Introduction

Interest in and awareness of deliberative mini-publics (DMPs) is expanding (OECD 2020a). These processes combine the random selection of citizens with deliberation that usually results in the production of policy recommendations to present to decision makers. These design features lead advocates of DMPs to make some specific democratic claims about them. First, they aim to bring together ‘representative subset of the wider population’ (Curato et al. 2021, p. 3) through the random stratified selection of participants. Advocates claim that this gives DMPs a particular claim to inclusiveness, representativeness and legitimacy (Fishkin 2018)—though all these claims are contested in the literature (e.g. Chambers 2009; Lafont 2019; Steel et al. 2020). Second, DMPs involve carefully structured and designed deliberations—informed discussion typically defined by diverse viewpoints, mutual respect and exchange of arguments. The suggestion is that deliberation in this format can produce epistemically robust outcomes, most often in the form of informed and reasoned input into existing decision-making processes through recommendations or guiding principles (Curato et al. 2021).

Deliberative mini-public practitioners and researchers have developed well defined and often highly structured process designs and models for DMPs, such as Citizens’ Juries or Citizens’ Assemblies (see e.g. Curato et al. 2021; Ryan and Smith 2014). They are often (but not always) convened or commissioned by public authorities (Caluwaerts and Reauchamps 2015) such as local councils, state or national governments, specific government departments or legislative bodies. Mini-publics are only one form of deliberative practice, and the practice and theory of deliberative democracy extends beyond this specificity. Yet, DMPs are certainly having a moment. High profile, national level DMPs have taken place in several Organisation for Economic Co-operation and Development (OECD) countries and in some cases have resulted in notable legislative changes. This increased public and political profile warrants increased scrutiny of their ethics and governance.

The design and delivery of the process itself is often delegated to professional deliberative engagement specialists (hereafter ‘practitioners’), defined as ‘an individual working in the public or private sector who is paid to design, implement, and/or facilitate participatory forums’ (Bherer et al. 2017a, b, c. p. 1). In this article, I explore how practitioners identify and attempt to overcome the ethical issues that they face in deliberative practice through a qualitative thematic analysis of interviews with 25 practitioners. Practitioners sit at a critical intersection between the demands of commissioner–clients, and the democratic and emancipatory goals of deliberative democracy (Lee 2014). Existing research suggests that being engaged in a commercial relationship with the commissioner as their client puts practitioners in an awkward position (Lee 2017). They face possible loss of income and future work if they do not meet certain demands (Bherer et al. 2017b). At the same time, their professional identity is ‘closely associated with the realisation of positive democratic values’ (Bherer et al. 2017a, b, c, p. 7), meaning that they must balance meeting the expectations of the citizen participants, commissioning body, and the normative ideals of the process itself.

Whilst previous research has shed light on how practitioners perceive themselves and their role (Cooper and Smith 2012; Friess and Herff 2023), less is known about how they navigate these tensions on the ground. This is an important line of enquiry because the deliberative mini-publics community of practice seems to operate in a way that runs counter to the existing evidence on the commercialisation and professionalisation of public participation. This literature highlights the tension between commercial and democratic values (Bherer et al. 2017a, b, c; Christensen and Grant 2020; Lee 2014), and the tendency of commercialisation to close down collaboration and reflexivity (Chilvers 2013, 2017). But anecdotal experience of the deliberative mini-publics community of practice suggests otherwise; that practitioners and scholars are highly reflexive, characterised by frequent learning calls, conferences, briefs and reports discussing perceived failures and lessons learned. This led Hendriks and Carson (2008) to define deliberative democrats as a community of practice rather than an industry. Our findings illustrate the capacity of deliberative practitioners to engage in reflexivity, which is especially important as interest in DMPs grows: as greater resources and political interests are devoted to them they face greater risks of being compromised.

The paper proceeds as follows. I first review the literature on the professionalisation and commercialisation of public participation, and outline the development of the deliberative mini-publics community. I then briefly outline the methods taken in this research. I go on to identify three ethical pressure points that practitioners encounter: pressure to steer the direction of a DMP; pressure to achieve impact and realise transformation; and pressure from operating in a commercialised context. Following this I outline the practices that interviewees utilise to try and address ethical issues. Practitioners employ a range of tactics in response to ethical issues, many of which rely on their experience and judgement, and some of which are more formally codified. However, not all challenges are surmountable, and interviewees were not naïve to the limitations of their power integrating DMPs within existing power and governance structures (Boswell et al. 2022). These findings demonstrate the reflexivity of practitioners and advances our understanding of ethical issues in citizen deliberation.

The birth of an industry?

