Over the past two decades, deliberative democracy – ‘a normative theory of democratic legitimacy based on the idea that those affected by a collective decision have the right, opportunity, and capacity to participate in consequential deliberation about the content of decisions’ (Ercan et al., Reference Ercan, Hendriks and Dryzek2019: 23) – has taken center stage within democratic theory. Deliberative democracy is a ‘talk-centric’ account of democratic theory (Chambers, Reference Chambers2003: 308). But is everyone able to talk? Has everyone heard? Studies show inequality between interlocutors; we hear some voices while others are silenced. How do unequal power relations affect deliberation in small-group settings? What steps can we take to ensure that deliberation is democratic? This paper provides an answer to both these questions.
Although many scholars of deliberative democracy have shifted their focus to conceptualizing larger deliberative systems (Afsahi, Reference Afsahi2020; Beauvais and Warren, Reference Beauvais and Warren2019; Chambers, Reference Chambers2013; Mansbridge et al., Reference Mansbridge, Bohman and Chambers2012; Niemeyer, Reference Niemeyer2014; McKay and Tenove, Reference McKay and Tenove2020), small forums remain a pertinent area of study. Research from conflict-ridden contexts such as Northern Ireland (Luskin et al., Reference Luskin, O’Flynn and Fishkin2014), Bosnia-Herzegovina (Dejaeghere and Vermeersch, Reference Dejaeghere and Vermeersch2017), Colombia (Jaramillo and Steiner, Reference Jaramillo and Steiner2014), the Republic of Ireland (Suiter et al., Reference Suiter, Farrell and O’Malley2016), and Belgium (Caluwaerts and Reuchamps, Reference Caluwaerts and Reuchamps2018) shows that small forums can ‘promote recognition, understanding, and learning’ by bringing people of different backgrounds and opposing values together to make decisions (Dryzek et al., Reference Dryzek, Bächtiger and Chambers2019: 1145). However, exploring how we can make deliberation in small forums more democratic and inclusive remains a necessary task.
Deliberation that is inclusive, respectful, and constructive is not ‘self-generating’ (Levine et al., Reference Levine, Fung and Gastil2005: 3); it requires careful planning and numerous interventions to deliver on its promises. Facilitation is a commonly employed – and yet understudied – intervention (Guttman, Reference Guttman2007). Although most deliberative forums are facilitated, scant attention has been paid to the effects of different facilitation styles on the quality and success of deliberation (for an exception, see Gordon et al., Reference Gordon, Haas and Michelson2017).
This paper fills this gap by introducing and measuring the efficacy of new facilitation games with the aim of improving democratic outcomes of deliberation. I argue that facilitation, as commonly understood and practiced, is often not enough to counteract the imbalances in power and influence within deliberation, as evidenced by the gaps between men and women in speaking time and deliberative influence within facilitated deliberative forums (Mendelberg et al., Reference Mendelberg, Karpowitz and Goedert2014a) – gaps that are similarly likely to exist under conditions of racial and religious diversity and inequality. In such circumstances, using innovative facilitation games can be advantageous. Games change the dynamics of deliberation and help participants internalize positive deliberative behaviors. I demonstrate this by drawing on my past experiences as a facilitator and introducing two games: deliberative worth and simulated representation.Footnote 1 The former pushes participants to remain aware of their behavior patterns within deliberation by asking them to choose the best deliberator at the end of each round of deliberation; the latter encourages empathy and perspective-taking by partnering participants and asking them to represent one another’s viewpoints for a portion of deliberation. I use original data gathered from proof-of-concept experiments to demonstrate the efficacy of these games in increasing the propensity of good behavior for all and reducing the behavior pattern gap between visible minority and ethnic majority participants. By my reckoning, this is the first study comparing different facilitation games’ capacity to improve the behavioral content of deliberation and yield more democratic outputs.
Problem of power and the role of facilitation
A frequently lodged critique of deliberative democracy is its disregard of the salient perniciousness of power imbalances between participants. Scholars have noted that marginalized groups, especially women, often have unequal voice and influence within deliberation (for a brief discussion, see Beauvais, Reference Beauvais2019; Cornwall and Goetz, Reference Cornwall and Goetz2005; Hall, Reference Hall2007; Mendelberg, Reference Mendelberg2002; Sanders, Reference Sanders1997; Yaylacı and Beauvais, Reference Yaylacı and Beauvais2017; and Young, Reference Young and Benhabib1996). In conditions of racial and religious diversity, where participants do not have relational positional equality (Anderson, Reference Anderson2010), deliberative processes can become similarly biased in favor of the more powerful participants, further entrenching the marginalization of visible minorities. This runs against the promise of deliberative democracy to eradicate ‘the discriminatory effects of class, race, and gender inequalities’ (Gutmann and Thompson, Reference Gutmann and Thompson2004: 48, 50). In such cases, the role played by facilitators in ensuring the equality and openness of the deliberative process becomes ever-important.
