Introduction
Deliberative democracy at its heart is concerned with connecting citizens to politics, but there is a legitimate concern that not all citizens can, will, or want to participate in deliberation. The ideal of inclusiveness in deliberative democracy (Dryzek, Reference Dryzek2009) is what grants legitimacy to political deliberation because all those affected should have the right to have their voices heard and every perspective should be equally considered (Dryzek, Reference Dryzek1990; Gutmann and Thompson, Reference Gutmann and Thompson1996). This demands that participation extends beyond the privileged and politically engaged, to marginalized citizens and minorities; and that different views and ideologies are represented in political deliberation (cf. Young, Reference Young2000; Dryzek and Niemeyer, Reference Dryzek and Niemeyer2008).
Thus far, research on willingness to deliberate reveals a mixed picture. Two studies suggest a low willingness to participate in political deliberation (Hibbing and Theiss-Morse, Reference Hibbing and Theiss-Morse2002; Mutz, Reference Mutz2006), while a third is more positive (Neblo et al., Reference Neblo, Esterling, Kennedy, Lazer and Sokhey2010). One factor that distinguishes the third (Neblo et al., Reference Neblo, Esterling, Kennedy, Lazer and Sokhey2010) study is that it offers a proposal for engaging in citizen deliberation under the auspices of a deliberative minipublic.
Minipublics are a subset of democratic innovations that involve participation of ordinary citizens in small-group deliberation applied to collective issues, with a view to influencing policy. Specific definitions of a minipublic vary considerably (see Ryan and Smith, Reference Ryan and Smith2014), but the particular form in focus here comprises randomly selected members of the public (Smith, Reference Smith2012) participating in a structured process that seeks to achieve high-quality deliberation (Ryan and Smith, Reference Ryan and Smith2014).Footnote 1
Mutz (Reference Mutz2006) suggests that it is this very type of public ‘cross-cutting’ deliberation that is unappealing to (United States) citizens, in contrast to Neblo et al.’s (Reference Neblo, Esterling, Kennedy, Lazer and Sokhey2010) more optimistic finding in respect to deliberation in a (hypothetical) minipublic. Studies in respect to recruitment for actual deliberative minipublics themselves reveal a mixed picture. Citizens appear willing to participate in broad terms (Smith, Reference Smith2012), although the extent of this enthusiasm is strongly contested (e.g. Jacquet, Reference Jacquet2017).
These conflicting findings may reflect an incomplete model of the factors that drive deliberative participation. In this paper I explore the question through the lens of personality, which is a well-established and useful construct for understanding a variety of behaviors, including political participation.Footnote 2 Its application yields the potential to build a more nuanced, finer-grained picture of deliberative participation, beyond broad categorical claims that are insensitive to context (e.g. Jacquet, Reference Jacquet2017).
Personality also provides a useful reference point for interrogating representativeness claims concerning minipublics, and the factors contributing to systematic under- or over-representation of particular categories. This includes the standard claim of descriptive (demographic representation), but also the claim of ‘discursive representation’ ensuring that all relevant discourses/viewpoints are adequately represented (Dryzek and Niemeyer, Reference Dryzek and Niemeyer2008).
Here I examine the impact of personality on willingness to participate in political deliberation by focusing on the interplay between personality and situational features, that is personality niching effects. Another aim is to contribute to a more thorough analysis of personality by adding individual variables to the analyses that are assumed to modify personality niching effects and to consider citizens’ motivations to accept or decline an invitation to participate in a deliberative minipublic.
The effect of personality on willingness to participate is demonstrated in the context of a deliberative minipublic (Uppsala Speaks) conducted in Sweden in February 2016, on the issue of begging by internal EU migrants within the Uppsala municipality, to which citizens were recruited via responses to a mail-based population survey.
The Swedish case study expands the range of existing research on political participation beyond the United States. The example also demonstrates the causes of differential participation across personality types, and potential remedies via mechanisms that structure the deliberative situation to encourage more widespread participation.
The article proceeds by first outlining the ideal of inclusiveness in deliberation. The theory of personality niching is then discussed, followed by a detailed presentation of the Uppsala Speaks case study. Discussion then turns to the situational properties of the case study. These are used in the first instance (1) to explore the relative effects of personality traits on citizens’ willingness to deliberate. This is followed by a demonstration of the utility of personality niching via analysis of (2) how personality interacts with individual characteristics and (3) expressed motivations to accept/decline the invitation that correspond to situational features. Finally, a general discussion considers the implications of personality niching for deliberative theory and practice.
Deliberation and inclusiveness
Deliberative democracy is supposed to be, at its core, inclusive – in addition to involving deliberative and consequential dimensions (Dryzek, Reference Dryzek2009).Footnote 3 Complete inclusion in deliberation is problematic, because the ideal of having all those who will be impacted by a political decision deliberate together is empirically impossible (Goodin, Reference Goodin2000). This means that deliberation in smaller groups must be able to draw on representativeness claims. In the case of minipublics – the predominant practical manifestation of deliberative ideals – the gold standard of representation has tended to be descriptive, such that a given assembly of citizens is demographically representative of the wider population (Mansbridge, Reference Mansbridge1999). More recent, is the influential argument that a deliberative forum should be ‘discursively constituted’ to represent all relevant discourses or arguments (Dryzek and Niemeyer, Reference Dryzek and Niemeyer2008). Whatever the established criteria, the extent to which certain kinds of citizens are predisposed to participate in deliberation – beyond well-established demographic explanations – is an important question, as is whether there are systematic factors that influence non-uniform participatory patterns.