The expansion of civic engagement initiatives over the past two decades has led to a booming field of practice and practitioners, dubbed by some as an ‘industry’ (Hendriks and Carson 2008; Lee 2015) that has become professionalised and commercialised (Bherer et al. 2017a, b, c). As a result, deliberative democracy has seen increasing consolidation of a community of practice (Hendriks and Carson 2008), inhabited by professional deliberative engagement specialists, scholars, activists and international networks and organisations. This group is made up of professionals who specialise in the design, delivery, study and promotion of participatory processes, including DMPs. Academics also inhabit this space, either solely as researchers or as ‘pracademics’ collaborating in both research and practice (see Powell et al. 2018). There are global networks and organisations facilitating connections across geographical and professional boundaries. They frequently promote or advocate for participatory activities through networking, producing best practice guidelines, training programmes and research on participation. Some prominent examples include the International Association for Public Participation (IAP2) and the Democracy R&D Network, who promote deliberative and participatory processes around the world. IAP2 has a broad scope of activities (beyond DMPs) and their activities include training, annual awards, and copyrighted protected tools and guidelines for evaluating participatory practices (IAP2 2023). Democracy R&D has a narrower focus on sortition-based deliberative processes only, and its activities are focused on learning across the network, capacity building and advocacy.

Across this community of practitioners, academics and advocates, there are different viewpoints on the purpose, role and ideals for deliberative mini-publics (Parry and Curato 2022). Lively contestation around their democratic value has long been present in the deliberative mini-publics community, but as the practice of DMPs widens, further reflection on this heterogeneity is needed. Practitioners are advocates of the process, so they inevitably need to highlight the good, at least in public. At the same time, they have a more grounded understanding than anyone of DMPs’ limitations. The findings in this paper surface the lived experience of practitioners as they navigate these limitations and ethical issues. How they deal with these constraints shapes the future trajectory of deliberative mini-publics.

There is an apparent tendency in the literature on democratic innovations to under report failures or negative effects of participation (Curato et al. 2021; Hanson 2018; Spada and Ryan 2017). Existing research suggests that this is exacerbated by the commercialisation of participation, where a community of practice closes ranks around perceived failures to protect their professional reputations (Chilvers 2017). From his study of UK public dialogue networks, Jason Chilvers (2013, 2017) suggests that this limits the types of self-critique that practitioners engage in, focusing on process design or quality, rather than structural or ethical constraints that more fundamentally challenge the premise and conditions of participation. Similarly, Caroline Lee argues that practitioners (in North America) generally work within the confines of existing systems and organisations, where participation extols the virtues of democratic emancipation, but at the same time must remain managed and manageable (Lee 2014, p. 493).

This is especially relevant as DMPs become more popular and face increased ethical risks. The professionalisation and commercialisation of public participation makes deliberative participation more susceptible to ethical risks such as co-option, manipulation or undue political influence (Bherer et al. 2017a, b, c; Hendriks and Carson 2008; Lee 2015). For some, the establishment of an ‘industry’ operating in a commercial context is directly in tension with principles of deliberative democracy (Hendriks and Carson 2008). Practitioners may place the interests of their client—usually the commissioning body—above the deliberative and democratic features of the process (Bherer et al. 2017a; Christensen and Grant 2020). Attempts to steer a DMP by the commissioning body could compromise the core features of a DMP, namely claims of inclusiveness and representativeness achieved through sortition, and well-informed deliberation (Curato et al. 2021).

As a researcher with practice experience in this field, my own experience of the deliberative mini-publics community of practice challenges these critiques. The community is awash with training calls, webinars, workshops and conferences to share learning and enhance cross-fertilisation of knowledge, but further research is needed to establish if this corroborates Chilvers’ (2013, p. 3–5) findings that learning ‘remains largely instrumental, crowding out and denying deeper forms of reflectiveness and reflexivity’. Our findings suggest that this conclusion might do a disservice to practitioners in the deliberative mini-publics community. Practitioners work at the coalface of deliberative mini-publics, in the thick of it, and we know that they hold strong personal commitments to deliberation and democracy (Cooper and Smith 2012; Friess and Herff 2023). Yet despite their role as ‘crucial agents’ (Willis 2018) in designing and implementing DMPs, we know surprisingly little about the everyday practices that practitioners engage in (Escobar 2022).

This paper advances knowledge of ethical issues in deliberation by surfacing the grounded experiences of practitioners as they navigate the tensions outlined above. It demonstrates practitioners’ willingness to address ethical issues head on, and their suggestions for mitigating ethical risks. It also highlights the kind of reflexivity that some practitioners engage in, which, counter to previous studies (Chilvers 2013, 2017; Lee 2014, 2017), often involves reflections about the structural constraints that DMPs face and the limitations of their individual agency to deal with ethical dilemmas. In doing so, it brings practitioners’ experiences to the fore. This provides a fruitful foundation for further dialogue and reflection on deliberative mini-publics between theory, empirical research, and practice—a ‘necessary recursivity’ (Mansbridge 2022) for the field as it matures further.

Methods

This paper analyses 25 semi-structured interviews with practitioners and advocates of DMPs. Interviewees’ work usually involved direct engagement by the commissioners of deliberative mini-publics: they are paid to provide a service (Bherer et al. 2017a, p. 5), although many were involved in other forms of deliberative work. The interviews focused specifically on deliberative mini-publics, which is only one form of deliberative practice. Many participants had experience in other types of deliberative and participatory processes and brought this experience into the discussion when relevant. Interviewees were based in Europe, Australasia, North America and Latin America. These regions include the OECD countries where an acclaimed ‘deliberative wave’ has been posited (OECD 2020a). Europe and Australia were overrepresented in the sample, and are especially relevant given that DMPs here are most often commissioned by public authorities and design and delivery is ‘outsourced’ (Christensen and Grant 2020; Wan 2018) to professional practitioners.