Facilitators can kickstart conversations (Ryfe, Reference Ryfe2006: 87), add nuance to the dialogue by acting as a devil’s advocate or by posing hypotheticals (Beauvais, Reference Beauvais, Bächtiger, Dryzek and Mansbridge2018), and ‘[clarify] the contours of [a] particular perspective’ by ‘[asking] probing questions’ (Levine et al., Reference Levine, Fung and Gastil2005: 10). They are tasked with intervening if conflict arises between interlocutors (Moore, Reference Moore2012). Furthermore, they are believed to be able to combat polarization and moderate attitudes (Bächtiger et al., Reference Bächtiger, Dryzek, Mansbridge, Bächtiger, Dryzek and Mansbridge2018; Grönlund et al., Reference Grönlund, Herne and Setälä2015). Facilitation is also needed to ensure that the conversation proceeds constructively (Menkel-Meadow, Reference Menkel-Meadow2011) through timely participation (Ryfe, Reference Ryfe2006) and by reframing or rephrasing participants’ comments and conclusions in order to find common ground. Most pertinently, facilitation is believed to empower individual deliberators so that they can participate effectively (MacKenzie, Reference MacKenzie, Bächtiger, Dryzek and Mansbridge2018). Facilitators, in this sense, ensure equity in speech (Beauvais, 2018) and fairness of procedure by reminding participants of the rules of deliberation, like respect and inclusion (Smith, Reference Smith2009).Footnote 2 In short, ‘facilitated deliberation is better than unfacilitated deliberation for ensuring both that everyone can speak and that all sides of the debate are heard’ (Beauvais and Bächtiger, Reference Beauvais and Bächtiger2016: 10). It is clear that facilitation is often the unsung hero of successful deliberations.
Facilitation is also wide-ranging and can vary from minimalist/passive to interventionist/involved (see Dillard, Reference Dillard2013; Smith, Reference Smith2009). Kara Dillard (Reference Dillard2013, 220) identifies three models of facilitators: the ‘turn-taking enforcer’, who operates as a passive facilitator; the ‘designated driver’, who pushes the conversation forward; and the ‘quasi-participant’, who takes on the most involved role, which may include interpreting or rephrasing conversations.
Despite the variation, most facilitations fall on the more minimalist/passive side, where facilitators simply call on reserved participants, keep time, or remind participants of the rules of deliberation. Such interventions may be insufficient in counterbalancing inequality in power and authority embedded within deliberative processes. Accordingly, the deliberative behavior of visible minorities and ethnic majority groups will likely reflect their positional inequalities and create a gap similar to the gendered behavioral pattern gap between men and women in deliberative engagements (Afsahi, Reference Afsahi2021; Beauvais, Reference Beauvais2019; Mendelberg et al., Reference Mendelberg, Karpowitz and Oliphant2014b).Footnote 3 I expect that ethnic majority participants within a deliberation engage in negative deliberative behaviors such as interrupting others or dominating speech – much like men. Likewise, I expect visible minorities to behave similarly to women and engage in positive deliberative behaviors like rule-following.
Benefits of gamification of deliberation
Games can ensure a good deliberation when facilitation alone is insufficient. The most common way we think of games is that of ‘system[s] in which players engage in […] artificial conflict, defined by rules, that results in […] quantifiable outcome[s]’ (Salen and Zimmerman, Reference Salen and Zimmerman2003: 80). They impose limitations on who can play, for how long, and where. Within games, ‘an absolute and peculiar order reigns’ to ensure that the game is not spoiled (Huizinga, Reference Huizinga1980 [1949]: 10). However, not all games are built upon conflict. The primary mechanism in a game like Dungeons and Dragons, for example, is cooperation and collaboration rather than conflict.Footnote 4 In such ‘cooperative games, players cooperate rather than compete. Everyone wins – or loses – together. The outcome is either communal joy or reciprocal consolation with shared commitment to do better next time’ (Lyons, Reference Lyons2022: 16). In recent years, there has been a rise in serious games ‘designed purely for learning purposes’ (Din et al., Reference Din, Baig and Khan2023: 1). Also known as applied games, serious games can help game players learn or practice something of importance through gameplay. They can be an ‘essential and versatile resource[s] for […] teaching because of their unique ability to mix serious subjects with entertainment’ (Ullah et al., Reference Ullah, Amin and Munsif2022: 193; see also Lopez-Merino, Reference Lopez-Merino2023).
The effectiveness of games combined with their entertainment value has made them popular in different fields, including civic engagement. Subhodeep Jash, for example, makes a case for incorporating game mechanics into civic life to reshape citizen participation. In order to ‘motivate civic activism’ and address declining participation and trust in government, he suggests the use of badges, rankings, rewards, and points in the non-game context of civic interactions. Looking at cities like Gent, Belgium, and Tlalnepantla, Mexico, Jash concludes that gamifying civic participation helps draw citizens into public decision-making by giving them opportunities to be heard and incentives to embrace their roles as rights-bearing and responsible members of society (Jash, Reference Jash2019).
Meanwhile, Josh Lerner identifies five different types of games that can be used in deliberative and participatory settings and which have proved to make democratic engagements ‘more effective, transparent, and fair’ (Lerner, Reference Lerner2014: 5). These are animation games, team-building games, capacity-building games, analysis games, and decision-making games. For instance, animation games get people to move around and greet each other in a low-stakes, playful way. For example, in Toronto’s participatory budgeting event, the ‘Good Day’ handshake game had everyone stand up, join hands, and twist around the room together. It ‘broke down barriers by connecting people physically’ (Lerner, Reference Lerner2014: 176). As Lerner argues, just a few minutes of this activity noticeably energized the group, helping people feel more alert and comfortable, which led to more open and engaged participation throughout the meeting.
Team-building games, such as the puzzle challenge used in the Rosario Hábitat participatory workshops, helped participants practice collaboration and problem-solving and taught them skills to work together more openly and better listen to each other. The game split residents up into groups and asked them to assemble neighborhood puzzles. Residents soon discovered that some pieces were with other teams, requiring the teams to communicate and cooperate across groups (Lerner, Reference Lerner2014: 119–147).
