reciprocity is at the core of the deliberative democratic account of justification.

Participants commented on how deliberation surfaced and clarified those motives and affective considerations. ‘I suspect that I possibly would not have been involved if it had been some other fairly vanilla-type subject, but it is certainly one that I have got reasonably strong views on’ (Online participant 5). The topic had moral implications, which brought out diverse feedback and questions on the matter (Online participant 6).

Many participants also thought it important that the constituents also shared the reasons for their views and preferences, even if they agreed with each other. ‘We were a pretty homogeneous group […] There weren’t strong opinions on either end of the spectrum. Most people were generally for it…. But we still learned things from each other’ (In-person participant 11). People knew different things or had different questions that made them think more broadly than if they were simply being asked to respond to a survey.

1. iii) Active Listening

Our analysis reveals that active listening is a key feature of authenticity in the context of DTHs. Elected representatives must demonstrate that they have been listening and reflecting. Research indicates that people, especially in direct interactions, are often good at detecting insincerity and attempts at manipulation (Goupil et al., 2021).

Demonstrating active listening can take different forms (e.g., admitting limited knowledge when responding to some questions or concerns). In our case, this admission was received positively by those who mentioned it (Online participant 7). Relatedly, the MP taking in questions and replying with well-thought-out responses that directly referred to how participants put them indicated active listening (Online participant 2). With active listening, representatives become aware of viewpoints they otherwise would not know and would not be able to represent. One participant reflected on the MP saying, ‘well, that’s a perspective I had not thought of, or that is a good question’ […] ‘I would say this indicates to me that people are coming up with more potential implications than he has already thought of’ (Online participant 4).

The requirement of inclusivity

The DTHs organized as part of CTP consisted of residents of a constituency, their elected representative, and a moderator. Inclusion as a deliberative ideal can refer to the demographic characteristics of those present, with the proviso that all citizens in the room have an equal chance to give voice and be listened to (Young, 2002). But inclusion can also refer to relevant knowledge and relevant voices. When it comes to relevant knowledge, we should not automatically take that to mean that experts should play a greater role. As Sprain and Reinig (2018) point out, expertise can create ‘hierarchies that foreclose participation’ in deliberative forums, with expert information being seen as yielding obvious solutions.

Our participants did not perceive any shortcomings in terms of the demographic characteristics of those present, or the degree to which all had a chance to give voice and be listened to. However, concerning relevant knowledge, participants felt there were issues only medical experts specializing in mitochondrial donation can respond to. Many participants expressed concerns about the absence of experts in the room (not to participate in the same way as the citizens, but to provide information). ‘Inclusion’ accordingly should also be understood as applying to access to reliable and relevant technical information and points of view from experts on issues that feature prominent technical facets. Some participants regarded their representative as a legal expert for his experience in navigating legislation and using his experience to guide participants in how policies are created and how to translate their ideas into legislation (In-person participant 9).

While the MP acted as an expert, it was still not sufficient for some constituents. As one participant explains:

… [the MP] said that it was through that forum that he was going to make his decision. I think that concerned me because it seemed to be a collection of people who were asking questions and there weren’t really any experts that knew the answers. I was just a little bit concerned at the approach. I think it’s really, really good to have everyone there and to get community engagement, involvement. I think that the concept is good, but I was just a bit concerned that there were a lot of questions asked and some other people were answering those questions, but whether it was from a very well-understood background, I just wasn’t sure. It just felt like a collection of people who were passionate and interested, but there didn’t seem to be a lot of, I guess, science or information to support those discussions or those decisions … I came home and I was just like, ‘Oh, if that’s what they’re generating their decisions on, I’m a little bit concerned because it was a lot of uninformed but very caring and thoughtful people in the room.’ (In-person participant 8)

Thus, for a subject like mitochondrial donation, the forums could have included different expert perspectives beyond the background materials (e.g., experts from the medical, legal, and ethical fields along with the broader community). Participants suggested that this would help balance the information shared in the town hall so that participants could make more informed decisions (In-person participant 9; Online participant 12).

Participants also noted the absent voices of people with experiences of mitochondrial disease. This resonates with common practice in citizen deliberation to over-sample populations directly affected by the issue at hand. Including affected people has a democratic-epistemic virtue. Reasons and considerations from those affected by the disease and whose lives will be impacted by the legislation should be included in making this legislation. While some participants were familiar with the lived experiences of those with mitochondrial disease, those who were not emphasized the need to include affected people:

I had a friend who died from mitochondrial myopathy a few months ago, so I’m pretty aware of what that involves or what happens there and how serious this whole thing is. And there was no one there who really knew an awful lot about it, it struck me. There were people there who knew about it. Had a lady from the audience from a meeting, gave a pretty good description of how mitochondria works, and that sort of thing. I thought it would have been more helpful to have somebody there who knew about the subject, who knew more about what that was. Because I think the people that went there were people who, either have some experience of it, or know of it, or know of people who have, or family members who’ve had it (In-person participant 1).