These interviews were part of a broader research project into the ethics and integrity of DMPs, with three researchers conducting interviews. We recruited interviewees through our existing connections in the deliberative mini-publics community, and further people through snowballing. We are also members of this community, and we were known to some interviewees. When there was a closer working relationship between an interviewee and one of the research team, another member conducted the interview. Interviewees were aware of and interested in the aim of the research and often asked about plans for the work, and findings so far. Our shared background and knowledge meant that we could empathise with some interviewees’ difficult experiences and recognised the sensitivity of the topic, having been involved in the design, implementation and evaluation of DMPs ourselves. As interviewers we are, thus, not disinterested or detached parties from the community of practice which we are researching, nor did we set out to give this impression.

We designed interview questions to elicit interviewees’ experiences of ethical issues in deliberative mini-publics. Interviews were semi-structured to allow us to probe further on certain issues as they came up, and for interviewees to bring in their own views on the questions themselves (Swain 2018). Interviews took place between November 2021 and November 2022. They were audio recorded, transcribed and anonymised before being entered into NVivo for qualitative thematic analysis. A hybrid approach to qualitative thematic analysis enabled the use of both deductive and inductive reasoning (Swain 2018). Four broad themes formed the basis of the initial analysis. These were the range of experiences with commissioners, responses from practitioners, structural limitations, and broader context and reflections. These themes were based on both existing literature and the research aims of developing our understanding of the ethics and integrity of DMPs. Based on the initial analysis, I identified the three ethical pressure points and responses presented below.

Ethical issues in deliberative mini-publics

The analysis is presented in two main sections, ethical issues and responses to them. Responses from practitioners do not necessarily address each pressure point in turn, but rather summarise the red flags and red lines that practitioners follow to anticipate ethical issues both before and during a deliberative mini-public. Practitioners’ broader reflections are included throughout these two main sections. I first present three types of pressure that practitioners experience: pressure to steer the process, pressure to achieve impact and transformation; and the pressures of commercialisation. These three pressures can occur at different points before, during and after the deliberative mini-public, come from different actors, and from structural constraints faced by commissioned DMPs.

Pressure to steer the process

They were really worried that the jury were going to say no…and that I had to give them a maybe. And I said, I’m sorry. I’m sorry. You know, we’ve been through this process over 6 days. This is—I am holding this space. That’s the jury’s call. And if you ask them a yes or no question, and they feel they’re not being heard, they’re going to answer no... it was awful, the most horrendous experience I’ve ever had.

This is how MaggieFootnote 1 described one of the most challenging experiences of her career, facing significant pressure from a commissioning authority to steer the direction and outcome of the DMP she was delivering. ‘It’s not my job to engineer you an outcome’ was her reply, echoing similar responses from other interviewees.

Several other interviewees reported attempts by authorities to steer the process. This could occur at different points in the DMP. First, the selection of expert witnesses and provision of information to participants was cited as a particularly sticky point. For most this was described as ‘heated discussions’ but not insurmountable, whereas for a few there were more overt attempts. Recent research has found that the selection of experts and information in DMPs happens largely without transparency or meaningful oversight (Roberts et al. 2022). This type of influence threatens some of the core claims of a DMP—that the deliberation is well informed and inclusive, and legitimate (Curato et al. 2021). A second point where this pressure could be felt was commissioning authorities trying to change the content of final reports written by the citizen participants, with one interviewee recalling an authority threatening to withhold payment from contractors until their demands were met. Third, a few interviewees told us about incidents where stakeholders or interest groups had attempted to influence citizen participants directly (rather than practitioners being on the direct receiving end). To this end, some interviewees recounted measures they took to try and protect participants from this kind of influence. Fourth, participants themselves may try to steer the process. We heard about two cases where, through a genuine sortition process, participants with vested interests happened to be selected, and tried to undermine the deliberative process from within.

Interviewees reflected on why various actors try to exercise undue influence over a DMP. There is a degree of unpredictability to a deliberative process that can alarm commissioners. Some interviewees felt that fear of public backlash can hamper the governance and progress of a DMP because commissioners are not open and honest with practitioners about the politics and political stakes around the process and issue. This paves the way for further problems when commissioners are unwilling to relinquish power over certain aspects of the process. This was significant for some interviewees because,

…In the end, power is at the heart of this. I think it, you know it all comes back to power and who has power and how you know if something has deliberative integrity, it is in some way, going to behold somebody who’s got power, it’s going to constrain them in some way.

In other words, if a process does not successfully demand that decision makers to give up some of their power, then its integrity is compromised. Critics of DMPs are likely to shrug their shoulders at this finding: some have long argued that DMPs fail to shift to existing distributions of power (e.g. Fuji Johnson 2015; Hammond 2021). Yet, for some practitioners, this is an important ideal to pursue in practice. This issue is distinct from the ongoing debate over the level of power that DMPs should be endowed with—whether they should be decisive or consultative (Rangoni et al. 2021). Commissioning authorities can attempt to exercise excessive control over a process regardless of its level of empowerment, although a closer tie to decision-making may raise the stakes (Mansbridge 2022).