Lerner also outlines how capacity-building games like the jeopardy game used in Toronto’s Participatory Budgeting program helped participants gain the information and skills they needed, while analysis games, such as the matching puzzle in the same engagement, helped participants break down complex problems into smaller chunks. Finally, Lerner talks about decision-making games like speed dating that help move participants towards identifiable outcomes. A number of elements remain the same across these games. First, games, Lerner insists, should be fun and can make the democratic process, even under hostile circumstances, fun! Second, rules are important, and unwritten rules can marginalize people and create hierarchies. Third, ensuring tangible outcomes helps prevent gameplay from becoming trivial and disempowering (Lerner, Reference Lerner2014).
Games, whether through mechanisms of conflict or collaboration (Huizinga, Reference Huizinga1980 [1949]: 10), create awareness of the behavior of oneself and that of others and impose procedural fairness, as ‘the essence of play is that the rules be kept – that it be fair play’ (Huizinga, Reference Huizinga1980 [1949]: 52). As such, games are designed to push their players to become ‘more cognizant of the social processes in which they interact’ (Gordon et al., Reference Gordon, Haas and Michelson2017: 3791). But games can also make the process more enjoyable and accessible and their participants more attentive (Lerner, Reference Lerner2014: 149–172).
By gamifying the discussion – ‘[using] […] game design element in [the] non-game [context]’ of deliberation (Deterding et al., Reference Deterding, Dixon and Khaled2011: 9) – we can make the process more fun while imparting key knowledge about the rules and aims of the process of deliberation. Here, I highlight three ways games can help facilitators execute deliberation more democratically and empower those facing positional inequality within these settings. First is that of regulating the talk. Facilitators often have to perform a difficult balancing act: ensuring that ample conversation takes place while encouraging the participation of those who are less inclined to talk. The downside of prioritizing equal and effective speaking time for all participants is that facilitators may have to limit the speech of some in favor of others. This can make deliberation overly constructed and not organic enough. Using games can kickstart the participation of those less inclined to speak and regulate the speaking time and behavior of those interlocuters who may dominate deliberation.
Second, by establishing ground rules, games help facilitators avoid the uneasy task of micromanaging unruly participants. A facilitator who is committed to ensuring equal and effective participation of all interlocutors may be tasked with having to focus their efforts on one or two persons within a group by reminding them to respect speaking time or wait their turn. This can have negative consequences if these participants feel they have been singled out. The facilitator’s interventions may be viewed as undue influence by participants who have more structural power – often white, male, better-educated, and with more linguistic and conversational fluency. This can disrupt the deliberation, diminish the quality of conversation, and alter conversational dynamics. In such a case, facilitation games can improve the dynamics without singling out any particular interlocutor. Moreover, these activities put the participants in charge of ensuring that these games are carried out successfully, thereby reducing the direct interference of facilitators (for an account of these concerns, see Moore, Reference Moore2012; Thompson and Hoggett, Reference Thompson and Hoggett2001; Morrell, Reference Morrell, Bächtiger, Dryzek and Mansbridge2018).
Third, games can help fix malignant dynamics of a deliberation, should they arise, in ways that simple facilitation cannot. When negative behavioral patterns are established in a deliberative setting despite the presence and intervention of the facilitators, continuing deliberation as is might not be beneficial. The use of games can reset deliberation and introduce the possibility of establishing new and better behavioral patterns in these settings.
In as much as games can be instrumental in achieving a more inclusive and egalitarian deliberation, they should not be seen as an alternative to skilled facilitation. Instead, they should be used in conjunction with facilitation when needed and decided by the facilitators in charge (e.g., Jungkunz, Reference Jungkunz2013).
Facilitation games
In this section, I outline two facilitation games that can be utilized in deliberative settings. These games can increase instances of positive deliberative behaviors and ensure equal voice for racial minorities within deliberation. They should be seen as tools that can be deployed by facilitators or conveners of deliberation in moments when more minimalist/passive facilitation is not enough.
Deliberative worth games
The first game is a deliberative worth game, which reinforces adherence to the ground rules of deliberation by asking participants to be cognizant of each other’s conduct during discussions. Deliberative worth games are particularly useful when there are high instances of disrespect and rhetorical action, as evidenced by participants cutting each other off or dominating the deliberation, for instance.
Prior to the deliberation, participants receive a list of expected behaviors and ground rules for engagement (see online Appendix, A.13). This list sets out basic norms, such as listening attentively, waiting for one’s turn to speak, engaging respectfully, providing reasons for one’s views, and building on what others have said. On deliberation day, the group spends the first 20 minutes reviewing these rules together and discussing any changes or additions they would like to make. This collaborative rule-making is not just a procedural step – it is essential for fostering a sense of ownership and legitimacy in the process (Lerner Reference Lerner2014). Once the group has settled on a final set of rules, these become the reference point for the rest of the session.
Deliberative worth games ask participants to pay attention to the ideas expressed by their fellow participants and the manner in which they are deliberating. Since all participants have either been involved in designing the rules of deliberation or have been made aware of them from the beginning, this is an easy task. Facilitators may only need to keep track of negative behaviors like interrupting or dismissing others and be attuned to the positive atmosphere created by participants who follow the rules of discourse by waiting their turn or building on another person’s position. At the end of each round of deliberation, participants write down the name of the person they believe most contributed to creating a positive deliberative atmosphere. When participants select the best deliberator at the end of each round, they do so with these collaboratively amended rules in mind. Their responses are then anonymized and shared with the group, and the person with the most votes is deemed the best deliberator. The emphasis here is on democratizing behavior – respectful engagement, careful listening, providing explanations, and building constructively on the ideas of others – rather than simply rewarding the loudest or most assertive voice. This game can be introduced either at the beginning of a deliberation or when the dynamics of deliberation make it necessary.
