Abstract
This research examined the role that group dynamics played in the 2016 US presidential election. Just prior to the election, participants were assessed on perceived self-similarity to group members’ views, perception of own leader’s prototypicality, perceptions of social values, and strength of support (attitudes). Results indicated that Democrats were more inclusive, seeing more similarity between themselves and members from the outgroup political party, while Republicans displayed more ingroup solidarity and negative attitudes toward outgroup members. Trump was viewed as a more prototypical leader by Republicans than Clinton was by Democrats. These results may help to explain the perhaps surprising fragility of Democrat voters’ support for Clinton.
Introduction
On the face of it, the 2016 US presidential election might have looked like an “upset” victory for Donald Trump. However, it is possible to see that Trump and the Republicans would clearly win, and that the Democrats and Clinton would lose, if viewed from a social psychological perspective. If we approach the election as a contest between groups, with competing factions, and if we carefully consider group dynamics, it may not be so surprising after all.
This research reports on a survey of likely voters conducted in the week leading up to the US presidential election of 2016. Using this sample of voters from both political parties, we sought to examine the prospect of a “winner-takes-all scenario” on the social identities of respondents’ own social group (Republican or Democrat), their perceptions of members of the other group/party, and their respective leadership candidates. In viewing the election as a contest between social groups, in addition to examining the perceived similarity the members had with these social groups, we may better understand the outcome of Trump’s election.
We relied on the group dynamics literature, specifically social identity theory (SIT), (Hogg and Abrams, 1988; Tajfel and Turner, 1979), as well as the values/belief literature (Rokeach, 1973). According to SIT, one reason why people join social groups, Republican or Democrat in this case, is because groups provide people with a sense of “belongingness” and offer them opportunities to draw both personal and collective esteem (Abrams and Hogg, 1988). Moreover, in the classical statement of SIT (Tajfel and Turner, 1979; see Hogg and Abrams, 1988), it is proposed that groups contribute to their members’ social identity by being positively distinctive from outgroups and by winning either objective or social competitions with these outgroups. However, the theory does not directly address situations in which the goal of the competition is the right to represent the individual members of both groups, and it does not consider how this might affect the victor’s perspective on the intergroup relationship. Within the context of the US election, and because of the increasing polarization of the two political parties, both Republican and Democrat groups were highly salient and thus would offer people a powerful group membership and a sense of belonging. SIT would expect that when identification is strong, members should report strong similarity between their own views and those of members of their ingroup. Conversely, members should feel very little connection to outgroup members of the other political party.
However, within the context of a competition in which the future of both groups is uncertain, and because of the ups and downs of the conflict, it could be very difficult for people to be assured of the stability of either group. This might affect the way in which members perceive others from their ingroup. To this end, Haslam and Reicher (2007), argue that group members are motivated to take advantage of opportunities to act as “identity entrepreneurs.” That is, they are motivated to psychologically incorporate the outgroup as part of their ingroup identity, particularly if they feel that their own group truly represents the best interests of both groups, as a superordinate whole. In other words, they see their ingroup beliefs as more “justified,” and valid (cf. Marques et al., 2001), and are likely to engage in ingroup-projection, whereby group members assume that their group is representative of the wider American culture (see Wenzel et al., 2008). While this strategy has a direct benefit for the individual—boosting the member’s sense of personal worth—it may come with high costs for the collective by eroding ingroup solidarity in relation to the original ingroup as members become more open to identifying with the outgroup.
Another reason why people adopt political affiliations is to reflect their personal values (Rokeach, 1973). Values capture enduring interpersonal orientations and principles (see Schwartz, 2012). While a number of values dimensions have been proposed over the years, several recent studies indicate that these reduce to two orthogonal dimensions (see Duckitt, 2001, for a review; see also Braithwaite, 1994; Rokeach, 1973). In the area of moral psychology, Haidt and Graham (2007) characterize these two values dimensions as “individualizing,” or an approach that focuses on individuals as the locus, and “binding,” an approach that focuses on the group as the locus. Relevant to the election, individualizing has been associated with liberal ingroups, and binding has been associated with conservative ingroups (also see Conover and Feldman, 1981; Haidt and Joseph, 2004). This mirrors earlier work in which liberal ingroup values were more likely to reflect equality and an openness to change, while conservative ingroup values were more likely to include security and a desire for order and uncertainty reduction (Jost, 2017; Jost and Napier, 2012). Taken together, the pattern emerging is that liberal ingroups are more likely to be characterized as endorsing “equality” and “individualized,” or idealism-centered values, whereas conservative ingroups are more likely to value “support for security and order," with the group focused on relativist-centered behavior, or behaviors that seek to protect the ingroup relative to other outgroups (see Graham et al., 2013; McHoskey, 1996).