Research on deliberation has identified patterns of participation based on individual-level characteristics – primarily demographic and political, but also motivational – which tend to distort representation (Mutz, Reference Mutz2006; Neblo et al., Reference Neblo, Esterling, Kennedy, Lazer and Sokhey2010; Karjalainen and Rapeli, Reference Karjalainen and Rapeli2015). There is a strong argument that participation is uneven, and those who do participate tend to be drawn from the politically privileged, which means that deliberation might have the unintended effect of amplifying prevailing political inequality instead of addressing it (Sanders, Reference Sanders1997; Young, Reference Young2000). Hence, the factors affecting the inclusiveness and representativeness of minipublics need to be understood.
Personality, representativeness, and implications for deliberation
Recent research demonstrates the importance of personality traits in explaining political participation more generally and alludes to their potential utility for understanding the issue of minipublic participation (Mutz, Reference Mutz2006; Mondak et al., Reference Mondak, Hibbing, Canache, Seligson and Anderson2010; Gerber et al., Reference Gerber, Huber, Doherty and Dowling2011a). It also demonstrates how personality ‘play[s] an important role in shaping the kinds of conversations citizens engage in and the setting for those conversations’ (Hibbing et al., Reference Hibbing, Ritchie and Anderson2011: 621).
The potential implications of personality-driven self-selection in minipublic deliberation are twofold. The first concerns implications for the goal of discursive representation, where all relevant political discourses (or attitudes) are ideally represented in a deliberative exchange (Dryzek and Niemeyer, Reference Dryzek and Niemeyer2008). Because personality (which is relatively stable) has been shown to be antecedent to political attitudes (Caprara et al., Reference Caprara, Schwartz, Capanna and Vecchione2006; Gerber et al., Reference Gerber, Huber, Doherty, Dowling and Ha2010), it offers a potential proxy for understanding the broader challenges for achieving discursive representativeness, and whether and why individuals holding certain positions may be less willing to represent them in deliberative exchanges. To the extent that personality and attitudes are correlated, it also circumvents the problem that Dryzek and Niemeyer (Reference Dryzek and Niemeyer2008) themselves contend with, where certain discourses are transformed out of existence by deliberation, bringing into question the legitimacy of their representation in the first place.
Second, personality stands to illuminate not only representative dynamics, but the related discursive dynamics – the manner in which deliberation proceeds. Although this is a relatively unexplored question in relation to deliberation, the strength of behavioral associations with personality from the wider personality literature suggest effects of uneven composition of personality on the dynamics of deliberation – which can affect, for example, levels of factors such as altruism and empathy or orientation toward practical or integrative solutions (Halfhill et al., Reference Halfhill, Sundström, Lahner, Calderone and Nielsen2005). It is this kind of dynamic that is related to deliberative pathologies, such as the so called ‘law of group polarization’ (Sunstein, Reference Sunstein2002), where outcome is a function of group composition. Improved understanding of self-selection based on personality helps to explain the deliberative dynamic as well as how process is related to outcome. These insights speak to the heart of a number of legitimacy questions associated with deliberation. And they are of critical importance in relation to deliberative design, in particular when it comes to facilitation strategies.
Personality niching: traits and situations
The ability of personality to predict participation operates via the interaction of personality and situation in shaping motivation. A useful approach for understanding the interplay between personality and situations is that of ‘personality niching’. Personality niching refers to individuals selectively choosing situations that evoke conduct in line with their dominant personality characteristic, while situations less attuned to one’s personality are avoided (Blass, Reference Blass1991; Asendorpf, Reference Asendorpf2009; Wagerman and Funder, Reference Wagerman and Funder2009). It refers to the interplay between personality and contextual factors via the activation of different motivations to participate (Blass, Reference Blass1991; Asendorpf, Reference Asendorpf2009). Niching helps illuminate the impact of factors, such as variation in institutional and cultural status of particular kinds of deliberative forums.
Accounting for situations – ranging from personal to institutional and cultural settings – contrasts with the tendency to draw universal claims from single country studies on deliberative participation, which (implicitly) ignore these situational differences. Just as there are likely to be differences across settings in relation to deliberative practices (Sass and Dryzek, Reference Sass and Dryzek2014), so too will there be differences in how practices are perceived and enacted (Goodin and Dryzek, Reference Goodin and Dryzek2006), and in how these settings translate into participation. Such variation means that a ‘one situational context fits all’ account of deliberative participation is unlikely to suffice – just as there is no ‘one size fits all’ in the application of deliberative theory to practice (e.g. Thompson, Reference Thompson2008).
This means, for instance, that deliberative events in the United States and Sweden are likely to result in different patterns of participation – even for the same personality type – because citizens’ understandings and expectations of what a minipublic is, their role in the democratic process, and the appreciation for that process will vary across different contexts. The task of both deliberative theory and practice is to adapt to a range of contexts – which, to date, has been lacking in the participatory dimension of deliberative theory.
To determine the impact of personality niching, we must first identify the most relevant domains of personality as well as the most important situational properties that can provide a structure for the analysis.
Personality traits
This study adopts the Five-Factor Model (FFM) of personality as it is widely recognized to capture the five most important characteristics of individuals’ behavior (Costa and McCrae, Reference Costa and McCrae1992). The FFM groups a wide array of personality traits into five main domains, and every domain is in turn composed of six sub-level facets. These five main trait domains include: Extraversion, Agreeableness, Conscientiousness, Emotional Stability, and Openness to Experience (a more detailed explanation of each trait domain is provided below).
The FFM was developed using a lexical approach (assessing the prevalence of adjectives about individuals’ behavior). And although the level of each personality type varies across cultures, the same five traits have been cross-culturally validated in explaining this variation (McCrae et al., Reference McCrae, Jang, Livesley, Riemann and Angleitner2001).
