Content uploaded by Ronald E. Riggio
Author content
All content in this area was uploaded by Ronald E. Riggio on Sep 26, 2018
Content may be subject to copyright.
Article
Them and us: Did Democrat
inclusiveness and Republican
solidarity lead to the 2016
US presidential
election outcome?
Julie Christian and Daniella Nayyar
University of Birmingham, UK
Ronald Riggio
Claremont McKenna College, Kravis Leadership Institute, USA
Dominic Abrams
Centre for the Study of Group Processes, Department of Psychology,
University of Kent, UK
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 mem-
bers’ 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.
Keywords
Leadership, social identity theory, social values, Republicans, Democrats
Corresponding author:
Julie Christian, School of Psychology, University of Birmingham, Edgbaston, Birmingham B15 2TT, UK.
Email: j.n.christian@bham.ac.uk
Leadership
0(0) 1–19
!The Author(s) 2018
Article reuse guidelines:
sagepub.com/journals-permissions
DOI: 10.1177/1742715018793733
journals.sagepub.com/home/lea
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 represen-
tative 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.
2Leadership 0(0)
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 conserva-
tive ingroups are more likely to value “support for security and order", with the group
focused on relativist-centered behavior, or ones 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 prefer-
ence 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 the ones 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
Christian et al. 3
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 per-
ceived 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 both 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 oppos-
ing outgroup (i.e., Democrat or Republican depending on the target, respectively).
•H2: Group distinctiveness and self-group similarity. Consistent with a perceived congru-
ence (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 out-
group 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 evalu-
ated 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
4Leadership 0(0)
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 ¼10.06 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 with-
drawal 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, percep-
tions of social values, and strength of support (attitudes) were assessed using the follow-
ing items:
1. Similarity between own views and those of ingroup members’ and those of outgroup mem-
bers’ 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)to5(very much).
2. 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 rate how much the same values are held by the other group
Christian et al. 5
(ingroup/outgroup order was randomized). Responses were recoded using five-point
Likert-style scales ranging from 1 (not at all)to5(very much).
3. 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)to5
(very much).
4. 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.
5. 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.
6. 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).
7. 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 (F
1, 268
¼13.013, p¼0.0001; M
Democrat
¼31.98, SD ¼8.82; M
Republican
¼36.24,
SD ¼10.29), and Republicans were more likely to report owning their own homes (F
1,
268
¼12.440 p¼0.0001; M
Democrat
¼.37 or 36.8%; M
Republican
¼.58 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.
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.
6Leadership 0(0)
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:F
1, 268
¼15.56, p¼0.0001; M
Democrat
¼4.11, M
Republican
¼3.73; Republican
ingroup instrumental values:F
1, 268
¼17.05, p¼0.0001; M
Democrat
¼3.89, M
Republican
¼4.29;
Republican ingroup terminal values:F
1, 268
¼20.70, p¼0.0001; M
Democrat
¼3.75,
M
Republican
¼4.31).
3
Only ingroup instrumental values for the Democrats were evaluated
similarly by both Republicans and Democrats (F
1, 268
¼1.40, p¼0.183; M
Democrat
¼3.97,
M
Republican
¼3.73).
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 influ-
ence 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 ¼0.45, p<.01; ingroup instrumental (D):
pr ¼0.44, p<.01; ingroup terminal (R): pr ¼.43, p<.01; ingroup instrumental (R):
pr ¼.44, p<.01). In contrast, among Republicans, ingroup similarity judgments were cor-
related significantly with perceptions of ingroup values (ingroup terminal (D): pr ¼0.39,
Table 1. Effects of sociodemographic variables and group membership.
Variable Social group M/% SD Df F
Age Democrat 31.98 .45 1, 268 13.013**
Republican 36.24 .36 1, 268
Gender Democrat 44.8 – 1, 268 2.18
Republican 36.3 – 1, 268
Ethnicity Democrat 69.6 – 1, 265 5.04*
Republican 82.2 – 1, 265
Education Democrat 76.8 – 1, 260 0.0034
Republican 78.8 – 1, 260
Marital status Democrat 28.8 – 1, 268 9.64**
Republican 47.3 – 1, 268
Home owner Democrat 36.8 1, 268 12.440***
Republican 56.2 1, 268
Prior voting behavior Democrat 76.8 1, 297 0.014
Republican 77.4 1, 297
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 pro-
portion of Caucasian participants); education: (0) not reporting college education (1) reporting college educated (per-
centage 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.