In this way, participants were a bit concerned that some voices, particularly the people who had experienced the disease or were at risk of their children having it, were not present in the forum. From an empathetic perspective, the issue is that the real-world implications this forum may have would be based just on the views of those who were present (In-person participant 10).

These insights on inclusion were enabled by some design shortcomings. Yet, citizens have also articulated a nuanced understanding of ‘experts’. It was not just the lack of experts and affected people (our design shortcoming) that participants critiqued. Participants reflected on the democratic and epistemic qualities of inclusion in which different types of experts are needed, to broaden the range of considerations in making decisions on contentious issues. ⁴

The requirement of (potential) impact

Finally, recursive representation requires the DTHs to have at least the potential for impact on formal decision-making processes (even if it may not be decisive in any given case). The power of the process and hence, the significance of participating in it matter greatly. This differentiates democratic innovations strengthening recursive representation from consultations that engage constituents without committing to taking their input into decision-making.

The perceived need to consult the electorate on the most controversial issues was prevalent among our participants (In-person participant 4). Specific emphasis was placed on the need for broader public consultation on moral issues because they require more than what a party line or single representative personally thinks (Online participant 13).

Some participants extended this thinking beyond conscience votes to moral issues more generally. Their reasons include that citizens need to be involved in these decisions and educated about the issues (Online participant 12), for the sake of diversity of questions and feedback (Online participant 6), and ultimately, because citizens are disconcerted about the disconnection between them and their representatives:

…. on contentious issues like this one, like abortion law, or gay rights…, all the contentious ones, there should be some sort of consultation with the electorate. It somehow has to be put in politicians’ heads that they represent an electorate, not to run the country the way they’d like to have it (Online participant 7).

Standard town halls and consultations are not designed to allow for this type of influence. Even with the outcome of the conscience vote on mitochondrial donation being contingent on how the parliament votes as a whole, a chain of accountability and legitimacy on that vote is still established between the constituents and their representatives in a consequential manner.

Many participants were motivated to participate because their representative in Parliament sent a formal invitation indicating that constituents could determine the direction of his conscience vote. Because Australian democracy usually lacks similar opportunities, some constituents considered this invitation a ‘privilege’ to be able to influence their local member, hear what other people had to say, and learn about how citizens can get more decision-making ‘control’ (Online participant 4; Online participant 2). Yet because DTHs have a potential contribution to recursive representation, citizens’ perception of the impact their participation could make also relies on the MP’s commitment to authentic communication, active listening, and reflection. In short, promised ‘impact’ alone does not suffice. At least in CTP, the features of authenticity and impact are interdependent.

This opportunity to influence parliamentary decision-making revealed that constituents wanted to be more involved in their community and in politics. From our interviews, constituents made the point that the public does not want to be deeply engaged on every single issue. They expect that representatives should be able to make decisions independently. Nonetheless, a conscience vote reflected something different in the eyes of participants:

[W]hen it comes to a conscience vote, it’s up for grabs. And I think then it’s important that the Member of Parliament knows what the people that they’re represented … The people they’re representing are thinking to help guide their own thoughts. Because I don’t think that with the conscience vote, they should just be going with their own conscience. There should be an element of representing the community as well (Online participant 3).

In other words, because a conscience vote has less to do with the party line, it means that the vote should be more directly informed by the considered views of the public.

Participation in this process was particularly appreciated because of its impact: CTP town halls are more than ‘just talk’. Some participants expressed their appreciation for interacting with and being listened to by their MP. One participant noted:

I really appreciated the opportunity to be involved and have a say…the ability to ask a question and it was more personal than following the local member on social media. You kind of react to it. You get to see what they are doing. You kind of comment on that issue. Whereas this town hall meeting before they are about to do something, before they are about to vote on proposed legislation. So, it was really good to kind of have input before they kind of do the thing…It made me feel like a valued member of the community, especially since I had just moved here a couple of months ago (Online participant 6).

It is worth recognizing the particular context in which the intervention was conducted that may have facilitated impact. As we have noted, the electorate was relatively highly educated, compared with other electorates in Australia. Because the electorate was in the nation’s capital, voters were more likely to have worked in the public service than might have been the case in other parts of Australia. Moreover, our intervention was on a conscience vote, in which the MP can vote according to the views of those who participated in the DTHs. News reporting was moderate, and not strongly skewed in one direction or another. The structure of the bill voting, either up or down, is also a factor - it was not an amendable proposal. In this sense, the issue was perhaps more straightforward than many other topics that come before Parliament.