In contrast to other kinds of individual characteristics (e.g. attitudes, belief systems, social dominance orientation), traits are ‘core’ constructs, theorized to be causally prior to other individual features (McAdams and Pals, Reference McAdams and Pals2006) and predictive of social conduct (Costa and McCrae, Reference Costa and McCrae1992; Roberts et al., Reference Roberts, Wood and Caspi2008), including political participation.Footnote 4
Situational properties
Critical situational properties refer to those attributes of a situation that are ‘psychologically relevant to people in general’ (Wagerman and Funder, Reference Wagerman and Funder2009: 30). The theory of personality niching posits that certain vital situational properties determine both the strength of association between personality and motivation to participate in political deliberation, as well as the nature of that association – personality traits map onto particular kinds of motivation to participate, or not. For example, Jacquet (Reference Jacquet2017) describes the kinds of reasons why citizens might choose not to participate in political deliberation, without accounting for how pervasive these demotivations might be for which particular individuals under what circumstances – his own research spanning three different contexts that might serve to confound his own generalized results.
By contrast, a personality niching approach seeks to understand situational dynamics. Three general situational properties that are used to understand personality niching include degree of autonomy, situational strength, and self-awareness. Together these influence the overall relationship between personality and behavior, that is, how situations can leave more/less room for the influence/expression of personality differences (Blass, Reference Blass1991).
The first general situational property – degree of autonomy – concerns the extent to which behavioral norms are voluntary or externally imposed. At one extreme, compulsory participation nullifies the role of personality, while increasing discretion and autonomy enhance the niching effect.
The second general feature concerns situational ‘strength’. A strong situation involves clear social scripts and well-defined norms of acceptable conduct, which, in turn, reduces the niching effect (Asendorpf, Reference Asendorpf2009). For example, the act of voting usually involves a highly structured, strong, and clearly scripted process, where everyone behaves consistently irrespective of personality. Conversely, weak situations lack definitive scripts, leaving more room for personal differences to influence behavior (Blass, Reference Blass1991: 406). Thus, situations with low strength increase the scope for heterogeneous personality effects because there is greater room for interpretation by the individual in anticipating what the situation will involve.
The third and final general situational property concerns the extent that a particular situation forces individuals to focus on their own behavior and contributions to the situation. Self-focus decreases when the situation involves routinized, concrete activities that guide individual conduct (Blass, Reference Blass1991). For example, voting, in addition to being a strong situation, is both anonymous and formalized and thus tends not to increase self-focus because individuals tend to adhere to the routinized practices of the situation thereby mitigating the effect of personality niching. In contrast, situations structured so that individuals have to actively decide for themselves how to behave within the situation, strengthen the personality niching effect. For instance, formalized kinds of group political deliberation – which involve direct group contact – are spontaneous to the extent that discussion requires interaction and responsiveness, thus potentially dramatically increasing self-awareness and the scope for personality to influence behavior.
Overall, the effect of personality niching in the Swedish case study – Uppsala Speaks – is expected to be modest, but a discernable effect is anticipated. To the extent that niching does occur, it is enhanced by the voluntary nature of participation – which was explicitly stressed in the letter of invitation (provided in the Online Appendix) – and because deliberative participation is expected to lead to heightened self-awareness. Strongly counteracting these is the very weak situational strength of Uppsala Speaks due to the novel nature of minipublics in Sweden (Svensson, Reference Svensson2007). Taken together, the niching results are likely to be conservative compared to the effect when minipublics become increasingly routinized.Footnote 5
Personality niching in the case of Uppsala speaks
In addition to the general situational properties, three properties of specifically political situations influence the role of personality traits in determining motivation to participate. These properties include: civic and social norms, instrumental benefits and interpersonal dynamics (Gerber et al., Reference Gerber, Huber, Doherty, Dowling, Raso and Ha2011b). All three properties – which are explained in detail below – are relevant to the decision to participate in Uppsala Speaks.
The operation of these niching-inducing properties is fairly complex. And it is uneven: not all personality traits come into play for a given situational property (Mondak and Halperin, Reference Mondak and Halperin2008; Mondak et al., Reference Mondak, Hibbing, Canache, Seligson and Anderson2010; Hibbing et al., Reference Hibbing, Ritchie and Anderson2011). The objective here is not to develop a comprehensive account of their effects, but to demonstrate the utility of the theory of personality niching.
In the following discussion, the anticipated niching impact of each property associated with Uppsala Speaks, are considered in turn.
Civic norms: The effect of civic norms in personality niching involves the way they differentially shape conformity behavior among different traits. In the case of deliberative participation, the influence works to the extent that deliberation is viewed as a civic responsibility.
The aforementioned situational weakness of Uppsala Speaks likely mitigates this effect, but the topic of EU migrant begging covered by the event potentially induces the norm of duty. The issue was salient in wider public debate at the time, where street begging has grown exceptionally in Sweden as a result of EU immigration policies. It is a keenly contested issue, with the effect that a sense of social responsibility among respondents induces a civic norm, even in the absence of a norm for deliberative participation.Footnote 6
To the extent that civic norms operate in this case, the effect is expected to be differential according to personality trait. The trait most affected by the presence or absence of civic norms of political participation is Conscientiousness (Mondak et al., Reference Mondak, Hibbing, Canache, Seligson and Anderson2010; Gerber et al., Reference Gerber, Huber, Doherty and Dowling2012: 853). Inter alia, it is associated with an ability to delay gratification, as well as a tendency to follow norms and rules. This includes norms relating to civic duties – contributing a ‘fair share’ to the communal workload. While this suggests a tendency for Conscientiousness positively relates to political participation, the empirical evidence is very mixed. Some studies have shown a weak positive association, while others show a negative relationship (Mondak et al., Reference Mondak, Hibbing, Canache, Seligson and Anderson2010).