Christian et al. 7
p<.01; ingroup instrumental (D): pr ¼0.33, p<.01; ingroup terminal (R): pr ¼0.31, p<.01;
ingroup instrumental (R): pr ¼0.39, p<.01), and Democrat instrumental and terminal
values were negatively correlated with Republicans’ perceived ingroup similarity (outgroup
terminal: pr ¼0.16, p<.01; outgroup instrumental: pr ¼0.23, p<.001), consistent with
our values congruence hypothesis (H1) for the Republicans and our strategic identification
assumptions (H3) for the Democrats (see Table 4).
Table 2. Descriptives for values, Democrats.
Ranking Ingroup terminal MSD
1 Equality 4.32 0.91
2 Freedom 4.05 0.91
3 A world at peace 4.00 0.96
4 Happiness 3.95 0.89
5 Family security 3.90 0.86
6 Self-respect 3.90 0.92
7 A world of beauty 3.74 1.07
8 A sense of accomplishment 3.73 0.95
9 Inner harmony 3.65 0.97
10 Wisdom 3.64 1.01
11 National security 3.60 1.02
12 Social recognition 3.60 0.95
13 A comfortable life 3.53 0.87
14 True friendship 3.50 0.99
15 Mature love 3.48 1.10
16 Pleasure 3.47 0.90
17 An exciting life 3.19 0.89
18 Salvation 2.89 1.18
Ranking Ingroup instrumental MSD
1 Broadminded 4.00 1.00
2 Helpful 3.98 0.94
3 Capable 3.95 0.94
4 Logical 3.94 0.87
5 Intellectual 3.93 0.99
6 Responsible 3.90 0.93
7 Ambitious 3.80 0.95
8 Self-controlled 3.79 0.91
9 Forgiving 3.76 1.01
10 Polite 3.70 0.98
11 Clean 3.69 0.99
12 Courageous 3.69 1.03
13 Independent 3.63 1.07
14 Loving 3.62 1.01
15 Cheerful 3.60 0.97
16 Honest 3.60 1.01
17 Imaginative 3.58 1.02
18 Obedient 3.23 1.06
8Leadership 0(0)
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 pro-
totypicality 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 ¼0.28, p<.01). Thus, for Republicans but not Democrats, the more that
respondents viewed their leader as prototypical, the more they also regarded themselves
Table 3. Descriptives for values, Republicans.
Ranking Ingroup terminal MSD
1 National security 4.61 0.65
2 Freedom 4.44 0.88
3 Family security 4.39 0.88
4 A sense of accomplishment 4.22 0.93
5 Self-respect 4.11 0.94
6 Happiness 4.10 0.92
7 Wisdom 3.95 1.03
8 A comfortable life 3.93 0.90
9 A world at peace 3.68 0.94
10 Social recognition 3.67 1.03
11 Salvation 3.66 1.07
12 Mature love 3.60 1.13
13 Inner harmony 3.55 1.12
14 True friendship 3.54 1.12
15 Equality 3.42 1.09
16 A world of beauty 3.39 1.11
17 Pleasure 3.33 1.12
18 An exciting life 3.18 1.1
Ranking Ingroup instrumental MSD
1 Responsible 4.33 0.88
2 Ambitious 4.26 0.94
3 Capable 4.26 0.91
4 Logical 4.24 0.93
5 Independent 4.13 1.06
6 Self-controlled 4.11 0.95
7 Courageous 4.09 0.97
8 Intellectual 4.07 0.98
9 Clean 4.01 1.02
10 Honest 4.00 1.03
11 Helpful 3.86 1.06
12 Polite 3.78 0.97
13 Loving 3.61 1.04
14 Obedient 3.61 1.06
15 Cheerful 3.55 1.05
16 Broadminded 3.53 1.12
17 Forgiving 3.43 1.10
18 Imaginative 3.42 1.11
Christian et al. 9
Table 4. Descriptive statistics and variable intercorrelations (Republicans below the diagonal (n ¼145); Democrats above the diagonal (n ¼125)).