Closer to the domain of the present research, studies bridging interpersonal values (within the moral foundations work) and group processes have demonstrated that values mediate the strategies used by ingroups (Morris, 2018; Stewart and Morris, 2018). Stewart and Morris (2018), for example, examined the relationship between ingroup values of liberals and conservatives and their intergroup bias toward immigrants (the outgroup was not the opposing social group) in the UK. In a series of studies, the researchers found that liberal ingroup values of individualizing directed towards the outgroup mediated intergroup bias, helping to suppress the negative views of the immigrants. Conversely, however, conservative ingroup values magnified the difference between the ingroup and the outgroup thereby increasing intergroup bias against the immigrant outgroups (also see, Sparkman and Eidelman, 2016; Meertens and Pettigrew, 1997).
While the work of Stewart and Morris offers a concrete step forward, it also has some interesting limitations. For example, the strength of ingroup and outgroup similarity were not directly measured. Also,the research was not framed within the context of an intergroup competition between liberals and conservatives. Finally, and importantly, it did not connect well to the well-established body of literature examining social identity and leadership, thereby overlooking the significant identity-entrepreneurial role of the leader as a member of the social group (Abrams et al., 2018; Hogg, 2001; Reicher et al., 2005; Van Knippenberg et al., 2004).
Measuring the linkages between the individual and the group is important, because it would provide insights into whether voting preferences more strongly reflect values preference or identity needs. If the key factor is values, then we would expect that the main basis for aligning with either the Democrats or Republicans is rooted in the evaluation of others as having a similar set of beliefs to those held by the self. If the key factor is group identity, then perceptions, values, and motives would be drawn from the group’s identity and would be a part of the construal of the identification, meaning that we could expect values to differentiate between the groups (Duckitt, 2001; Graham et al., 2013). This would be detected by assessing similarity with other ingroup members and, subsequently, self-prototypicality. A central question for social identity theory would center on the extent to which the individual internalized the beliefs of the group and thus reported a stronger sense of being “representative” for the group. For example, if self-prototypicality and perceived leader prototypicality are both high, this would tend to indicate that these individuals would support the leader, whereas if one or both is lacking, then the leader may be more vulnerable to voter apathy or disengagement. Thus, the inter-relationships between values, perceived similarity to other members and one’s sense of representativeness may have a direct effect on support for leadership candidates and contest outcomes.
To answer these questions, we conducted a survey with a sample of US citizens in the week leading up to the US presidential election of 2016. We looked at the differences between Republican and Democrat supporters in the associations between perceived values, self-group similarity, and self and leader prototypicality perceptions of one’s own social group/political party. We proposed a number of hypotheses:
H1: Value congruence. According to SIT, both Republicans and Democrats will report strong similarity with other ingroup members and low similarity in views with the opposing outgroup (i.e., Democrat or Republican depending on the target, respectively). H2: Group distinctiveness and self-group similarity. Consistent with a perceived congruence (Rokeach, 1973), we would expect that ingroup values could motivate the formation of social groups, and therefore, social values should differentiate between social groups. That is, we would anticipate that the pattern of particular values that are perceived to best represent Republicans would be different from those that are perceived to best represent Democrats. However, the strength with which the key values are associated with their respective groups should not differ. Also, if values congruence is a key reason for group membership, then we would also expect that respondents who perceived their ingroup values as more distinctive would also perceive greater similarity between their own and ingroup views. H3: Strategic identification. Consistent with a strategic identity hypothesis (Marques et al., 2001), ingroup values will correlate with ingroup perceived similarity where the group highly identifies with the ingroup construct. However, we reasoned that Democrats’ more overarching values may reflect their motivation to accommodate outgroup members, as suggested by Haslam and Reicher (2007). Therefore, Democrats may strategically envisage stronger perceived self-similarity with outgroup members. However, this strategy also implies lower self-ingroup prototypicality, because these more inclusive values imply a less distinctive ingroup prototype. H4: Leader prototypicality. Turning to the relationship between leader perceptions, leader prototypicality should affect leaders who are perceived as more prototypical and evaluated more positively by group members. Although Trump may be regarded as highly idiosyncratic in many ways (such as his prior career path and media coverage), and Clinton much more conventional because of political background, these characteristics do not necessarily imply that they would be viewed as prototypical of their parties. Indeed, based on their values and capacity to capture the central goals of their parties, it is plausible that Trump could be regarded as more highly prototypical among Republicans than Clinton is among Democrats. Moreover, because we contend that Republicans are more likely to pursue a path of distinctiveness and differentiation from the outgroup, the greater distinctiveness of Trump would contribute more strongly to perceptions of his prototypicality. Clinton’s more inclusive approach, perhaps ironically, would make it harder for Democrat supporters to view her as a distinctive, and thus prototypical, leader.