The idea behind this game is simple. Instead of intermittently remembering the rules or being reminded of them by a facilitator, these games push deliberators to remain continuously attentive to the rules of discourse throughout deliberation. When repeated, this helps participants internalize these rules, thereby raising the quality of deliberation and creating a space where inclusive, respectful, and constructive deliberation can occur.
This game draws on the logic of face-saving strategies. Face-saving strategies refer to the measures we take in social interactions where specific rules of conduct have been established to claim and maintain a positive image of ourselves (see Goffman, Reference Goffman, Laver and Hutcheson1972; Reference Goffman1967). While the particular face-work – strategies aimed at keeping up our self-image in front of others – might look different depending on the cultural context, the underlying logic persists (Ting-Toomey, Reference Ting-Toomey, Littlejohn and Foss2009: 372).
Simulated representation
The second game is a simulated representation game, where participants represent another’s point of view with the aim of improving the quality of dialogue and cultivating empathy. These games are particularly useful when the deliberation is facing instances of cognitive apartheidFootnote 5 or hermeneutical exclusion,Footnote 6 evidenced by participants ignoring what is said instead of reflecting on it or continued misunderstanding between participants. In these games, participants are partnered up to discuss the topic one-on-one. They are provided with a guide as to the questions they can ask one another and instructed to learn as much as possible about their partners’ position and the rationale behind it. This is followed by a simulated representation round. Instead of Person A presenting and defending their own position, they are now tasked with presenting, explaining, and defending the position of Person B. Participants are instructed to remain in character throughout this simulated deliberation. After one round of simulation, deliberation resumes normally. This game is best deployed after one full round of deliberation during which disagreements between participants have become evident. It can be used intermittently during deliberation as a way to get participants to see each other’s point of view.
Simulated representation games are based on role-playing games in education (Blatner, Reference Blatner2005; Sumler-Edmond, Reference Sumler-Edmond2013; Wender, Reference Wender2014) and perspective-taking or imagined contact practices in psychology (Ku et al., Reference Ku, Wang and Galinsky2010; Shih et al., Reference Shih, Wang and Bucher2009; Wang et al., Reference Wang, Kenneth and Ku2014). Recently, some deliberative democrats have looked into the efficacy of role-playing or perspective-taking exercises in deliberation (Morrell, Reference Morrell2010; Muradova, Reference Muradova2020). Most notably, as Gordon, Haas, and Michelson show by using role-playing games in a series of participatory budgeting meetings in New York City, ‘role-playing games can enhance creativity in deliberation, specifically as empathetic identification with others’ and even ‘[eliminate] informational and reputational pressure to keep information hidden’ (Gordon et al., Reference Gordon, Haas and Michelson2017: 3800).
Simulated representation games are meant to interrupt the normal course of deliberation and our usual ways of listening and responding to one another. The one-on-one discussion increases the stakes of listening. Knowing that you need to present and defend another person’s position incentivizes deep listening. The simulated portion of deliberation encourages perspective-taking on the participants’ part, as they will have to take steps to think in their respective partners’ terms to respond to the new issues and questions. Participants are also encouraged to not only hold a particular position but to also defend and justify it in the face of questions from others.
Both games are particularly useful in contexts with deep diversity and inequality – when the particular topic of deliberation is likely to generate contentious dialogue and the ethnic, racial, and religious diversity of the deliberative cohort and their positional inequalities can undermine open, respectful, equal, and constructive deliberation (Afsahi, Reference Afsahi2022).
Deliberative worth and simulated representation are not simply activities but full-fledged games. Each creates a bounded, rule-governed environment where participants strive towards clear goal outlines (winning the best deliberator in deliberative worth exercises and staying true to the views of their role-playing partner), and outcomes are tracked and recognized. Both games incorporate specific game mechanisms, or ‘rules, structures, elements, and processes that designers use to shape gameplay’ (Lerner, Reference Lerner2014: 53). In deliberative worth games, participants are guided by explicit rules for interaction and take part in rounds where they recognize each other’s positive contributions through voting. This introduces collaborative competition and ties participation to measurable outcomes – namely, who is recognized for fostering a constructive climate. In simulated representation games, participants are assigned to advocate for a partner’s viewpoint, tapping into the conflict existing between the paired partners that requires deep listening and empathy. Participants are aware of how long the simulation lasts and know that their performance will be recognized and potentially corrected. This creates a feedback loop and a sense of accountability for participants.
Based on this, I hypothesize:
H1: Visible minority participants are more likely to engage in positive deliberative behaviors – being more respectful or empathetic than ethnic majority groups, reflecting their positional inequalities – in facilitated deliberative forums.
H2: Gamified facilitated forums are more likely to increase instances of positive deliberative behavior for all participants and decrease instances of negative behavior, such as interrupting others or dominating the discussion, than facilitated forums with no games.
H3: Facilitative games are more likely to decrease the behavioral gap between visible minority and ethnic majority participants in their respective deliberative behaviors than a facilitated design with no games.
H4: Groups using facilitative games are more likely to see positive outcomes of deliberation – increased political efficacy, knowledge gains, and willingness for future participation – than groups that do not use such games.Footnote 7
Research design
To examine the effects of facilitation games on deliberative processes and outcomes, I ran eight discussion groups where the participants were tasked with deliberating on whether the province of British Columbia, Canada, should permit the resolution of certain civil law cases through religious arbitration – a sensitive political topic. 40 undergraduate and graduate students from a large public university in Canada participated in 3 deliberative experiments.Footnote 8 The study was approved by the Behavioural Research Ethics Board of the University prior to participant recruitment. The Appendix (A.1; also see A.15) accounts for the key ethical concerns – discomfort and privacy – with this study and the steps taken to address them.