Variable MSD12 345678
Ingroup similarity 3.79 (1.72) 0.97 (0.83) – 0.01 0.11 0.14 0.08 0.08 0.09 0.131 n.s.
Outgroup similarity 1.79 (3.88) 0.80 (0.91) 0.27* – 0.44*** 0.45*** 0.43*** 0.44*** 0.11 0.525***
Ingroup instrumental
values Democrat
3.81 (3.97) 0.92 (0.84) 0.39*** 0.23* – 0.71*** 0.59*** 0.73*** 0.04 0.206*
Ingroup terminal
values Democrat
3.75 (4.11) 0.79 (0.78) 0.33*** 0.22* 0.73*** – 0.66*** 0.70*** 0.14 0.248**
Ingroup terminal
values Republican
4.34 (3.77) 0.68 (0.78) 0.39*** 0.13 0.46*** 0.51*** – 0.77*** 0.27** 0.283**
Ingroup instrumental
values Republican
4.29 (3.91) 0.80 (0.78) 0.31*** 0.26* 0.60*** 0.61*** 0.72*** – 0.21* 0.321***
Leader prototypicality 3.27 (3.52) 1.02 (0.94) 0.28** 0.02 0.29*** 0.33*** 0.23* 0.24* –
0.267*
Self-prototypicality 3.58 (3.70) 1.00 (0.92) 0.626*** 0.121 n.s. 0.515*** 0.519*** 0.433*** 0.409*** 0.375*** –
*p<.05; **p<.001; ***p<0.0001. n.s.: not significant.
10 Leadership 0(0)
as like other ingroup members, suggesting a greater sense of coherence of the leadership and
membership (M
Democrat
¼3.52, M
Republican
¼3.27; 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 relation-
ship 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) ¼15.475, p<.001; and Leader, F(2, 296) ¼148.581, p<.001, Wilks k¼.499, and sig-
nificant simple effects of Leader for both the similarity to the ingroup (F(1, 297) ¼200.010,
p<.001) and the outgroup (F(1, 297) ¼220.462, p<.001), meaning that leaders were impor-
tant for shaping responses both to peers and the opposition. There was also a significant
effect for the covariate, strength of support (F(2, 296) ¼12.573, p<.001, Wilks k¼.922).
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
Table 5. Means for perceived similarity of views (ingroup and outgroup) by leader.
Ingroup similarity Outgroup similarity
Candidate MSDM SD
Hillary Clinton 2.03 1.12 3.55 1.19
Donald Trump 3.79 0.96 1.80 0.79
Figure 1. Effects of conditions on similarity of views.
Christian et al. 11
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 prototypi-
cality revealed a marginal main effect for group membership, F(1, 247) ¼3.54, p¼.069;
M
Democrat
¼3.72; SD ¼.92; M
Republican
¼3.60; SD ¼1.00), indicating that perceptions of
leader prototypicality do not differ strongly between the two parties. However, strength of
support, F(10, 247) ¼4.00, p<.05; M
Democrat
¼63.78; SD ¼29.75; M
Republican
¼72.69;
SD ¼26.98), was a significant covariate, indicating that commitment to the party and per-
ceptions 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) ¼0.72, p¼.705; M
Democrat
¼3.72; SD ¼.92; M
Republican
¼3.60; SD ¼1.00)
(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 cam-
paign, 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 similar-
ities 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 differ-
entiated 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
12 Leadership 0(0)
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 strat-
egy that is equally deliberative as that by 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 individu-
alizing (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 uncer-
tainty and more resilient to change. This, in turn, motivates them to create a flexible super-
ordinate 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 cohe-
siveness and hampers the chance of a “win.”
A last aspect that we might consider is the extent to which the group’s identity is inter-
nalized 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 out-
lined 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.
Christian et al. 13
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 differ-
ences 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 interper-
sonal 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 evi-
dence 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 inclu-
siveness strategy as applied here is that rather than members drawing a sense of distinctive-
ness 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
14 Leadership 0(0)
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 contrib-
uted 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 Barbera
´et al., 2015; Piurko et al., 2011). Time will tell how this is
implemented.
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.