However, as mentioned, there are no such wider civic norms signaling that participation in an event such as Uppsala Speaks, are sanctioned behavior given its relatively novel nature in a Swedish context. Thus, a negative correlation between Conscientiousness and participation is expected (Mondak et al., Reference Mondak, Hibbing, Canache, Seligson and Anderson2010: 97).
Interpersonal relationships: Next, a specific property of political situations is the kind of interpersonal relationship a situation demands of its participants. Participation in politics is strongly influenced by the expectations about the interpersonal relationships and social dynamics the situation encourages. For deliberative participation, the most notable is the anxiety producing prospect of what Mutz (Reference Mutz2013), refers to as ‘exposure to cross-cutting’ deliberation, where conflict avoidance hinders participation.
Conflict avoidance in political settings is likely to be even more pronounced in Sweden, with a tendency for consensual political interaction avoiding interpersonal conflict (Rosenberg, Reference Rosenberg2002). This factor, combined with the weak situational strength of Uppsala Speaks, meant that it was important to clearly explain to those invited what was being asked, explaining the potential interpersonal dynamics. This was achieved via careful drafting of the invitation letter to prospective participants – mailed out to 4000 citizens living in the Uppsala region, Sweden (see Online Appendix) – to give a strong sense of the deliberative situation, and thus induce a more consistent niching response. The letter stressed the pro-social dimension (citizens coming together) and role of facilitators in managing the group dynamics (thus reducing potential for conflict).
Nevertheless, there remained scope for prospective participants to anticipate potential conflict. Previous studies show that the potential for interpersonal conflict and distress negatively impacts on participation by Agreeableness in political discussion (Mondak and Halperin, Reference Mondak and Halperin2008; Mondak, Reference Mondak2010; Gerber et al., Reference Gerber, Huber, Doherty and Dowling2011a) because of aversion to interpersonal tension (Antonioni, Reference Antonioni1998; Gerber et al., Reference Gerber, Huber, Doherty, Dowling, Raso and Ha2011b). Alternatively, Agreeableness entails a pro-social dimension – altruism, caring, and a positive outlook on others (Graziano et al., Reference Graziano, Habashi, Sheese and Tobin2007). The overall effect suggests ambivalence when it comes to political participation, with ‘a taste for participation and a possible distaste for politics’ (Mondak and Halperin, Reference Mondak and Halperin2008: 346). Given the controversial topic of Uppsala Speaks, reduced participation by agreeable individuals is expected.
A different dynamic is anticipated for Extraversion and Openness. Extraversion is positively associated with the social enjoyment of political participation (Mondak and Halperin, Reference Mondak and Halperin2008; Gerber et al., Reference Gerber, Huber, Doherty, Dowling, Raso and Ha2011b), the frequency of political discussion (Mondak and Halperin, Reference Mondak and Halperin2008), large social networks (Mondak et al., Reference Mondak, Hibbing, Canache, Seligson and Anderson2010), and a proclivity for political discussion, especially in formal contexts (Hibbing et al., Reference Hibbing, Ritchie and Anderson2011: 610). Similarly, Openness is associated with large social networks and a proclivity for interpersonal interactions (Mondak and Halperin, Reference Mondak and Halperin2008) – including political discourse and civic engagement (Vecchione and Caprara, Reference Vecchione and Caprara2009; Mondak, Reference Mondak2010; Gerber et al., Reference Gerber, Huber, Doherty, Dowling, Raso and Ha2011b). Extraversion and Openness are thus expected to be overrepresented.
Prior studies have found a positive association between Emotional Stability and participation in social and collective activities (Caspi et al., Reference Caspi, Chajut, Saporta and Beyth-Marom2006; Vecchione and Caprara, Reference Vecchione and Caprara2009) – explained by an ability to cope with stressful situations – including engagement in cross-cutting political discussion (Hibbing et al., Reference Hibbing, Ritchie and Anderson2011: 619; Gerber et al., Reference Gerber, Huber, Doherty and Dowling2012).
Thus, overall, the interpersonal dimension is expected to increase participation for Extraversion, Openness, and Emotional Stability, while the inverse is expected for Agreeableness.
Instrumental benefits: Instrumental benefits that motivate participation include both direct and indirect forms. Uppsala Speaks involved both, with a direct benefit in the form of a stipendiary payment of 2000 Swedish Kronor and indirect benefits in the form of an anticipated influence of citizens and/or decisions in the municipality they are engaged in, the process and the resulting ‘citizens report’ – which would also be distributed to the electors and more generally to the public – even though the discussions would not automatically translate into decisions.
The instrumental dimension most strongly affects conscientious individuals because they tend to be meticulous, goal-directed, and most sensitive to the opportunity costs of participating. Although, direct economic incentives are not expected to influence conscientious individuals, they are potentially motivated by the informal benefits (such as participation in a formal decision). However, given the vague prospect of influence an overall under-representation of conscientious individuals is expected.
The forgoing discussion implies the following hypotheses in respect to Uppsala Speaks and personality niching (expressed as overall effect by personality type, and associated niching-driven conduct):
Hypothesis 1 Extraversion is positively related to willingness to participate because of the anticipated interpersonal relationships.
Hypothesis 2 Agreeableness is negatively related to willingness to participate because of the type of interpersonal relationship.
Hypothesis 3 Conscientiousness is negatively related to willingness to participate because of lack of civic norms and (instrumental) influence.
Hypothesis 4 Emotional Stability is positively related to willingness to participate because of the anticipated interpersonal relationships.