The sample consists primarily of undergraduate students, with one early-stage graduate student, resulting in a relatively narrow age range – making age not a meaningful factor in this study. Moreover, while this study focuses specifically on the role of ethnicity in shaping deliberative dynamics, a separate publication from this project examines gender differences in the same experimental setting and finds broadly similar patterns in deliberative behavior (Afsahi, Reference Afsahi2021). Given the modest sample size (40), further subgroup or interactive analyses by gender or age within the current study were not statistically robust. The demographic breakdown of the sample, including gender and age, is provided in Appendix A.3.
Participants were randomly assigned to three different deliberation settings on three separate days. Both games were played in separate sessions to see their respective efficacies compared to control conditions as well as to one another. The first session – November 1st, 2015 – consisted of 14 participants in three smaller groups deliberating in the presence of a facilitator but without any games.Footnote 9 The second session – November 7th, 2015 – was comprised of 16 participants in three smaller groups and utilized a deliberative worth facilitation game throughout the day.Footnote 10 The final session – November 8th, 2015 – involved 10 participants deliberating in two smaller groups. Simulated representation games were utilized after one round of deliberation.Footnote 11
Participants received an information packet a week before deliberation. The packet contained a pamphlet with basic information about the Canadian legal and justice system, including information about (religious) arbitration practices in Canada, the schedule for each day, and an account of the standards and process of deliberation. Each deliberation session started at 10 am and ended around 2:15 pm. Trained facilitators presided over all sessions, and audio-recording devices were placed at each table.Footnote 12 The investigator (author) was not present in the room for any of the sessions in order to reduce preconceptions about each session. Participants were asked to fill out a pre-deliberation questionnaire before each session commenced and a post-deliberation questionnaire at the end of each session.
It is important to note that participants’ positions on the central issue – whether religious arbitration should be allowed – were not strongly polarized at the outset. In the pre-deliberation survey, 5 out of 40 participants responded ‘never’, 30 responded ‘sometimes’, and 5 responded ‘always’. This distribution suggests that most participants held moderate views, with only a small number at the extremes.
Effect of games on the process of deliberation
Data
To test H1, H2, and H3, which concern the process of deliberation, I used qualitative data from the deliberative discussions. The data is comprised of coded transcripts of all three deliberation sessions.Footnote 13 The sessions were coded for their overall quality as well as the propensity of participants to engage in positive and negative deliberative behaviors as identified by the literature on deliberative democracy.
According to the literature, positive deliberative behaviors encompass providing reasons, showing respect, considering and integrating others’ perspectives, demonstrating sincerity, displaying empathy, and engaging in meaningful dialogue (Afsahi, Reference Afsahi2021: 12). Meanwhile, negative deliberative behaviors occur when participants fail to justify their positions, process information with bias, engage in cognitive apartheid, show disrespect, practice hermeneutical exclusion, employ rhetorical tactics, and participate in unproductive dialogue.Footnote 14 Participants’ behaviors were coded accordingly.
Drawing on existing literature and my experience as a facilitator, I provide an account of the particular indicators for each positive deliberative behavior (Table 1) as well as negative deliberative behaviors (Table 2).Footnote 15
Table 1. Indicators of positive deliberative behaviors in deliberation
Table 2. Indicators of negative deliberative behaviors in deliberation
A number of steps were taken to code the transcriptions according to these indicators. A series of preparatory codes was applied while transcribing the discussions.Footnote 16 In the subsequent cycle of coding, nVivo – a Computer Assisted Qualitative Data Analysis Software – was utilized to deploy hypothesis codes based on the literature and the first round of coding and to code the transcripts. This process was repeated three more times with all identifying information and previous codes removed to reduce coder bias as much as possible.
Results
Do ethnic majority and visible minority participants exhibit different tendencies to engage in positive versus negative deliberative behaviors?Footnote 17 As Figure 1 highlights, there were significant differences between the ethnic majority and the visible minority participants. In the control group, 87% of all coded speech acts by visible minorities were identified as positive deliberative behaviors. Contrastingly, only 49% of speech acts by ethnic majority participants were identified as positive deliberative behaviors.
Figure 1. Gap between visible minority and ethnic majority participants in positive/negative deliberative behaviors under control conditions.
While exploratory, these results substantiate H1: visible minorities are more likely to engage in positive deliberative behaviors. Moreover, societal inequalities and power imbalances are reproduced within deliberative interactions. The tendency of ethnic majority participants to engage in negative deliberative behaviors – interrupting others or dominating the dialogue – reduces the speaking time, influence, and confidence of visible minority participants. This finding is congruent with similar findings on the gender gap in deliberation discussed in Section 1.
Are facilitative games successful at increasing positive deliberative behaviors for all and reducing this gap? Figure 2 provides an overview of the deliberation process with and without facilitative games of deliberative worth and simulated representation.Footnote 18
Figure 2. Comparing average positive/negative behaviors across three groups.
The results, while also exploratory, substantiate H2: facilitative games almost double the instances of positive deliberative behaviors. Deliberative worth games significantly increase adherence to the standards of deliberation. Similarly, simulated representation games help participants become better deliberators. There is no significant difference between the two games’ propensity to increase positive deliberative behaviors.
What effect do these games have on reducing the gap in behavior patterns between ethnic majority participants and visible minority participants? As Figure 3 highlights, through the use of games, the difference in the behavior patterns of visible minority and ethnic majority participants is substantially reduced (cf. Figure 1).
Figure 3. No gap between visible minority and ethnic majority participants in positive/negative deliberative behaviors when games are used.
With the introduction of the deliberative worth game, 92% and 95% of speech acts by visible minorities and ethnic majority participants (respectively) were coded as positive deliberative behaviors. A similar pattern holds for simulated representation (95% to 87%, respectively). Facilitation games can democratize deliberation by reducing the differences between visible minority and ethnic majority participants within a small-group discussion, which supports H3.