In a context where DMPs are becoming more mainstream, one interviewee noted that ‘it’s been harder and harder to ignore some of the power factors that are emerging’. They gave the example of a high-profile process that had been heavily influenced by politicians and bureaucrats in the design and the final outcomes. As interest in DMPs grows, they become bigger investments for commissioning bodies in terms of both resources and political stakes, relative to other forms of consultation. This is a departure from earlier days, where certainly in OECD countries, DMPs were more experimental and small scale, with commissioners enthusiastic to engage with an experiment in deliberative democracy for its own sake.

A few interviewees described how difficult it could be to get access to senior decision makers, who may not want to deal with an issue head on or were not fully invested in the deliberative process. In one example, the junior person that the practitioner dealt with was very enthusiastic about the concept of deliberation, but did not have sufficient decision-making authority, making it difficult for the practitioner to plan effectively. It turned out that more senior figures in the commissioning authority were actively hostile towards the deliberative process and wanted it to fail.

Pressure for impact and transformation

If the process does not have an impact on their lives that they want…there’s a question there…you know, you’ve asked them to make these sacrifices in order to participate, emotional or economic or whatever. And are you delivering the kind of the impact on people’s lives that that they are hoping for, and that you’re promising to some degree or another?

For David*, a DMP having meaningful impact for affected communities was a matter of ethics and integrity. When practitioners bring people into a deliberation, especially people who face significant barriers to participating in the first place, those people should be able to see and feel the results of their democratic labour. Many interviewees recounted their frustration about what happened—or rather didn’t—after the conclusion of the mini-public itself. They felt that they had delivered robust, empowering experiences for participants that went on to have very little impact in terms of recommendations or concrete results. From this perspective, achieving impact and making positive changes to peoples’ lives is not only an instrumental aim of a DMP, but a normative commitment. The impact of DMPs is constructed subjectively (Parry and Ercan 2024). For some interviewees it meant having concrete policy impact, as well as improving ongoing relations between citizens and decision makers. For others like David, it was about making a difference to everyday peoples’ lives as a result of their participation. Unlike the pressure to steer the process, the pressure for impact seems to come not only from commissioners or observers, but also from practitioners themselves, and their own hopes and ideals for what DMPs could potentially achieve.

In most cases, DMPs do not have accountability mechanisms built into them that enable the scrutiny of decision makers’ responses. Funding and resources usually end after the deliberative event, leaving little scope for follow up. There are a few exceptions emerging, such as the Ostbelgien model (Setälä 2020). Nonetheless, what happened to the output of a DMP was a point of frustration for most interviewees, and a point in the process where they felt they had relatively little power to do anything about it, compared to the process design. Interviewees reflected on the limitations of what they could achieve in a context where they were commissioned by an authority to design and deliver a DMP, and the structural constraints they had to work within. Their experiences reflect the tension within which commissioned DMPs are often situated—between democratic emancipation and elite control (Asenbaum and Hanusch 2021; Hammond 2021; Lee 2013, 2014). Integrating DMPs into broader political systems is a structural challenge, and ‘it’s kind of the central tension of our work…we do deliberative democratic experiences for people in the midst of political systems that are mainly not deliberative or democratic’.

These concerns went beyond the impacts of the DMP. The notion that DMPs should enable something transformative was shared by some interviewees. From their perspective, transformative potential does not refer to the transformative experience of participating in a DMP (e.g. Ehsassi 2022) but rather the potential of the deliberation to surface and challenge dominant economic paradigms, structural inequalities and systems of power. They felt that DMPs had genuine transformative potential, but that this was constrained by the commissioning context, not only in the impact of the process but also within the deliberative process itself:

My experience is that the framing and the commissioning determines what’s possible or not and potentially, that’s what limits the systemic transformative potential of these processes…I personally don’t think the problem is with the tool. I think the problem is actually with the context it operates in, the commissioning bodies, and who has the agenda-setting power.

From this perspective, a DMP’s transformative potential is constrained by an unwillingness on the part of commissioning bodies to address structural issues which may be outside the immediate remit set by commissioners but are nonetheless part of the overall context or framing of the issue under discussion. This can also be conceived of as another type of pressure to steer the process from commissioners, by limiting the agenda. But the interviewees who discussed this point did not describe this quite in the same way as in the section above. It occurred in more subtle ways and for a few people, were intrinsically related to the practitioners’ own personal beliefs around the transformative potential of deliberation. These issues were not uniformly salient across all interviewees, which highlights the points that different practitioners hold different ideals about DMPs, in turn shaping their perception of what is and is not an ethical issue.

One interviewee gave the example of closing down discussions around economic growth as a default paradigm during a deliberation about climate change. When these discussions are bracketed out, a DMP risks ‘fitting in what ought to be a political process into this very depoliticised policy space’ where there is little ‘space for real discussion or deliberation because it’s like there’s not much to discuss if you’re doing a little bit of dial turning’. Another interviewee had observed something similar:

What I have found to be challenging is, again, this is just my perception…is like a reductionism of complexity when thinking about issues…what I’m noticing is… often a marginalisation of those interdependent issues and challenges, so things like anything to do with oppression and power seem to be missing, like a lack of recognition.