Notes
1. There are large literature investigating changes to the female stereotype (Diekman and Eagly, 2000;
Eagly and Steffen, 1984) and women in leadership (see Eagly and Karau, 2002; Eagly and Carli,
2003; Ryan and Haslam, 2005). Both of these bodies of work are relevant for those interested in
further exploring the impact that such factors might have had on Clinton’s aspirations for leader-
ship. Also, in this study, please note that we are focusing on the extent to which the leader reflects
the group’s prototypic behaviors.
Christian et al. 15
2. We have recoded as interval variables as indicated in the footnote of Table 1. However, if
the analysis is conducted using two groups, Democrats and Republicans, and tested with a t-test
procedure, the results are similar: Democrat ingroup terminal values:M
Democrat
¼4.11,
M
Republican
¼3.73; t(269) ¼3.94, p¼0.0001; Republican ingroup instrumental values:
M
Democrat
¼3.89, M
Republican
¼4.29; t(267) ¼4.135, p¼0.0001; Republican ingroup terminal
values:M
Democrat
¼3.75, M
Republican
¼4.31; t(267) ¼6.177, p¼0.0001). Only ingroup instrumental
values for the Democrats was evaluated similarly by both Republicans and Democrats
(M
Democrat
¼3.97; M
Republican
¼3.82; t(269) ¼1.335, p¼n.s.).
3. The pattern of correlations was not significantly different when we did not control for the
strength of attitudes. For example, the correlation between ingroup similarity and leader pro-
totypicality is: r¼.272, p<0.001 for Republicans; for Democrats, the association between out-
group similarity and leader prototypicality is: r¼0.129, p¼n.s. (the outgroup being the
“ingroup” for the Democrats).
4. To check for possible restrictions in the range that might potentially have affected the correlations,
we examined the means and standard deviations associated with the individual variables used to
create the indices (above). A review of means showed that they were similar across both groups,
above the mid-point on the scale (five-point scale with average of top three distinctive items)
(Democrats: 4.09 (terminal), 3.61 (instrumental); Republicans: 4.23 (terminal), 4.27 (instrumental)).
Only low means, below mid-point, would signal a lack of consensus for either of the groups (see
Table 4). The standard deviations for the values variables (items used to create indices) were
somewhat narrower for the Democrat (0.89–1.00) participants than for their Republican counter-
parts (0.65–1.06) (see Tables 2 and 3). Overall then, there seems little basis to believe that either
mean level responses or differences in variance had a bearing on the pattern of correlations.
5. To test the significance of the difference between correlational coefficients, we compared the cor-
relation of each value with the ingroup similarity versus outgroup similarity (e.g., Democrats:
0.11, 0.44; see Table 4 for coefficients) The Z-test results for the Democrats are as follows:
Democrat instrumental values: Z¼4.55, p¼.00001; Democrat terminal values: Z¼4.89,
p<0.00001; Republican instrumental values: Z¼4.22, p<0.00001; Republican terminal values:
Z¼4.31, p<0.00001; Leader prototypicality and ingroup/outgroup similarity: Z¼0.16,
p¼0.87; Self-prototypicality and ingroup/outgroup similarity. Z¼3.44, p<0.001. Conversely,
the Z-test results for the Republicans are: Republican instrumental values: Z¼4.77, p<0.00001;
Republican terminal values: Z¼4.41, p<0.00001; Democrat instrumental values: Z¼5.25,
p<0.00001; Democrat terminal values: Z¼4.60, p<0.00001; Leader prototypicality and
ingroup/outgroup similarity: Z¼2.05, p<0.05; Self-prototypicality and ingroup/outgroup similar-
ity. Z¼6.96, p<0.00001.
References
Abrams D and Hogg MA (1988) Comments on the motivational status of self-esteem in social identity
and intergroup discrimination. European Journal of Social Psychology 18: 317–334.
Abrams D and Hogg MA (1990) Social identification, self-categorization and social influence.
European Review of Social Psychology 1: 195–228.
Abrams D, Randsley de Moura G, Marques JM, et al. (2008) Innovation credit: When can leaders
oppose their group’s norms? Journal of Personality and Social Psychology 95: 662–678.