Discussion
These proof-of-concept experiments demonstrated that standard facilitation alone was insufficient to eradicate the unequal power relations that resulted in a behavioral gap between visible minority and ethnic majority participants. The games were instrumental to democratizing deliberation by reducing negative behaviors that undermine the voice and influence of marginalized groups.Footnote 19 While games increased the overall quality of discourse, their effect was particularly useful and noticeable for ethnic majority participants.Footnote 20
A few examples from each of the small-group discussions can further elucidate the contributions of the games. In the deliberative sessions without games, participants exhibited negative deliberative behaviors by pushing for false consensus – undermining the democratic aspect of deliberation. There were 12 instances of such behavior under control conditions. By pushing for false consensus, participants signal to one another that they are not really paying attention to the disagreements or points of contention and are processing the information in a way that is biased in favor of what they want to hear. Consider the exchange below:
S1P3: Ok. So we all agree.
S1P1: I don’t agree with that.
S1P3: What?
S1P1: [laughs] I think you’re trying to come to consensus too quickly.
S1P3: I thought you said, I thought you said…I mean, I’m just trying to find common…
S1P1:….Yeah, and I think that’s the problem… Footnote 21
In this case, S1P1 calls S1P3 out on forcing consensus by constantly imposing their point of view onto that of S1P1 while summarizing the conversation.
The negative deliberative behavior of disrespect was frequently observed in the control group (n = 115), where participants often interrupted each other instead of waiting their turn to speak. There were 110 instances of participants interrupting each other. Consider the example below:
K1P4: It’s hard to tell though cause people can just as a personal thing, people just do it and…
K1P3: I know, like, like…
K1P5: Yeah, that’s what I am saying if it’s happening, how much is it happening? How much are we saying it happening and when. Not every single Muslim person, every Catholic, or, you know, Jewish, like…
K1P4: Honestly, not everyone would do that.
What this example also demonstrates is how negative deliberative behaviors can reinforce one another. As K1P4 attempts to make their case, they are cut off by K1P3, who is then cut off by K1P5. Before either K1P3 or K1P5 can finish their thoughts, K1P5 is cut off by K1P4. Such exchanges were rare when the deliberative worth and simulated representation games were utilized.Footnote 22
When participants engaged in deliberative worth exercises, they often undertook self-facilitation by encouraging others to participate by either asking them what they thought or attempting to rephrase or repeat what others had said for the group at large. Overall, there were 57 instances of self-facilitation in deliberations that used deliberative worth games.
Consider this example: at one of the tables, the participants were discussing the possibility of instituting limitations on religious arbitration by using the Canadian Charter of Rights and Freedoms. The specific question was to limit discriminatory practices on inheritance. One of the participants, S7P1, was particularly silent during the conversation. This prompted self-facilitation by another participant:
S7P3: What do you think [S7P1]?
S7P1: I’m not too sure.
S7P3: You’re not sure? What do you, what’s, what’s the, what’s the worry that you have?
[…]
S7P3: what about, what about inheritance?
[…]
S7P1: My, my opinion is [that] religious arbitration, [when] just the two make a decision, to reduce the [costs] according to another religion; […] enumerates, just reduce the cases [decided] according to the religion. So, um, religious arbitration, may or may not protect people [when it comes] to punishment.
S7P3: So, are you saying you are ok with the inheritance here or…?
S7P1: Yeah but not punishment.
This exchange encouraged S7P1 to be more involved in deliberation without active moderation from the facilitator.
Additionally, there were 120 instances of participants listening to one another and building on what the previous person had said. Consider the exchange below:
K7P6: Yeah, I’m, like, even, like, um, I don’t know what to call it, Canadian law or civil law, or whatever. And that’s still interpreted too. So, like, even normal law, it’s like, one judge will see the law one way and one sees it another way. So, in some ways, interpretation of religious law isn’t necessarily […] just, there’s, there is no one definition of any law [too]. It’s always interpretative. Um, maybe you could have some kind of training or even certification; so it would verify credentials or something. This might be an option.
K7P1: So, like trained arbitrators?
K7P5: Exactly.
K7P6: Yeah.
K7P2: There should be a qualification that [sort of, test or] something like that. So before [you do] arbitration, the arbitrator have to be qualified based on some standards or regulations, so in that case the [.] would have some, [a program] I guess, […].
K7P6: Yeah, for sure.
[Facilitator asking if K7P4 has something to say]
K7P4: No, it’s already been touched on. It was just that both people are gonna agree on the same arbitrator, so, like you said, it’s, it’s gonna be, they’re verifying that person.
K7P6: Yeah.
K7P4: And then they’re putting the decision in their hands. So, they’re probably gonna think that person is best.
K7P5: Yeah, exactly.
K7P3: I also feel like having, like, maybe a list of options provided by the, um, somebody. [LAUGHTER]
K7P4: As suggestions?
K7P3: Yeah.
K7P1: So that they if they don’t know anyone, it would be helpful.
This example illustrates a fluid conversation that transitions from K7P2 and K7P6 talking about qualifications and training for arbitrators to K7P4 emphasizing the choice of the people in picking the arbitrators. K7P3, adding to this, talks about having a list. K7P4 asks whether the point would be to suggest arbitrators to people seeking arbitration, and K7P1 justifies this suggestion: if people do not know an arbitrator beforehand, the list would make it easier for them to find and choose one. This is an example of six participants reflecting on the discussion and making one complete point on one theme.
The simulated representation game also led to interesting effects, like opening participants up to see the validity of opposing views. There were a total of 16 instances where participants amended their views – or rather expanded the sphere of reasonable positions – through the course of deliberation.