Several interviewees who raised these points had worked as facilitators, and they observed how micro-deliberative details, design and facilitation could work in ways that marginalised certain viewpoints, intentionally or unintentionally. The ways in which designers and facilitators summarise or paraphrase participants’ arguments, for example, could serve to reframe an issue (Landwehr 2014). The ever-present threat of being short on time could be another justification for parking some concerns and never getting around to discussing them fully.

Power in a DMP is exercised in different ways. For our interviewees, time was the most frequently mentioned constraint they faced in general. Insufficient time, or strict timelines, are a compounding structural constraint because they impact so many other aspects of the process and participants, including the freedom to explore interdependent and systemic issues, which may take deliberation away from the immediate goal of addressing a boxed-off remit. Despite being mentioned as a major constraint, some interviewees took limitations on time as a given, the reality of fitting into established systems of electoral and policy cycles. It was most often framed as a potential red flag or something to be negotiated at the start of a project—a design feature—rather than something that could be potentially weaponised as a form of undue influence, or a structural constraint. Yet, time is power in a deliberative mini-public. Control of that time is one way in which DMPs can remain managed, circumscribed spaces for decision makers. This resonates with Caroline Lee’s (2014, p. 498) assessment of public participation professionals as working within the ‘competing logics of democracy and rationality’. In this context, practitioners struggle to reconcile the democratising, empowering and transformative work they believe that deliberation could achieve, with the requirement that deliberation remains managed and manageable, within a restricted framework set by those in power.

Some interviewees talked candidly about these restrictions in broader terms. Their views challenged the boundaries of deliberative mini-publics and reflected on the contexts in which they are deployed (Chilvers 2017, p. 131, 132). For example, most interviewees were careful to qualify their support for DMPs with caveats and heeds for caution. This is notable in a context of DMPs’ increasing popularity, and the potential for marketing claims to overtake knowledge claims:

I couldn’t be just a purely naïve advocate of mini-publics as something that will change democracy forever. I believe in... mini-publics as a source of, marginal change in the political system. But at the same time, if I want to take seriously their potential, I have to produce critical accounts also.

Another cautioned against attributing ‘a kind of folk heroism’ to DMPs: ‘I think we have to be careful with their claims to legitimacy overall and we have to be careful to ascribe them overriding mandates’. These qualified claims demonstrate practitioners’ grounded reflexivity on the limits of DMPs, and their realistic expectations for what they can achieve.

Pressure of the market

If I wanted to be rich or have a big business I would not be a specialist in democracy, I can tell you that for sure! I do this work because it’s how I can use my skills and talents to improve and strengthen democracies and because it aligns so deeply with my values. My values drive everything I do.

Unsurprisingly, many interviewees, like Annie* above, actively contested the claim that commercialisation means that practitioners are motivated by profit over democratic ends. The sentiment shared above was echoed by interviewees who talked about their personal values and commitments to deliberative democracy as important motivating factors in pursuing this line of work. Others talked about their own personal integrity, and how this supported them in standing up to certain ethical pressures.

There was no consensus on commercialisation of the field. A few explained that their organisation being a non-profit was important in this regard, and they expressed some doubt about the motivations of commercial businesses in the field. Others dismissed this suggestion, with one person speculating that the distinction between profit and non-profit was ideological rather than grounded in reality: ‘non-profit organisations still rely on an income, and they need to generate funding, and they also need to meet funding targets. So the question for me is, what is it about? What is it about a commercial organisation that seems to generate such discomfort?’.

A few people felt like commercialisation and competition held back collaboration in the field. Several interviewees mentioned the importance of transparency and collaboration for the general advancement of the field, which required things like sharing best practice and materials, and lessons from things going wrong, because ‘you learn as much from the bad ones if not more…you know, as a field, you’ve got to own up to what you get wrong’. Nonetheless, a couple of interviewees felt that there was resistance to this kind of openness:

It makes zero sense to me that we are…so protective of our materials that we can’t pick and choose all the best stuff and repackage into the greatest thing for all of us to be using…that just that upsets me to no end.

A more prominent concern for some interviewees was that, as DMPs became more popular, new providers could enter the market. Some fear that new, large firms coming into the field from other areas, might not have the same level of commitment to deliberative democracy that they have (Chilvers 2017), which could in turn make them more susceptible to ethical risks. This is compounded by the fact in many cases, how to deal with ethical issues is not codified and knowledge is held by highly experienced individuals. A fear for some interviewees is that organisations that lack this level of experience may not be quite so scrupulous. For most people who mentioned this, it was more of a worry than an observation, but a couple of people cited instances where they felt that sub-par processes had been run using the deliberative name but not being up to scratch. This contributed to their worries that deliberation could be appropriated and mislabelled, tarnishing the overall reputation of deliberative mini-publics.