1
In sum, adopting an intergroup perspective allow us to answer critical questions about the importance of the group and its values in the competition for leadership. The group context is often overlooked, but it may be critical in understanding the outcome of this and future elections.
Method
Participants
Two hundred and ninety-nine participants were recruited through Amazon’s social research platform, MTurk. The mean age was 34.52 years (SD = 10.06 years). The sample was largely comprised of single (86.2) people reporting to be educated to at least a high-school diploma level or greater (86.2%), and approximately half were Caucasian, men, and owned their own homes. Of the participants, 51% were reportedly Democrats, while 49% indicated that they were Republicans, and 77.2% had voted in the last election. Each participant was paid $3 for his/her participation.
Procedure
Participants responded to an advertisement asking for supporters of either candidate: Clinton or Trump. They were told the study assessed their social perceptions and that they would be asked to comment on the values of both Democratic and Republican social groups. Next, procedures about the confidentiality, data storage, and those of withdrawal from the study were explained. Written informed consent was collected. Participants indicated which candidate they planned to vote for, responded to voting behavior questions about the candidate they were supporting, and were asked to complete a series of measures assessing group perceptions. Upon completion, participants were thanked and debriefed.
Measures
Perceived similarity of group members’ views, perception of leader prototypicality, perceptions of social values, and strength of support (attitudes) were assessed using the following items:
Similarity between own views and those of ingroup members, and those of outgroup members’ views. Single item measures were used, assessing the extent to which the participants viewed themselves as similar to the ingroup and the outgroup (Abrams and Hogg, 1990). We asked, “How similar are your views to other Clinton/Trump supporters?” and with the outgroup, “How similar are your views to the views of other Trump/Clinton supporters?,” scored using a Likert type scale ranging from 1 (not at all) to 5 (very much). Ingroup and outgroup values. Drawn from Rokeach’s (1973) framework, we measured 18 terminal values (i.e., a desirable end state) and 18 instrumental values (i.e., preferable modes of behavior). Participants were asked to complete this process twice, once to rate the extent to which they perceived the values to be held by the ingroup (both terminal and instrumental values), and then to rate how much the same values are held by the other group (ingroup/outgroup order was randomized). Responses were recoded using five-point Likert-style scales ranging from 1 (not at all) to 5 (very much). Own leader’s prototypicality. The prototypicality of the leader (candidate corresponding to the Party) was assessed by asking: “To what extent do you think that the candidate embodies this group’s norms?,” scored on a five-point Likert scale from 1 (not at all) to 5 (very much). Ingroup self-prototypicality. Group members’ prototypicality for their own party (Van Knippenberg, 2011) was assessed by asking: “As a prototypical member of the [Democrat/Republican as appropriate] party, I represent the interests and opinions of the group well,” scored on a five-point scale from not at all to very much. Strength of support for own candidate. Strength of support for the candidate was captured using a scale ranging from 0 (lowest level of support) to 100 (highest level of support), with every 10% on the scale being indicated. Prior voting behavior. Past behavior was measured using a single item: “Did you vote in the last (general) election?” All items were scored using binary coding: 0 (no) and 1 (yes). Sociodemographic characteristics. Participants were asked to indicate their age, gender, ethnicity, educational achievement, marital status, and home ownership, using open ended questions. Data were coded for analyses. (See the footnote of Table 1.)