In the following passage, S8P6 – a strong supporter of religious arbitration – admits that he recognized a flaw in his opinion as a result of being paired up with S8P1 – an ardent opposer of religious arbitration – during the simulated representation game.
S8P6:…but then, the only thing I’ll say is that I’ve been I in the sense that I put in so much trust into the fairness of the religious arbitrators. Believing that they would try to the best of their ability to be very, um, unbiased and fair to both parties. And deciding strictly based on, strictly based on what the rules are and what their opinions are.
While opinion change is not always and necessarily a positive outcome of deliberation, the ability to see beyond one’s own positions and see the logic, reasonableness, and validity of other people’s positions is. This example demonstrates how games make this possible.
There were also many instances (n = 30) where participants engaged in productive dialogue by offering concessions or demonstrating a willingness to compromise. For example:
S8P1: I would say, maybe we could implement religious arbitration. Since, it’s already there, it’s no point to analyze it. But then, I would make point in getting rid of it at some point in the future.
and,
C8P1: Ok. Now I concede that that’s a good example where arbitration could work.
and,
C8P2: But I’m saying, I’m, I am […] willing to compromise.
These results highlight that the gamification of deliberation can equalize behavior between participants with positional inequalities and increase the propensity of positive deliberative behavior for all.
Effect of games on the outcomes of deliberation
Data
This section examines the effects of facilitation games on the outcomes of deliberation. The data in this section was gathered through pre- and post-deliberation questionnaires administered during each session. Many of these pre- and post-deliberation survey questions were drawn from sample questionnaires from the Participedia Project.Footnote 23 Below, I highlight some of the interesting results from this study.
Results
The first item under examination is political efficacy. One aim of deliberation is to boost political efficacy and enhance individuals’ confidence in their ability to participate in similar processes specifically and in politics overall. I operationalized this by asking the participants about the extent of their agreement with the following statement before and after deliberation: ‘Sometimes politics and government seem so complicated that a person like me can’t really understand what’s going on’. The variable was coded from ‘Strongly agree’ (-2) to ‘Strongly disagree’ (+2).
As Figure 4 demonstrates, facilitated deliberation helps increase political efficacy. The largest increase – 0.44 points (from 0.12 to 0.56; p = 0.03) – was achieved when the facilitative game of deliberative worth was utilized. The smallest effect was seen under control conditions (0.21 to 0.36; p = 0.43). Under simulated representation, participants started the day with the lowest levels of political efficacy but experienced moderate gains (0.3 points; p = 0.28).Footnote 24
Figure 4. Effect of games on political efficacy.
There were larger increases in political efficacy when the two facilitative games were utilized. What are the specific mechanisms at work that explain this difference? It is difficult to pinpoint one particularly. Perhaps the overall congenial atmosphere created by engaging in the games enables participants to view contributions to political conversations and decision-making as easy and manageable. This result, while preliminary, substantiates an earlier hypothesis (H4) that games will have a marked impact on a key outcome of deliberation: political efficacy.
The second item under examination is gains in knowledge. Knowledge gains as a result of deliberation signal that ample dialogue has occurred. Do facilitation games have a positive impact on the ability of participants to learn through the process of deliberation? Participants were asked a series of questions checking their factual knowledge on religious arbitration and the legal factors surrounding the issue before and after deliberation.Footnote 25
Figure 5 summarizes the results. The largest gains in knowledge – 0.24 (from 0.47 to 0.71; p = 0.24) – was seen under control conditions. When the games were used, knowledge gains were less noticeable, with the facilitative game of deliberative worth producing the smallest effect (0.58 to 0.61; p = 0.88). The gains were also small when simulated representation was used (0.56 to 0.64; p = 0.77). This result does not support the earlier hypothesis (H4).Footnote 26
Figure 5. Effect of games on knowledge gains.
These results caution us against assuming that games will have a positive effect on all aspects of deliberation and raise the question of why participants learned and retained more information when they deliberated under control conditions. This may be explained by research that shows that many learn better under more contentious circumstances (Brader, Reference Brader2005; Marks et al., Reference Marks, Manning and Ajzen2012; Morey, Reference Morey2017). It is possible that the more contentious atmosphere under control conditions leads participants to learn and remember more from deliberation. One can point to the cognitive effort imposed by participation in the games as another potential explanation. Since participants are preoccupied with the rules of the game, they are less attentive to new information.
The final item under examination is willingness to participate in future deliberations (see Figure 6). Many scholars have noted that willingness for future engagement is a key desired outcome of deliberation (Timotijevic and Raats, Reference Timotijevic and Raats2007; Stromer-Galley and Muhlberger, Reference Stromer-Galley and Muhlberger2009).
Figure 6. Effect of games on participants’ future willingness for deliberation.
Figure 6 demonstrates that the use of games increased willingness for future participation by a small amount (0.15 for deliberative worth and 0.21 for simulated representation). While preliminary, this points to the potential positive effect that games can have within a deliberative setting.Footnote 27 Once again, it is difficult to point to a single mechanism at work that could explain this difference. But perhaps the use of games disrupts the monotony of deliberation and leaves individuals with a more positive outlook on the process.
As noted above, there was little polarization in participants’ views on whether religious arbitration should be permitted prior to deliberation. There was little change in participants’ overall positions following deliberation: post-deliberation, 6 said ‘never’, 28 said ‘sometimes’, and 6 said ‘always’. However, while the overall position remained stable, participants’ views on specific rationales became more nuanced. Support for considering religious arbitration as part of Canada’s commitment to religious freedom increased (mean increase = 0.3, t-test p = 0.021), and there was also a significant increase in recognition that religious arbitration could undermine equality between persons in Canada (mean increase = 0.375, t-test p = 0.023). These shifts suggest that deliberation prompted greater engagement with the complexities of the issue, even if it did not shift overall positions.