Hypothesis 5 Openness is positively related to willingness to participate because of the anticipated interpersonal relationships.
These general effects are similar to those hypothesized by Gerber et al (Reference Gerber, Huber, Doherty, Dowling, Raso and Ha2011b), with respect to participation in local politics, particularly for the case of Extraversion being strongly associated with involvement in deliberative participation.
Personality measures and data collection
Data concerning willingness to participate in Uppsala Speaks and personality was collected as part of a recruitment survey. The survey, completed by both those willing and unwilling to participate, was administered by Statistics Sweden (SCB), mailed to 4000 households in the Uppsala Municipality and completed by ~1200 individuals – of which 200 accepted participation and 1000 whom did not. All respondents were offered a small payment for survey completion irrespective of whether they accepted participation.
Due to a number of research requirements the overall survey used to collect the data was comparatively demanding – 12 pages in length – including demographic, political, motivational, psychological dimensions, and attitudinal variables, along with items used for obtaining measures for the five personality traits. For this reason it was necessary to use a brief version of personality survey in the form of the Ten-Item Personality Inventory (TIPI; see Online Appendix, Table I).Footnote 7
In addition, register variables (obtained from the SCB) pertaining to age, gender, education level, income level, number of dependent children, and membership in political or civil society organizations have been used as control variables.
Results I: interplay of traits and situations
The effect of personality traits on citizens’ willingness to participate in Uppsala Speaks is summarized in Table 1. As expected, Extraversion and Openness are both positively associated with willingness to participate. Emotional Stability is positively associated, without attaining statistical significance. Conscientiousness is negatively associated, as expected. So is Agreeableness albeit without significance.Footnote 8
Trait effects on willingness to deliberate
Standard errors in parentheses.
All models include controls for age, gender, and residential municipality (by municipality dummies).
*P<0.10, **P<0.05, ***P<0.01.
Table 1 shows that personality explains over 5% of the variation in the acceptance of the invitation to participate in Uppsala Speaks. While not particularly strong, the results were expected to be conservative, given the comparatively weak situation of the case study (see above) and the potential attenuation effects of personality niching in responding to a relatively long survey. However, as will be seen, explanatory power increases once personality niching is accounted for.
It should be noted that there was some attrition from among the pool of participants who initially accepted the invitation, but this was relatively small (22 participants) and the withdrawals occurred at different phases. Timing varied from shortly after the survey to after attending part of the minipublic event. Analysis of this attrition was not possible for this data, although personality is likely to play a role for attrition in the same way as for recruitment.
Personality niching and individual variables
While personality helps explain participation, there are a number of individual variables that potentially interact with personality traits and modify the observed dynamics. In prior research on personality and participation, a number of individual variables assumed to impact willingness to deliberate have been noted.
Political Efficacy refers to an individual’s belief in the own capability to understand politics and assumed important to citizens’ willingness to engage in political participation. However, there are competing theoretical expectations to what extent it increases or decreases motivation to deliberate (Neblo et al., Reference Neblo, Esterling, Kennedy, Lazer and Sokhey2010: 572). Compared to other political settings, deliberation is often deemed to be a way to empower citizens hence it might also appeal to those who feel powerless and confused about politics. Adding personality to the study, patterns of general political participation has been linked to Political Efficacy, which in turn is related to Extraversion and Openness (Vecchione and Caprara, Reference Vecchione and Caprara2009). Translating these results to deliberative modes of participation, I expect extraverts’ enthusiasm and assertiveness in social encounters to be attenuated if they feel disempowered and confused in a particular (deliberative) situation. In other words, disempowered extraverts are expected to stay away from deliberation because they cannot act as assertively and passionately in discussions as they wish. Open individuals, by contrast are not constrained by submissiveness, and are accordingly interested in participating in deliberation regardless of Political Efficacy.
Need for Judgment is another individual-difference variable of relevance that describes an individual’s predilection for judgment and partisanship. It has been shown to positively relate to a (hypothetical) willingness to deliberate (Neblo et al., Reference Neblo, Esterling, Kennedy, Lazer and Sokhey2010: 574). Therefore, in the case of Uppsala Speaks, a correlation between Extraversion and Need for Judgment explaining individuals’ willingness to participate is anticipated. This is because when Extraversion is combined with a high Need for Judgment, individuals are likely to be opinionated and seek opportunities to air their views in public.
Mondak (Reference Mondak2010: 97) has argued that Conscientiousness combined with a particularly well-developed sense of Civic Duty increases participation because these individuals are driven by an urge to ‘do the right thing’. Accordingly, I hypothesize that Civic Duty interacts with Conscientiousness to overcome an otherwise negative view of participating in politics.
Finally, the negative relationship between low Emotional Stability (i.e. Neuroticism) and participation is expected to reverse when combined with high Opinion Strength. While, these individuals associate negative effects such as fear, hatred, and anger, with participating in social events, feeling strongly about an issue overcomes aversion to personal distress. This is supported by tentative evidence from another case study of this effect (Jennstål and Niemeyer, Reference Jennstål and Niemeyer2014) where an opportunity to express frustration in a small, safe, setting compared to established political modes, overrode potential aversion to the deliberative situation, which is also likely to apply to Uppsala Speaks.
Together, these observations inform the following hypotheses:
Hypothesis 6 There is an interaction between Political Efficacy, Extraversion, and willingness to deliberate due to a strong sense of empowerment and assertiveness.
Hypothesis 7 There is an interaction between Need for Judgment, Extraversion, and willingness to deliberate because of the anticipated social influence.
Hypothesis 8 There is an interaction between Civic Duty, Conscientiousness, and willingness to deliberate when deliberation is deemed a personal responsibility.