Overall: the participating constituents regarded the DTHs as a ‘new’ participatory process enabling them to connect with their representative in a different way. This perception raises a question about the types of mainstream political engagement, such as emailing, telephoning, or speaking in person with their representative at a street stall, and the engagement gaps between citizens and their representatives in public office which do not create connections as directly as a representative democracy would. Even for those who had participated in consultative processes before, this was their first time directly engaging with their representative and an altogether different type of experience. The presence of their representative in the room, even online, was more significant than the more mainstream and routinized forms of engagement (e.g., responding on social media or to correspondence that tends to come from staffers on behalf of the representative) (In-person participant 4).

Conclusion

We have sought to assess whether, under favorable conditions, directly representative deliberative innovations can work well in a strong party, parliamentary system like Australia’s. Our findings suggest that the DTHs can indeed function well in such systems and enable recursive representation, but that claim comes with some prompts for further research.

Inclusion is another caveat, especially in exploring the question of who should be included in DTHs. Town Halls that are oriented more toward deliberation in advance of decision-making may often require over-sampling of affected people to inform decisions reflecting their lived experience (in this case, those suffering from genetic disease). Diverse expert knowledge and perspectives also need to be included, in this example: medical, legal, and ethical. ⁵ Depending on the issues at hand, inviting experts to DTHs where elected representatives meet their constituencies, can potentially strengthen rather than undermine the connection between MPs and citizens. Naturally, this could have changed the mood of the conversation, since a ‘clash of experts’ might have created a more tense environment. Future research on DTH variants incorporating one or both features should assess whether doing so alleviates the inclusion deficits perceived by participants in this forum.

Crucial for the role of representatives in DTHs to advance recursive representation is to decenter their style of political communication from persuasion to listening. In our case, the process was less about a platform for the MP to persuade his constituency and more about an opportunity for participants to express their views and ask questions about the issue. Participants valued active listening in this communicative interaction (discussed in our analysis of ‘authentic engagement’). In fact, participants viewed the MP’s willingness to listen to a variety of viewpoints and acknowledgment of his limited knowledge on the subject were his key strengths in demonstrating his genuineness to participants. This feature of CTP, the MP not indicating a stance, changes the dynamic of what counts or how to evaluate directly representative democracy as a deliberative process. We further suggest that authenticity is part and parcel of CTP. Because political parties were not implicated in the discussion, the process was not about partisan persuasion.

Finally, the forum was not only a space for constituents to merely voice their concerns but for their contributions to have a direct influence on the MP’s conscience vote. The degree of anticipated ‘impact’ on the deliberators’ part, then, is a novel design feature. We find that the prospect of influencing the political outcome by constituents augmented the connection and motivation to participate in this process with the MP. Future research should analyze both the normative conditions under which such influence might be warranted, and the empirical effects that foreseeable impact has on motivation at the individual level as well as the group level deliberative dynamics.

More important than the effect on the motivations of individuals is impact on the deliberative qualities of the political system. In a parliamentary system no less than elsewhere, DTHs can provide a powerful way to enhance meaningful connections in the deliberative system, linking representatives and their voters in new deliberative ways.

Beyond conscience votes, we can envision several other entry points for DTHs in a strong party system: private member bills, leadership authorizing backbenchers to convene DTHs as ‘advance sensors’ on emerging issues, local constituency-based issues, party leaders convening national DTHs on issues in their portfolio. These options would insert DTHs at different stages of the legislative process, in the development of options, as opposed to our conscience vote case which was directly prior to the parliamentary vote. It may also be possible to insert DTHs ‘downstream’ from policymaking, in the delivery of public services and implementation of policy. At the conclusion of our mitochondrial donation case, the health directorate responsible for implementing any new law at the state level indicated that the DTH process would be used for engaging the local population. The substantive focus of the DTH would change in these different options, but whether this would have any effect on the quality of the interactions, and their impact on participants and recursive representation more generally, would require further experimentation and research.

Overall, we have shown the relevance of deliberative ideals to improve practices of democratic innovations that address the growing disconnect between representatives and their constituents. Our study emphasizes the prospects of DTHs for enabling recursive representation with specific attention to the impact that design can have on how DTHs are experienced by the public. Michael Saward (2021) suggests that design is about providing plans or ideas for a purpose, in a particular context. This entails a co-evolving relationship between problems and potential solutions. As a result, there is no such thing as a final design, but rather a commitment to create and re-design processes that provoke social and political problem-solving. At the same time, it is unwarranted to accept democratic minimalism considering the success of experiments in directly representative innovations. The crises of representative democracy require multi-faceted responses and we have demonstrated the contribution of one such response anchored in recursive representative-constituency relationships.