Limitations
While the experiments in this paper serve as proof-of-concept tests, they have limitations. The experiment has a relatively small sample size (n = 40).Footnote 28 Moreover, all participants are residents of British Columbia. While the topic of deliberation can cause contentions, Vancouver – and Canada more generally – is a conflict-free and developed context. Replicating this study in a developing context or a conflict setting, such as those in Northern Ireland or Colombia, would shed further light onto the broader applicability of the games.
In addition, the sample is made up of university students, who are generally more accustomed to structured discussion and familiar with deliberative norms as a result of participating in seminar discussions. While this study has shown that students may have a propensity to engage in negative deliberative behaviors, I expect an increase in negative deliberative behaviors for all participants when introducing these games beyond the experiment. That said, I do not expect the general public to struggle with following the basic rules of deliberation or inability to follow the rules of the games. Evidence from Grönlund, Setälä, and Herne’s citizen deliberation experiment shows that providing clear discussion rules – even without active facilitation – was sufficient to prevent polarization and promote deliberative virtues such as respect and justification among ordinary citizens (Grönlund et al., Reference Grönlund, Setälä and Herne2010). This suggests that most people, regardless of educational background, can follow the rules when they are made explicit. Where I do anticipate some difference is in the uptake of the simulated representation (role-playing) game: adults with more entrenched views may be more hesitant to role-play someone else’s perspective. Still, the broader evidence from Josh Lerner’s work on participatory games in citizen engagement shows that people, for the most part, were able to participate and that the games were effective across a variety of settings and populations. Thus, while some adaptation may be necessary, I am confident that the positive effects of facilitation games observed here would largely translate to broader contexts.
Conclusion
How do unequal power relations affect deliberation in small-group settings? What steps can we take to ensure that deliberation is democratic? This paper has shown that unequal power relations – in this case, between visible minority and ethnic majority participants – can have serious consequences for deliberation.
Inequality outside of deliberative settings can, and often does, imbue the dynamics within deliberative settings. While those with less structural power – in this case, visible minorities – follow the rules of deliberation and engage in positive deliberative behaviors, those who have more structural power – ethnic majority participants – reduce the deliberative influence of the former by cutting their speaking time, dominating speech, and ignoring what has been said. That is, some are heard, and some are silenced.
What steps can we take to democratize deliberation? This article has focused on one such step and demonstrated that facilitation games can increase positive deliberative behaviors for all participants and reduce the behavioral gap between visible minority and ethnic majority participants within deliberation. These results highlight the efficacy of these games in ensuring that marginalized groups are given a chance to speak and are heard on an equal footing, making deliberation more democratic.
What does this mean beyond small-group settings? What might this imply for the role and frontiers of facilitation within complex democratic systems with built-in inequality regimes? Mini-publics ‘supplement existing institutions and practices’ and serve as laboratories for testing best practices of deliberation and facilitation (Beauvais and Warren, Reference Beauvais and Warren2019: 908).
In this case, we see that facilitation games can push participants to internalize norms that are needed for good deliberation instead of simply adhering to them. When utilized outside of mini-publics but in similar settings such as schools, universities, organizations, and even legislatures, these games can be fruitful in training individuals for deliberation in the larger deliberative system. We can learn to engage in deep listening by presenting and defending a contra-point for a period of time. We can learn to be cognizant of the effects of our behavior on others around us when engaging in conversations with them, online or offline.
Crucially, by emphasizing the internalization of norms instead of strict adherence to them, facilitative games help us envision how to best protect good deliberators from predation. This requires us to recognize the times when we need to stop being ‘good’ deliberators. First, internalization of norms of deliberation will better equip us to engage in conscious non-adherence to norms when the content of our interlocutors’ speech invariably undermines the possibility of deliberation (i.e., Neo-Nazis in Charlottesville). Second, internalizing the norms of deliberation through repeated facilitation games allows us to identify the circumstances where strict adherence to the norms would further marginalize a minority group by sanctioning or limiting, for example, protest speech (i.e., Black Lives Matter). In both cases, being a good deliberator who has internalized the norms of deliberation requires an active non-adherence to the norms of deliberation.
Supplementary material
The supplementary material for this article can be found a https://doi.org/10.1017/S1475676525100297.
Data availability statement
Please note that the deliberative experiments and survey analyses reported in this article were conducted under Ethics Certificate H13-03158 (June 12, 2014). In accordance with the approved protocol, deliberation transcripts, respondent identities, and IP addresses cannot be shared to protect respondent confidentiality. The anonymized dataset from the pre- and post-deliberation questionnaires is available as a STATA file through the European Journal of Political Research’s website (hosted by Cambridge University Press).
Acknowledgements
This paper has been several years in the making, and I am especially grateful to two anonymous reviewers and to Nicole Curato for their excellent comments and suggestions, which have greatly improved the final version. I would also like to thank Şule Yaylacı and Anastasia Shesterinina for their immensely constructive feedback on earlier drafts. I am grateful to Vaishnavi Panchanadam and Erik Severson for their excellent research assistance. Thanks as well to Miriam Dajczgewand Świętek Milstein for being a sounding board in the early stages of writing. Finally, I owe a special debt of gratitude to Barbara Arneil, Mark Warren, and Fred Cutler, whose support was instrumental in carrying out the larger research project.
Funding statement
This research was partly funded by the Social Sciences and Humanities Research Council of Canada (Grant # 17W32308).
Competing interests
The author declares no conflicts of interest.
Open access