Hypothesis 9 There is an interaction between Opinion Strength, Emotional Stability, and willingness to deliberate because of the anticipated interpersonal relationships and social influence.
To account for these interactions the survey included the following variables: Political Efficacy, Need for Judgment, Civic Duty and Opinion Strength toward the issue (see Online Appendix for measurements).Footnote 9
Results II: interplay of traits, situations, and individual characteristics
The results of testing the four hypotheses above (Hypotheses 6–9) are shown via six regression models in Table 2. Regression models 2–5 testing each of the specifically hypothesized interactions in turn, with model 1 repeating the test for general effects and model 6 testing the combination of all effects together.
From Table 2 it is noteworthy that none of the variables (Political Efficacy, Need for Judgment, Civic Duty, or Opinion Strength) are significant predictors of participation on their own. However, as predicted, when combined with a personality trait, interaction effects can be observed in the case of Extraversion and Need for Judgment (models 3 and 6), Conscientiousness and Civic Duty (4 and 6), while the expected interactions between Extraversion and Political Efficacy (3, 6), and Emotional Stability and Opinion Strength were not observed.Footnote 10
Interaction effects between traits and individual-difference variables on willingness to deliberate
Standard errors in parentheses.
All models include controls for age, gender, and residential municipality (by municipality dummies).
* P<0.10, ** P<0.05, *** P<0.01.
The finding of the interaction between Extraversion and Need for Judgment in explaining willingness to participate confirms Neblo et al.’s (Reference Neblo, Esterling, Kennedy, Lazer and Sokhey2010) prior finding for Need for Judgment. But it also sheds new light on why extraverts seek out deliberation. The trait domain of Extraversion involves two aspects, enthusiasm and assertiveness, and depending on what side dominates one can expect slightly different drives to participate (DeYoung et al., Reference DeYoung, Quilty and Peterson2007: 883). The findings here suggest that it is the latter aspect–assertiveness–that is the driving force behind extraverts’ willingness to participate. While enthusiasm is linked to friendliness and sociability, the aspect of assertiveness is conceptually closer to Need for Judgment, both involving a desire to take a strong stand.
Another notable finding is the relationship between Conscientiousness and Civic Duty, which shows how Conscientiousness inverts where there is a high level of Civic Duty. In other words, when conscientious individuals exhibit a belief in citizens’ responsibility in politics, the negative effect of Conscientiousness on willingness to participate dissipates. This demonstrates how Conscientiousness operates in concert with an individual’s attitudes about civic duties.
Finally, model 6 combines all possible combinations of individual traits and intersections with personality. Notably, the overall explanatory power of the model (just below 10%) is almost double that of personality alone (model 1). Thus, accounting for interaction between personality and other individual characteristics appears to improve understanding of who is willing to engage in deliberation.
Personality niching and motivational drivers
A major benefit of personality niching lies in better understanding the specific motivations that drive participation or refusal to participate in deliberation, which directly inform strategies for correcting systematic under-representation. The results thus far demonstrate how traits influence willingness to participate, and how certain individual variables modify these effects. But to fully understand why and how personality influences individuals’ response to the invitation, there is a need to look at specific motivational reasons.
In other words, if personality matters it should be evident not only in individuals’ decision to participate but also in how they explain their decision.
The relationship between motivations to accept/refuse participation and personality was tested via 12 motivational items included in the recruitment survey (the items are listed in the Online Appendix, Table IV).Footnote 11 The results in this section pertain to two groups of respondents – those who accepted the invitation to participate (200) and those who declined (1000) – with each group indicating their motivating/demotivating factors for their decision.
This analysis is more exploratory than the preceding ones. As such the following discussion does not develop testable hypotheses. Table 3 summarizes the results of the exploratory analysis (reasons to accept are shown in bold, and to refuse in italics) according to three types of niching effect described above (civic norms, interpersonal relationship and instrumental dimensions). Overall, Table 3 shows considerable variation in the ability of specific motivations to explain personality niching. Most prominent is the expectation of ‘enjoyment’, which is strongly associated with three traits (Extraversion, Openness, and Emotional Stability) and negatively associated with Conscientiousness, explaining 19% of variation. The results are broadly consistent with the forgoing discussion in relation to personality niching, which I will demonstrate by considering the effect for each personality trait in turn.
Trait effects on motivations to accept and refuse to deliberateFootnote a
Standard errors in parentheses
a Motivations to participate are in plain type. Motivations to refuse are shown in italic
Extraversion
The results in Table 3 are consistent with the general hypotheses in relation to Extraversion indicating a willingness to participate. Extraverts like social situations and would be motivated to participate simply by the sheer ‘enjoyment’ of participating (Mondak and Halperin, Reference Mondak and Halperin2008; Gerber et al., Reference Gerber, Huber, Doherty, Dowling, Raso and Ha2011b). Even those extraverts who are not particularly instrumentally driven to influence outcomes will have a disposition toward exerting influence over others (Mondak and Halperin, Reference Mondak and Halperin2008: 357). Strong self-esteem and a sense of efficacy in political discussions mean that they unlikely to refuse because they feel ‘it is impossible to discuss’ the issue with others, or due to ‘lack of knowledge’. On the other hand, they are attracted to the prospect of ‘influencing peers’.
Table 3 shows results for Extraversion, which are broadly consistent with what is expected: the pro-social dimension motivates participation because of ‘enjoyment’. Notable is how extraverts who abstain give the reason ‘lack of time’, while they view all other motives as irrelevant, that is it is not because they are troubled by self-doubts in respect to knowledge about the issue, nor because they are ‘uninterested in topic’ (Table 3). This could be interpreted as an acknowledgment of social norms promoting urgency of the topic, which is constituent with a ‘lack of time’ dominating non-participation.