Results
Preliminary analyses
To examine whether there were significant differences in socio-demographic characteristics profiles by group membership, we recoded membership to form a continuous variable and performed an analysis of variance (ANOVA) (see Table 1). 2 Consistent with the literature (Taylor, Funk, & Craighill, 2006; Cohn, 2018; Dimock et al., 2014), there were three main significant differences between Democrat and Republican participants with regard to both age and homeownership status (see Table 1). Democrats were significantly younger than Republicans (F1, 268 = 13.013, p = 0.0001; MDemocrat = 31.98, SD = 8.82; MRepublican = 36.24, SD = 10.29), and Republicans were more likely to report owning their own homes (F1, 268 = 12.440 p = 0.0001; MDemocrat = .37 or 36.8%; MRepublican = .58 or 56.2%). Turning to ethnicity and marital status, Democrat voters were less likely to be Caucasian or to be married than their Republican counterparts. The distributions indicated that there was not a great deal of variability within the samples, however.
Effects of sociodemographic variables and group membership.
Note: Group membership was recoded (0) non-Democrat, (1) Democrat; gender: (0) non-female (1) female (percentage reported reflects proportion of women); ethnicity: (0) non-Caucasian (1) Caucasian (percentage reported reflects proportion of Caucasian participants); education: (0) not reporting college education (1) reporting college educated (percentage reported reflects proportion of those reporting college education including Masters and PhD degrees); marital status: (0) non-married (including, single, divorced but not cohabitating) (1) married (percentage reported reflects married participants); age, home ownership, and behavior were continuous variables prior to analysis.
*p < .05; **p < .001; ***p < 0.0001.
Role of values
The means for each of the terminal and instrumental values attributed for each group were ranked (see Table 2). While there was some overlap in the items selected by each group (freedom and family security selected by each social group and therefore excluded from this analysis), there was also considerable divergence. For Democrats, the highest ranked ingroup values were “equality,” “broadmindedness,” “helpful,” “logical,” “happiness,” and “world at peace”; and for Republicans, the highest ranked were “national security,” “sense of accomplishment,” “self-respect,” “ambition,” “responsibility,” and “independent.” We used these two sets to represent “divergent” values for the two groups.
Descriptives for values, Democrats.
Descriptives for values, Republicans.
Next, to examine the mean outcomes on the divergent individual values items by group, we created four composite variables using the three most divergent value variables for each category: Democrat ingroup instrumental, Democrat ingroup terminal, Republican ingroup instrumental, and Republican ingroup terminal. Following the creation of the indices, we tested whether their means differentiated between the groups, with significant differences indicating divergence in values/beliefs between the Democrat and Republican participants.
The values significantly differed for three of the four indices (Democrat ingroup terminal values: F1, 268 = 15.56, p = 0.0001; MDemocrat= 4.11, MRepublican = 3.73; Republican ingroup instrumental values: F1, 268 = 17.05, p = 0.0001; MDemocrat= 3.89, MRepublican = 4.29; Republican ingroup terminal values: F1, 268 = 20.70, p = 0.0001; MDemocrat= 3.75, MRepublican = 4.31). 3 Only ingroup instrumental values for the Democrats were evaluated similarly by both Republicans and Democrats (F1, 268 = 1.40, p = 0.183; MDemocrat= 3.97, MRepublican = 3.73).
To better understand the relationships between these values and self-ingroup similarity, self-outgroup similarity, and self-ingroup prototypicality, we also examined the correlations for Democrats and Republicans separately. We controlled for prior voting behavior and the strength of the support for the candidate, because both variables might systematically influence ingroup values. 4 Therefore, we used a partial correlational analysis procedure. Among Democrats, outgroup similarity judgments were correlated significantly with perceptions of ingroup values (ingroup terminal (D): pr = 0.45, p < .01; ingroup instrumental (D): pr = 0.44, p < .01; ingroup terminal (R): pr = .43, p < .01; ingroup instrumental (R): pr = .44, p < .01). In contrast, among Republicans, ingroup similarity judgments were correlated significantly with perceptions of ingroup values (ingroup terminal (D): pr = 0.39, p < .01; ingroup instrumental (D): pr = 0.33, p < .01; ingroup terminal (R): pr = 0.31, p < .01; ingroup instrumental (R): pr = 0.39, p < .01), and Democrat instrumental and terminal values were negatively correlated with Republicans’ perceived ingroup similarity (outgroup terminal: pr = −0.16, p < .01; outgroup instrumental: pr = −0.23, p < .001), consistent with our values congruence hypothesis (H1) for the Republicans and our strategic identification assumptions (H3) for the Democrats (see Table 4).