Abrams D, Travaglino GA, Marques JM, et al. (2018) Deviance credit: Tolerance of deviant ingroup
leaders is mediated by their accrual of prototypicality and conferral of their right to be supported.
Journal of Social Issues 74: 36–55.
Barbera
´P, Wang N, Bonneau R, et al. (2015) The critical periphery in the growth of social protests.
PLOS ONE 10(11): e0143611.
Braithwaite V (1994) Beyond Rokeach’s Equality-Freedom Model: Two-dimensional values in a one-
dimensional world. Journal of Social Issues 50: 67–94.
16 Leadership 0(0)
Christian J, Nayyar D, Riggio RE, et al. (Unpublished) “Bridging the divide”?: Intergroup magna-
nimity by Brexiteers following the UK referendum. European Journal of Social Psychology
(in review).
Cohn N (2018) California results are consistent with a competitive house race. The New York Times,6
June. Available at: https://www.nytimes.com/2018/06/06/upshot/california-results-are-consistent-
with-a-competitive-house-race.html
Conover PJ and Feldman S (1981) The origins and meaning of liberal/conservative self-identifications.
American Journal of Political Science 25: 617–645.
Diekman AB and Eagly AH (2000) Stereotypes as dynamic constructs: Women and men of the past,
present, and future. Personality and Social Psychology Bulletin 26: 1171–1188.
Dimock M, Doherty C, Kiley J, et al. (2014) Political polarization in the American public. Pew
Research Center, 12 June.
Duckitt J (2001) A dual-process cognitive-motivational theory of ideology and prejudice. In: Zanna
NP (ed) Advances in Experimental Social Psychology (vol.33). Cambridge: Academic Press,
pp. 41–113.
Eagly AH and Carli LL (2003) The female leadership advantage: An evaluation of the evidence. The
Leadership Quarterly 14: 807–834.
Eagly AH and Karau SJ (2002) Role congruity theory of prejudice toward female leaders.
Psychological Review 109: 573–598.
Eagly AH and Steffen VJ (1984) Gender stereotypes stem from the distribution of women and men
into social roles. Journal of Personality and Social Psychology 46: 735–754.
Graham J, Haidt J, Koleva S, et al. (2013) Moral foundations theory: The pragmatic validity of moral
pluralism. In: Devine P and Plant A (eds) Advances in Experimental Social Psychology (vol.47).
Cambridge: Academic Press, pp. 55–130.
Haidt J and Graham J (2007) When morality opposes justice: Conservatives have moral intuitions that
liberals may not recognize. Social Justice Research 20: 98–116.
Haidt J and Joseph C (2004) Intuitive ethics: How innately prepared intuitions generate culturally
variable virtues. Daedalus 133: 55–66.
Haslam SA and Reicher SD (2007) Identity entrepreneurship and the consequences of identity failure:
The dynamics of leadership in the BBC Prison Study. Social Psychology Quarterly 70: 125–147. doi
10.1177/0902727250707000204.
Hogg MA (2001) A social identity theory of leadership. Personality and Social Psychology Review
5: 184–200.
Hogg MA and Abrams D (1988) Social Identifications: A Social Psychology of Intergroup Relations and
Group Processes. London: Routledge.
Jost JT (2017) Ideological asymmetries and the essence of political psychology. Political Psychology
38: 167–208.
Jost JT and Napier JL (2012) The uncertainty-threat model of political conservatism. In: Hogg MA
and Blaylock DL (eds) Extremism and the Psychology of Uncertainty. West Sussex: Wiley-
Blackwell, pp. 90–111.
Krishnan VR (2001) Value systems of transformational leaders. Leadership & Organization
Development Journal 22: 126–132.
Marques JM, Abrams D, Paez D, et al. (2001) Social categorization, social identification, and rejection
of deviant group members. In Hogg MA and Tindale S (eds) Blackwell Handbook of Social
Psychology: Group Processes. Malden: Blackwell Publishing, pp. 400–424.
McHoskey JW (1996) Authoritarianism and ethical ideology. The Journal of Social Psychology
136: 709–717.
Meertens RW and Pettigrew TF (1997) Is subtle prejudice really prejudice? The Public Opinion
Quarterly 61: 54–71.