Less expected is that the extraverted appears motivated by ‘influencing politics’. This is possibly due to interaction with Need for Judgment where there is a tendency to hold strong opinions. However, this requires further confirmation and analysis.
Agreeableness
While Agreeableness appears thus far not to play a role in deliberative participation (see Table 1), I have anticipated that the prospect of conflict is a strong demotivating factor for agreeable individuals. This may help explain the result in Table 3, where the prospect of interpersonal conflict is expressed in the observed demotivational association with ‘lack of knowledge’ enhancing the effect of anxiety. Again, this observation would need replication and more detailed assessment as part of future studies.
Conscientiousness
Overall, as expected, Conscientiousness is associated with lower deliberative participation (Table 1). And the results in Table 3 are consistent with the hypothesized reasons why. Conscientious individuals, being meticulous and goal-directed are not supposed to be motivated by the ‘enjoyment’ of participating nor value it as a ‘learning opportunity’ – although the latter association is not significant.
The instrumental dimension is, in theory, much more important and the prospect of making a concrete impact on politics potentially could be a motivation. But the lack of formal influence of this particular deliberative event mitigates this effect, and it shows in the results. A consistent outcome is that the ‘payment’ is not a motivator in cases where non-instrumental goals predominate. For similar reasons, abstention should be explained by normative standpoints (such as ‘politics is private’ or the ‘irrelevance of the topic’), although the results do not support this conclusion. Looking at Table 3, the anticipated demotivational dimensions of being ‘uninterested in the topic’ or ‘politics is private’ are not observed, but niching does occur. If a busy conscientious individual is going to participate, then ‘enjoyment’ is not a motivating factor. The instrumental dimension did feature, as expected, albeit only weakly via the demotivational dimension of ‘already knows what should be done’ (Table 3).
Emotional stability
An easy-going style, unhindered by anxiety of interpersonal settings or concern about the costs (or potential instrumental benefits) should predict higher participation by the emotionally stable, but no overall general effect was found (Table 1). Nevertheless, anticipated effects in relation to expressed motivations were observed. Strongest of these should be an attraction to potential ‘enjoyment’, consistent with Hibbing et al.’s finding that ‘emotional stability helps individuals overcome a natural tendency towards discussing politics [only] with likeminded individuals and leads to more conversations with casual acquaintances and individuals holding differing viewpoints’ (Reference Hibbing, Ritchie and Anderson2011: 619). And this is indeed the case in the findings in Table 3. Also anticipated was an inverse demotivating effect of ‘impossible to discuss’ where emotionally stable individuals are not troubled by the interpersonal relationships. This is also consistent with (although not anticipated) negative relationship with ‘payment’, which is consistent with the generally easy-going nature of emotionally stable individuals to not be instrumentally driven.
There are two unanticipated findings in relation to Emotional Stability. Although, the domain of civic norms is not expected to feature in respect to Emotional Stability, the principle ‘politics is a private’ was strongly rejected as a reason to abstain from participation. Also, as for extraverted individuals, time scarcity emerges as a main reason to why emotionally stable individuals refuse to participate. I will return to these findings below.
Openness
Openness was expected and confirmed as positively related to participation in deliberation, anticipated to be driven by a curious, creative disposition. The motivation of deliberation as a ‘learning opportunity’ was both expected and observed, as was the expected ‘enjoyment’ of participation. The converse motivations of ‘lack of knowledge’ or ‘impossible to discuss’ with others were negatively associated with the choice to refuse, as anticipated (Table 3).
Neither civic norms nor instrumental benefits were expected to influence open individuals’ participation. Nevertheless, those who did participate were negatively associated with motivation to ‘influence others’ and those who did not participate were not demotivated by ‘already knowing what to do’. Also unanticipated is the negative role of ‘politics is private’ in demotivating participation, but it is consistent with the overall pragmatic nature of Openness.
Discussion
The haste in contemporary scholarship in deliberative participation to claim sweeping generalizable findings misses a good deal of variation and important nuances in describing the motivations and the role of particular situations and circumstances in participatory dynamics. And to the extent that the broad findings asserting non-participation hold they do not account for all possible worlds (or situations) motivating greater deliberative participation, nor inform mechanisms for realizing that world – which will vary depending on particular situational features that interact with personality type. The results above demonstrate the utility of taking into account the context and design features that differentially motivate citizens to participate according to personality type.
A notable feature of the findings for those personality traits positively disposed to participation – Extraversion, Openness, and Emotional Stability – is that they are also negatively disposed to the proposition that ‘politics is private’ in their participatory assessments. These individuals do not decline to participate because they view ‘politics as private’. Conversely, the remaining two traits – Agreeableness and Conscientiousness, which are overall negatively related to participation – exhibit a positive relationship to the ‘politics is private’ proposition.
If we accept the proposition that the item ‘politics is private’ captures a civic norm, it indicates that Extraversion, Openness, and Emotional Stability are relatively unaffected by the norm, while Agreeableness and Conscientiousness associates with recognition that privacy in respect to politics is important. This supports the assertion that traits are differentially sensitive to surrounding norms. Agreeable and conscientious individuals seem to be inclined to provide a more normative argument for their decision to abstain, while instrumental motivations (‘lack of time’) best predict why extraverted and emotionally stable individuals chose to decline participation. It makes sense that agreeable and conscious individuals are the most sensitive to surrounding norms – albeit for different reasons, where agreeable individuals are concerned with getting along with others, conscientious individuals want to do the right thing.