Descriptive statistics and variable intercorrelations (Republicans below the diagonal (n = 145); Democrats above the diagonal (n = 125)).
*p < .05; **p < .001; ***p < 0.0001. n.s.: not significant.
Finally, to explore group members’ assessment of leaders as “prototypical members” of the group (H4), we examined the partial correlations between ingroup/outgroup similarity and leader prototypicality. For Democrats, there was no relationship between leader prototypicality and similarity with either the ingroup or outgroup. For Republicans, there was a significant correlation between leader prototypicality and perceived similarity with the ingroup (pr = 0.28, p < .01). Thus, for Republicans but not Democrats, the more that respondents viewed their leader as prototypical, the more they also regarded themselves as like other ingroup members, suggesting a greater sense of coherence of the leadership and membership (MDemocrat= 3.52, MRepublican = 3.27; also see Table 4—relationship between leader prototypicality and perceived ingroup similarity). 5
Self-group similarity and perceptions of leaders’ prototypicality
A central question is whether perceived similarity with either ingroup or outgroup members is significantly dependent on the group (i.e., Republicans/Democrats), and whether this is impacted by either/both of the leaders. We also evaluated the extent to which this relationship may be explained by individual differences in using strength of support for the party (which we treat as a covariate) (H1 and H3). We conducted a 2 between (Leader; Clinton vs. Trump) × 2 within (Similarity: ingroup, outgroup) ANCOVA, using strength of support as a covariate (see Table 5 and Figure 1). There were significant main effects for Similarity, F(1, 297) = 15.475, p < .001; and Leader, F(2, 296) = 148.581, p < .001, Wilks λ = .499, and significant simple effects of Leader for both the similarity to the ingroup (F(1, 297) = 200.010, p < .001) and the outgroup (F(1, 297) = 220.462, p < .001), meaning that leaders were important for shaping responses both to peers and the opposition. There was also a significant effect for the covariate, strength of support (F(2, 296) = 12.573, p < .001, Wilks λ = .922).
Means for perceived similarity of views (ingroup and outgroup) by leader.

Effects of conditions on similarity of views.
A related question is whether the perceived prototypicality of the leader depends on the particular group. Given that the two groups have different values and that the Democrat values are more inclusive, a social identity perspective would suggest that prototypicality may be stronger for Republicans, because their identity and values priorities differentiation from others more strongly, whereas Democrats are seeking to reinforce the connection with their groups (H1).
An ANOVA comparing Democrats’ and Republicans’ perceptions of leader prototypicality revealed a marginal main effect for group membership, F(1, 247) = 3.54, p = .069; MDemocrat= 3.72; SD = .92; MRepublican = 3.60; SD =1.00), indicating that perceptions of leader prototypicality do not differ strongly between the two parties. However, strength of support, F(10, 247) = 4.00, p < .05; MDemocrat = 63.78; SD = 29.75; MRepublican = 72.69; SD = 26.98), was a significant covariate, indicating that commitment to the party and perceptions of leader prototypicality are positively related. However, there was no effect of political party membership with strength of their support on their own representativeness, F(10, 247) = 0.72, p = .705; MDemocrat = 3.72; SD = .92; MRepublican = 3.60; SD =1.00) (H1 and H3).4 Taken together, the evidence suggests that the Republicans see their leader as more prototypical and therefore support him, despite the fact that they—on average—report themselves as less “prototypical” (means above).