Morris DSM (2018) Moral and political cognition: Implications for improving intergroup relations, civic
engagement and reducing prejudice. PhD Thesis. University of Birmingham, UK.
Christian et al. 17
Piurko Y, Schwartz SH and Davidov E (2011) Basic personal values and the meaning of left-right
political orientations in 20 countries. Political Psychology 32: 537–561.
Reicher S, Haslam SA and Hopkins N (2005) Social identity and the dynamics of leadership: Leaders
and followers as collaborative agents in the transformation of social reality. The Leadership
Quarterly 16: 547–568.
Rokeach M (1973) The Nature of Human Values. New York: Free press.
Ryan MK and Haslam SA (2005) The glass cliff: Evidence that women are over-represented in pre-
carious leadership positions. British Journal of Management 16: 84–90.
Schubert TW and Otten S (2002) Overlap of self, ingroup, and outgroup: Pictorial measures of self-
categorization. Self and Identity 1: 353–376.
Schwartz SH (2012) An overview of the Schwartz theory of basic values. Online readings in Psychology
and Culture 2. Available at: https://doi.org/10.9707/2307-0919.1116
Sparkman DJ and Eidelman S (2016) “Putting myself in their shoes”: Ethnic perspective taking
explains liberal–conservative differences in prejudice and stereotyping. Personality and Individual
Differences 98: 1–5.
Stewart BD and Morris DSM (2018) Moving morality beyond the ingroup: Liberals and conservatives
show differences on group-framed moral foundations and these differences mediate threat and
intergroup bias.
Tajfel H and Turner JC (1979) An integrative theory of intergroup conflict. In: Austin WG and
Worchel S (eds) The Social Psychology of Inter-Group Relations. Monterey: Brooks/Cole
Publishing Co., pp. 33–47.
Taylor P, Funk C and Craighill P (2006) Are we happy yet. Report, Pew Research Center,
Washington, DC, 13 February.
Van Knippenberg D (2011) Embodying who we are: Leader group prototypicality and leadership
effectiveness. The Leadership Quarterly 22: 1078–1091.
Van Knippenberg D, Van Knippenberg B, De Cremer D, et al. (2004) Leadership, self, and identity: A
review and research agenda. The Leadership Quarterly 15: 825–856.
Wenzel M, Mummendey A and Waldzus S (2008) Superordinate identities and intergroup conflict: The
ingroup projection model. European Review of Social Psychology 18: 331–372.
Author biographies
Julie Christian is an applied social psychologist based at the University of Birmingham.
Prior to her appointment, she worked at both the London School of Economics and
Political Science and at Cardiff University. Julie has an established track record in research-
ing social identity, attitudes, leadership, and housing/communities. She was an associate
editor for the Journal of Community and Applied Social Psychology and Member of the
British Psychological Society’s (BPS) Social Psychology Section Board (2004-2006). And, as
a direct extension of her work, she has held research grant awards and executive/charity
board positions both in the US and UK.
Daniella Nayyar was awarded her Bachelor’s degree in Human Psychology from Aston
University in 2013. She is currently a PhD candidate at the School of Psychology,
University of Birmingham.
Ronald Riggio PhD is the Henry R. Kravis Professor of Leadership and Organizational
Psychology and former Director of the Kravis Leadership Institute at Claremont McKenna
College and a Visiting Scholar at Churchill College, Canbridge University. Dr. Riggio is a
leadership scholar with more than a dozen authored or edited books and more than 150
18 Leadership 0(0)
articles/book chapters. His research interests are in leadership, organizational communica-
tion, and social competence. He is part of the Fullerton Longitudinal Study, examining
leadership development across the lifespan (from 1 year of age and through middle adult-
hood). Besides research on leadership development, he has been actively involved in training
young (and not so young) leaders.
Dominic Abrams is Professor of Social Psychology and the Director of the Centre for Group
Processes at the University of Kent, Canterbury. A Chartered Psychologist, he is a Fellow of
SPSSI, SPSP, and the UK Academy of Social Sciences. He is co-founding editor of the
journal Group Processes and Intergroup Relations. He has published extensively in the
areas of social identity theory and intergroup relations, particularly working on the rela-
tionship between social inclusion and social identity.
Christian et al. 19