Navigating ethical issues

The two sections below describe how practitioners try to anticipate and avoid ethical issues before they arise, and how they navigate ethical issues during a project that is already underway. These include some formal procedures built into their planning and processes at the strategic and governance level, as well as more relational approaches. All these measures are ultimately underpinned by the principle of iterative practice: nearly all interviewees told us how they had refined, redeveloped, and tested their approaches over time. They took important lessons from negative experiences and iterated their approaches for future work. Practitioners have learnt from ‘big mistakes’ and use this knowledge to anticipate ethical issues. These practices remained, for most interviewees, informal and ad hoc. Only a minority had developed formalised procedures for navigating these risks. This illustrates the extent to which the ability to deal with ethical issues is contingent on individual and personal experience, rather than formal systems, within the practitioner community.

Red flags: anticipating ethical issues

Practitioners try to mitigate ethical issues by appraising possible red flags before taking on a project. Interviewees recounted a range of factors that helped them to decide whether to take on a project in the first place that helped to try and mitigate pressure to steer the process and pressure for impact. Time, resources, issue, scope and rationale were frequently mentioned as important components to evaluate upfront. Most interviewees also specifically mentioned that they were looking for the potential impacts a DMP could have, and how clearly the commissioning body could articulate that from the outset. This included looking for the potential beyond a one-off deliberation, the level of upfront commitment given by commissioners to responding to or implementing recommendations, and broader impacts in the community.

Most factors were not hard constraints but rather trade-offs that organisers always dealt with. Interviewees looked for the openness of the commissioning body to negotiating better conditions, such as more time. One described this as an ongoing process of trying to achieve a balance between best practice and ideals for deliberation, with making the process accessible to the commissioning institution. More than one person told us they had knowingly entered projects under less-than-ideal conditions with the aim of renegotiating, or had reached compromises with commissioners on certain features. ‘Constraints aren’t always bad. Constraints ought to impel creative responses’ one interviewee pointed out, whilst another suggested that adapting to constraints and problems as they arise—to ‘design on the fly’—could be quite gratifying. The agility displayed by deliberative practitioners to adapt process designs on the ground to local contexts, even as the deliberation unfolds, provides some counter to the critique that DMPs have become an ‘off-the-shelf’ or ‘cookie cutter’ that is deployed without consideration of local conditions.

There was, however, a limit to the kind of constraints interviewees considered acceptable. Several interviewees described ‘participation wash’ as a red flag, manifested as a commissioner being uninterested in genuine deliberation with meaningful impact (Bherer et al. 2017a, b, c). One example included being asked to run a DMP during an electoral campaign where the commissioner seemed motivated by the potential reputational gains, which the practitioner turned down. A related concern could be termed ‘deliberation wash’: using the term ‘deliberative’ to describe engagements that did not meet the criteria for a deliberative process. This is something that concerned some interviewees, who feared ‘massively damaging misrepresentation’ which could damage the credibility of genuine deliberative processes (see also Bachtiger et al. 2010). At the same time, a few interviewees wanted to point out that this was not to say that all public engagement should be deliberative, because there is a role of other forms of consultation and participation—just that it is damaging to label processes as deliberative if they are not. This concern has led to some emerging conversations about developing standards for deliberative mini-publics—a debate worthy of further research elsewhere (Parry 2023).

A lack of transparency was another red flag. We heard about commissioners wanting practitioners or participants to sign a non-disclosure agreement, which was a big no–no; it implied the commissioning body might want to hide something from the public and control aspects of the process. Transparency has long been hailed as a key ideal of deliberative democracy (Naurin 2004), but this usually refers to the overall benefits of opening up decision-making procedures to discussion and scrutiny through public deliberation. For our interviewees, enacting transparency meant making sure that everybody involved in a DMP had clear information and expectations about their roles and boundaries. To try and avoid some of the constraints outlined earlier, especially the exercise of undue influence, all those involved in a DMP needed to accept their role and what behaviours were ethically unacceptable during a DMP. One interviewee explained that ‘the best [practitioner] organisations I’ve worked with have always been very clear about the terms on which the deliberative engagement will take place’. This was formalised to varying degrees, from contracts and codes of conduct to information sessions for everyone involved.

Transparency also demands practitioners are clear upfront with commissioners about what can realistically be achieved given the time and resources available, and don’t over promise. Some interviewees opted for more relational and long-term approaches, such as capacity building with public servants and developing good working relationships with public authorities who they worked with repeatedly. When commissioning authorities—including bureaucrats who are often the most closely involved in the delivery of the DMP—made power dynamics and political stakes explicit early on–‘trusting in sharing with us [practitioners] their fears’, this built a foundation for a more constructive relationship. Interviewees recalled positive relationships with commissioners and public servants as being those with clarity and openness to ongoing negotiation on design and process. Some interviewees mentioned having this kind of openness from the outset help lessen the likelihood of commissioners trying to steer the process later on, because it enabled practitioners to design a process with these risks in mind.

Whilst these measures foster transparency between the commissioning body and the practitioner, they do not necessarily open up the governance of a DMP to wider public or even academic scrutiny. In the era of commissioned mini-publics, transparency is advocated as best practice (e.g. OECD 2020b), but this has, thus, far not been extended to the backstage workings and governance of a mini-public, although a minority of practitioners publish materials such as process designs. However, the processes behind selecting expert witnesses (Roberts et al. 2022) or preparing briefing materials (Mazur et al. 2023) remains largely opaque, and the backstage ethics of governing a DMP have, thus, far received little scholarly attention (Ross et al. 2023).