Discussion
In this paper, we applied an intergroup framework to explore the 2016 presidential campaign, in which the Republicans and Donald Trump were victors over the Democrats and Hillary Clinton. According to social identity theory (Hogg and Abrams, 1988; Tajfel and Turner, 1979), a central motive for groups and their members is to sustain and reinforce their ingroup identification. More recent theory argues that group members should take advantage of strategic opportunities to reinforce the overall legitimacy of their identity. Yet the implications of this motivation differ for Republicans and Democrats because of their different value priorities. As such, winning or losing a competitive contest concerns the collective future of both groups. Prior to the competition outcome, as was the case here, we would expect members to strongly endorse their own ingroup, reporting strong similarities with other members. This allows groups to be more cohesive and to have a competitive edge within the conflict setting. The data reflect this difference, with Republicans endorsing their ingroup and thus acting consistent with the theory’s tenets (support for H1), but Democrats utilizing an identification strategy that seems to involve embracing the outgroup (reporting high perceived similarity with outgroup members and not the ingroup), thus supporting H3.
The Democrat ingroup
On the face of it, one might ask whether there are alternative explanations for the Democrats' sensitivity to the opposition. For example, one could argue that a social group, with members that behave in such a manner, is “disorganized”; or that they “might not know or be able to detect” their own ingroup values (i.e., the ingroup message is unclear). Turning to the values inventories, we first note that the values measures differentiated between social groups (also see Duckitt, 2001; Rokeach, 1973). That is, the values for each group were divergent, with Democrats, for example, reporting “equality” and “broadmindedness” as central values (supporting H2). Moreover, the mean scores on the scales (strength applied to their distinctive values) amongst for the Democrats and the Republicans were not statistically different. For both groups, mean scores were above the midpoint on the scale. In other words, the perceived similarity with the outgroup members was not the result of the Democrats “not knowing what they believed.” Likewise, the mean scores on the scales indicate that the pattern was not the result of a lack of engagement; there were no systematically low scores on the indices. Moreover, the variances in the ratings of the values were broadly similar amongst both Democrats and Republicans. Thus, there is no reason to believe that Democrats were being less systematic in their perceptions or strategies than Republicans, and the data suggest that the Democrats are following a strategy that is equally deliberative as that of the Republicans.
This leaves us asking, “Why are the Democrats viewing such strong similarities between themselves and the Republican opposition?” A review of the individual values items reveals that “equality” and “broadmindedness” are closely aligned with liberal values of individualizing (Graham et al., 2013). These ingroup values convey an implicit sense of inclusion. To this end, Stewart and Morris (2018) demonstrated that ingroup values, emphasizing fairness and inclusion for outgroup members, led to a reduction in intergroup bias. In other words, the group’s values facilitate a sense of similarity and inclusion. Jost (2017) and others have echoed this theme, demonstrating that the social group is more open to tolerating uncertainty and more resilient to change. This, in turn, motivates them to create a flexible superordinate identity in which they can integrate others (supporting H3). Within the context of a national, intergroup competition, these same core values, however shape the Democrat ingroup into being too permeable. Their flexibility directly works against the group’s cohesiveness and hampers the chance of a “win.”
A last aspect that we might consider is the extent to which the group’s identity is internalized by the individual members, such that there is a sense of personal embeddedness. Here, the strategic identity hypothesis (Marques et al., 2001; H3) suggests that widening the scope of group membership—linked to the values of the ingroup—should mean that group members see themselves as less prototypic of their ingroup (less like other Democrats, because they are individualizing). In part, this would transpire because they are integrating others and facilitating others to become part of their group and of a stronger whole (also see Christian et al., Unpublished; Schubert and Otten, 2002). This sense of inclusion is also evident in their tolerance toward ambiguity and longer action times to novel stimuli and circumstance. When we accounted for any other personal attitudes that might be influencing their views and prior experiences in the analysis, there was no significant effect for self-prototypicality. Thus, the values are internalized by the members, as is the goal for party membership, which is to be “inclusive” and more tolerant.
The Republican ingroup
Consistent with SIT (Abrams and Hogg, 1988; Tajfel and Turner, 1979), the Republicans perceived themselves to be similar to other ingroup members, and they seemed to act to reinforce their ingroup identification (H1). For them, there is a clear distinction between Republicans and Democrats. In line with this, they cling to the ingroup identification, because it helps them in their efforts to preserve social structures and ingroup values. Similar in process, but not outcome, to Democrats, Republicans’ identity is likely facilitated by their distinctive ingroup’s values—“ambitious” and concerned with “security.” These values map well onto Graham et al.’s (2013) notion of “ingroup binding” values. As outlined earlier, ingroup binding is predicated on the notion of “ingroup purity,” or the sense that group members are motivated to enhance personal self-concept by protecting the group entity.