Given that Extraversion, Openness, and (to a lesser extent) Emotional Stability also predict political participation in other contexts, mainly the United States (e.g. Mondak et al., Reference Mondak, Hibbing, Canache, Seligson and Anderson2010; Gerber et al., Reference Gerber, Huber, Doherty, Dowling, Raso and Ha2011b), this suggests generalizability of these specific effects; that certain dispositions are less sensitive to context than others – confirmation of which will require further research.
The results have implications for the design of minipublics in respect to participation. Sensitivity to the time commitment for otherwise motivated individuals (emotionally stable as well as extraverted) in particular has implications. In the case of Uppsala Speaks – an event lasting for 3 full days – some hesitation was expected. This was indeed the case with an acceptance rate for participation of 5%. However, it should be noted that in other settings, such as Australia, citizens seems to be more inclined to engage in demanding political deliberation lasting 3–4 days with an acceptance rate upward of 30% not being uncommon (Lubensky and Carson, Reference Lubensky and Carson2013). The aforementioned novelty of deliberative events in Sweden likely contributed to this difference, although confirmation of this effect also requires more dedicated research.
What can be stated with some degree of confidence is the differential impact of personality on participation and the interaction with features of Uppsala Speaks, including its novelty and time demands. Agreeable and conscientious individuals were expected to show a relatively low interest in participating in this novel, vaguely defined and potentially conflict-laden setting – for reasons that vary according to expectations based on personality niching. Agreeableness was related to demotivation because of ‘lack of knowledge’, which, ideally, should not be a factor. The normalization of deliberative events could correct this phenomenon, enhancing, and promoting their reputation for being enjoyable and empowering (Curato and Niemeyer, Reference Curato and Niemeyer2013), thus neutralizing this effect.
Routinized deliberation also potentially neutralizes the tendency for conscientious individuals to refuse participation via a sense of responsibility and complicity with civic norms. And such routinization would also potentially neutralize the belief that they ‘already know’ what should be done – which is otherwise deeply problematic from a democratic perspective irrespective of their actual participation in minipublics – by demonstrating emerging discourses that are seen as publically valuable (MacKenzie and Warren, Reference MacKenzie and Warren2012).
In sum, context counts and it is very possible – if not likely – that the normalization of deliberative events stands to increase participation by both these types of individuals by strengthening the situational characteristics – something that certainly occurs when there has been direct experience with them (Curato and Niemeyer, Reference Curato and Niemeyer2013).
Personality niching also has implications for the development of targeted recruitment strategies beyond the demographic approaches to overcome gaps in participation. If we accept that inclusiveness in minipublics means to give voice to a variety of perspectives and attitudes (Dryzek and Niemeyer, Reference Dryzek and Niemeyer2008), then personality niching stands to provide an avenue to improving participation among different perspectives. One strategy to deal with the possible implications of personality niching could be via the specific framing of the invitation to participate in minipublics. Depending on the wording and the dimensions of deliberation that is emphasized, it is possible to target certain (underrepresented) personality types.
Conclusions
The assertion that personality plays a role in deliberative participation appears to be both valid as well as having practical implications for deliberative design. Personality is a predictor of deliberative participation in Uppsala Speaks in a manner broadly consistent with expectations. Additionally, that this effect varies according to critical situational features that differentially attract citizens according to personality type reinforces the importance of the approach and implications for design. Development of greater knowledge about these personality-based effects will serve to inform both scholars and practitioners about design choices in respect to recruiting for minipublics as well as facilitating wider involvement in political deliberation.
Taking seriously the role of personality as a predictor of political attitudes also serves to illuminate – where necessary, remediate – existing strategies for achieving ‘discursive representation’ in deliberative minipublics (Dryzek and Niemeyer, Reference Dryzek and Niemeyer2008). In the absence of a turn toward taking personality seriously in respect to deliberative participation (as for political participation more broadly) there is a risk of ‘creating a politically engaged citizenry whose views are not representative of the broader public’ (Gerber et al., Reference Gerber, Huber, Doherty, Dowling, Raso and Ha2011b: 704). The representative claims of particular deliberative designs will become increasingly difficult to sustain.
There are also important implications for deliberative dynamics, where certain personality types dominate deliberative interactions – in this study extraverts and open individuals – contributing to certain group effects and impacting the deliberativeness of outcomes.
This article represents an early contribution to understanding the role of personality in deliberative participation. There is considerable scope for more research that focuses on the interplay between attitudes, personality, and deliberative participation, accounting for the dynamics of personality niching. The same also applies to understanding the role of personality group composition and the process of dynamic and deliberative outcomes.
Short of these lofty goals, the key objective here has been to demonstrate the utility of personality niching for understanding and improving the dynamics of deliberative participation. These effects have been demonstrated in respect to Uppsala Speaks, despite involving a weak situation. It is arguable that the niching effects would be considerably stronger where deliberative participation is more common, or were it to become the case. Another important factor, which cannot be fully accounted for here in a single study, is that of cultural context and its likely impact on niching effects. Understanding these dynamics requires future comparative research, for example, comparing more directly the Swedish context to the situational characteristics of deliberative events taking place in other locations.
Nevertheless, despite the early stages of this research, this study demonstrates that a finer-grained analysis that is sensitive to both the whole suite of personality types and the particular features of deliberative settings needs to be taken into account when exploring the possibilities for improving and/or institutionalizing deliberative participation.
Acknowledgments
The author is grateful to Dr Simon Niemeyer for detailed recommendations offered. Also, the author wants to thank Marcus Österman who has contributed to the analysis of the data as a research assistant. The Swedish Research Council has funded the research.
Supplementary material
To view supplementary material for this article, please visit https://doi.org/10.1017/S1755773918000024