Red lines: navigating ethical issues

Sometimes, difficulties didn’t become apparent until a project was already well underway, demonstrating that not all ethical issues can be avoided or predicted in advance. Some interviewees told us how they had struggled with the decision of whether to stick with a job under difficult circumstances. Most had ‘never had an unresolvable problem’ and a few interviewees said that ‘sometimes it’s important to go ahead with a project in conditions that are unfavourable. In some ways it’s easier to opt out and say your ethics are above this, than it is to engage in the messy political reality’. This view recognises that working within constraints is par for the course for practitioners.

In some difficult situations interviewees had felt they ‘had to hold the line’ and stand up to pressure from commissioners, when certain lines had been crossed. This occurred most often in standing up to pressure from commissioners to steer the process. Most interviewees did not have a formal process established for this, but rather relied on their experience and reputation to uphold the integrity of the process. Interviewees told us about some cases where they could use their reputation and credibility in the field as leverage to negotiate better conditions, hold commissioning authorities to higher standards or push back in the face of pressure. As one interviewee with years of experience put it, ‘as well as being bad for my professional reputation…I’ve got my own values…if you want someone to do that kind of job [do what the commissioning body demands], find someone else’. These examples show that the level of experience that a practitioner has matters for deliberative integrity—those new to the field are less likely to have this kind of leverage or the security to turn down work.

When things did go badly wrong during a DMP project due to interference from commissioners or other parties, a few interviewees had threatened to pull out of the process if the behaviour was not rectified:

I said, basically that for me to keep working with them, I need to know that they’re going to behave well…I think these processes are very worthwhile, and if people are going to actively undermine it, I can’t work with you.

Just one had actually walked out of an active contract when ‘there was nothing that we could do to crawl back those integrity-based measures that were being broken’. In this case, the commissioning authority insisted on using a citizen deliberation method that practitioners felt were unsuitable for the affected community and issue under discussion, and it was this insistence that led them to walk away from the project.

Practitioners are not the only actors that can help to uphold the integrity of the process. Some interviewees emphasised the role of participants as important in this regard, for two different reasons. Some argued that organisers having a hands-off approach provided evidence to the broader public and observers that the deliberation was completely independent and hadn’t been interfered with by decision makers or practitioners (see also Bagg 2024). ‘Panellists very soon into the process start to be better defenders of the process than we are’, one pointed out. Participants recognise easily attempts by commissioners or stakeholders to steer them, and they both resist and report it. Another reason for giving participants a central role in navigating ethical issues is that they have greater legitimacy than professional practitioners, so their input into the process is preferable, and more democratic, to having the entire design and process as expert-led by practitioners. This entails having participants write their own ethics codes, have autonomy over expert selection, and have sole control over recommendations and final reports. But this kind of participant involvement remains relatively narrow, within the confines of a predetermined method, timeframe, and agenda.

Aside from the strategic and governing levels, practitioners also push back against pressure in more subtle ways. In response to the tendency some interviewees had observed of sidelining marginalised viewpoints or structural discussions, facilitators in particular attempted to counter this by supporting participants from marginalised backgrounds and opening up conversations about structural inequalities. In the past, neutrality or impartiality was often presented as the ideal to which facilitators should be working (e.g. Bherer et al. 2017b; Landwehr 2014). But several interviewees countered this notion, one stating ‘I am not neutral about power in the room’. This links to the transformative potential that some interviewees felt was constrained in a commissioned context because ‘the most transformative conversations tend to come from when you access this, kind of, deeper level’. This finding is corroborated by existing research suggesting that facilitators are not always ‘neutral actors’ (Curato et al. 2023: p. 18; Escobar 2019) in DMPs but may try to actively resist, in subtle ways, the constraints they face in a context when they do not control the agenda. But the interviewees that mentioned this also faced frustration that ‘there has not been space in the structures that I’ve been given to explore power in any meaningful way’. Here, interviewees felt the limits as to what practitioners can achieve through adapting process designs and navigating on the ground. Some interviewees doubted the extent to which they alone could deal with some of the more fundamental ethical challenges, and indeed, whether DMPs can ever deal with issues of power and marginalisation. As one interviewee put it, their power in these situations is ‘tentative’. Practitioners are not naïve to the contexts and boundaries they work within and demonstrate reflexivity and meta-deliberation (Dryzek 2009) on the conditions under which DMPs take place.

Conclusion

Deliberative practitioners have extensive experience of the range of ethical issues that DMPs face. Through iteration, they have developed ways to try and mitigate these risks. Practitioners leverage their experience and reputation as a way of resisting pressure from commissioners and other parties. The practices they have developed are dynamic and agile, responding to changing contexts and expectations.