Turning to the Republican responses, the mean scores were high and above the midpoint on the scales. That is, there was confidence in all of their responses. However, in examining the scores on the items, the evidence tells a story of more extreme response patterns for the Republican cohort. This highlights the group’s ability to cope with the interpersonal differences among its members. That is, despite the divergent personal attitudes of the members (more extreme values), the group processes acted to hold the ingroup together. Thus, it is not the value preferences that motivate the Republicans, but the sense of self betterment drawn from the ingroup identity that acts as a “glue” for the membership. It seems plausible that the more cohesive and unified nature of the Republican ingroup was therefore able to “deliver a win” in the face of the competition.
Leaders and ingroups
Strong leadership from prototypical leaders is frequently seen as an avenue to facilitate group cohesion—particularly if that is seen as a shortcoming. We, therefore, examined the perceived prototypicality of Clinton and Trump among their ingroups. Leaders who reflect the “most (proto)typical views of the ingroup,” often in ways that are more extreme than average group members, are those who are the most frequently endorsed by their groups (Abrams et al., 2008). However, this may depend on whether the group’s goal is to diverge from the outgroup in particular ways. The present context considered a situation in which the goal was overall leadership supremacy, and this raises the question of whether the group’s value system favors simple domination or a more inclusive absorption of the outgroup.
On closer review, investigating leader prototypicality, while controlling for interpersonal evaluations of the Republican group members, there is clear support that the ingroup viewed Trump as a “normative leader.” This is in line with Krishnan (2001) who argues that the match between campaign messages and the ingroup values would be enough for him to be seen as normative—this was enough to consolidate the ingroup with him during the competition (H4).
In the same way, we also reviewed the mean score, as well as the relationship between outgroup identification (the Democrats’ construal of their group) controlling for interpersonal preferences of the ingroup members. Similarly, a straightforward read of the mean on the scale would indicate that Clinton was evaluated as being a prototypical leader. However, when group-level perceptions (strength of member similarity; identification) are reviewed, controlling for the individual ones, there is a very different story—one in which she is not seen as normative (H4). We can only speculate as to the reasons motivating this. The evidence suggests that factors, such as “traditional career path,” might be important when the variable is not viewed within the context of ingroup values. However, when prototypicality of the leader is viewed within this framework the lens that is used by ingroup members is rooted in Democrat’s path of inclusiveness. As outlined above, the problem with the inclusiveness strategy as applied here is that rather than members drawing a sense of distinctiveness from the Democrat party affiliation, they draw esteem by integrating others into their party (also see Reicher et al., 2005). This emphasis on the collective approach, ironically, works against Clinton hampering Democrat supporters’ ability to perceive her as delivering on and embodying the group’s values. This outcome occurs because the group becomes too flexible with the inclusion of the opposition. And, in turn, the absence of this endorsement for the leader and the use of this integration strategy worked to undermine the chance of a Democrat “win.”
Conclusion
In closing, group processes, particularly social identity processes and group values, may have had a great deal to do with the outcome of the 2016 US presidential election. The Democrats’ approach, which valued inclusion and led to more positive views of outgroup members, compared with the stronger ingroup cohesiveness of Republicans, likely contributed to the Democrats’ loss in this competitive election. It is important to note that when groups must share a common environment after a competition, the inclusiveness approach of the Democrats would likely work to pull the competing parties together. However, a strategic and more inclusive approach that looks for cooperation before the competition is won results in too much integration and loss of momentum for the group.
It is also important to point out that the strategy of the Republican ingroup to “win” the election is not one that is optimal for the “battle” of holding leadership after the election. Social identity theory indicates that acquiring new members is done under a limited range of conditions. The strongest option for the Republicans, having “won,” would be to take a leaf out of the Democrats’ “playbook”—for continued ingroup validation and positive identity, the “winners” should embrace as many Democrat outgroup members as possible to grow the ingroup (also see Barberá et al., 2015; Piurko et al., 2011). Time will tell how this is implemented.
Footnotes
Acknowledgements
We would like to thank Joseph Cotler, Tierra Patmavanu, and Corinne Voigt-Hill for their assistance in earlier phases of this research. Portions of this paper were presented at the annual International Studying Leadership Conference, Richmond (Virginia), December 2017.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