Designing pathways for impact into a DMP is an example of this: in the early days of DMPs, focus was on demonstrating citizens’ capacity to deliberate and the quality of the deliberation (Rosenberg 2005). This emphasis has been well critiqued in the scholarly literature (e.g. Chambers 2009; Lee 2014; Polletta 2014), and research has moved towards understanding the impacts of deliberative processes beyond the forum itself (e.g. Jacquet and van der Does 2021; Parry and Ercan 2024) and their interaction with broader political systems and actors (e.g. Ghergina et al. 2023). The field of practice has likewise evolved, away from this earlier, proof-of-concept phase, to a point where practitioners need to demonstrate that these processes can produce significant enough impacts to justify the considerable investment they require. But to frame this as merely an instrumental shift does something of a disservice to the community of deliberative practice. Our interviewees articulated strong normative commitments to achieving impact and the transformative potential of DMPs, adapting their practices to try and secure that impact as a matter of integrity.

This research has limitations which should be addressed in future research. First, our interviewee sample was not representative of the full range of locations and contexts in which DMPs take place. Whilst interviewees were based in several regions around the world, Europe and Australia were overrepresented, and this analysis did not include interviewees working in authoritarian contexts. I have removed all identifiable information from the interview data presented here, including location, to protect the anonymity of interviewees. But this also serves to flatten out differences in the contexts that interviewees were working in. Given that the ways in which DMPs play out are highly context specific (Hanson 2018), it is reasonable to assume that ethics and integrity risks will also vary and this deserves due consideration in further research.

In addition, the focus here has been the relationship between practitioners and commissioning authorities, but in reality, these are just two types of actors in a complex web of interactions that includes stakeholders, bureaucrats, activists, academics, civil society organisations, citizens and others. Each of these actors brings their own ideals and expectations around the why and how of a DMP. Even when we just focus on practitioners and commissioners, there is huge variation. ‘Practitioners’ is an umbrella term that can include designers, facilitators, advocators, recruiters, and more—all with different levels of agency. Commissioners are likewise diverse, from local councils to parliamentary bodies, to national governments (Boswell et al. 2022). Moreover, even in the same commissioning context, authorities convene deliberative processes for a variety of reasons (Ghergina et al. 2023), and DMPs have a variety of different governance set-ups that could shape responses to ethical issues, such as oversight committees. Further comparative research is needed to understand the different ways in which ethical issues manifest across different commissioning contexts and cases, different relationship dynamics, as well as different ways of addressing them.

As the stakes of DMPs are raised through their increasing popularity and public profile, so too are the ethical risks they face. Addressing power is crucial to counter the risks of manipulation, co-optation, or corruption. Practitioners themselves have a lot at stake, not least professional reputation and securing an income. When faced with insurmountable constraints, practitioners actively grapple with how to respond. Their strong normative commitment to deliberation and democratic values goes beyond performativity. The research presented here has, by laying open the backstage process dynamics of a DMP (Meyer and Quattrone 2021), demonstrated that deliberative practitioners have the potential to push back against pressure to steer the process, and apply pressure to achieve impact and transformation. Sometimes it’s explicit and happens at the interface between authority and practitioner—refusing to give in to a commissioner’s attempts to control the information citizens receive. Sometimes it’s subtle and occurs at the micro-deliberative level—ensuring that a small group conversation about power does not get sidelined; making sure that marginalised voices at the table are heard equitably. Either way, deliberative practitioners play a crucial role in upholding the integrity of deliberative mini-publics, which will only grow in importance as the practice spreads.

Regardless of practitioners’ strong normative commitments to deliberative democracy, this leaves us with questions about the legitimacy of a small group of deliberative engagement specialists whose work is unregulated and who remain mostly unaccountable for the decisions they make in the moment, however, commendable. Some interviewees were aware of this dilemma, advocating more and more for citizen participants to take over as much of the deliberative process as possible because ‘they’re the ones with legitimacy’. Nonetheless, even this approach keeps deliberation within certain boundaries—the bounds of the chosen method; the bounds of time; the bounded remit. In other words, this empowerment may occur ‘in specifically delimited ways that do not challenge the power of their administrative sponsors’ (Lee and Romano 2013, p. 749; Hammond 2021). This may be further problematised if DMPs become more mainstream and the industry around them more standardised, with less flexibility for adaptation and agility in deliberative practice. Further, economic pressure means that practitioners might, despite their strong normative commitments to deliberative integrity, end up taking on work they feel less than comfortable about because they need to bring in an income.

This research has implications for the future of deliberative mini-publics in practice. It highlights the increasing ethical constraints that practitioners must deal with as DMPs become more popular, and their commitment to upholding integrity in practice. This runs counter to existing evidence that practitioners’ upholding of democratic ideals is only performative (Lee 2014). Rather, I find that practitioners can and do push back in practice. At the same time, some practitioners acknowledge the limitations of their agency in the face of structural ethical constraints, demonstrating reflexivity. This analysis demonstrates that, in line with previous research (Bherer et al. 2017b), practitioners have astute awareness of the different contexts in which mini-publics are deployed. Practitioners learn from difficult experiences and adapt, tweak and iterate their practices to meet both ethical challenges and changing contexts. These practices often take place behind the scenes and in the moment, under intense pressure. Yet in more or less subtle ways they shape the ways in which a deliberative mini-public unfolds, with potentially significant consequences for democracy.