A follower-centric approach to the 2016 US presidential election: Candidate rhetoric and follower attributions of charisma and effectiveness

Abstract

The 2016 US Presidential Campaign was followed worldwide due to the dynamic nature of the candidates and the controversy surrounding them. This study takes a unique look at the candidacy of Donald Trump through the eyes of his followers. Using both qualitative data collected from campaign speeches and quantitative data collected from followers, our study examines (1) the positive/negative valence of the rhetoric used by both presidential candidates and (2) the relationship between follower characteristics, perceived threat of social groups, and attributions of charisma and effectiveness to Donald Trump. The results of this study suggest that Trump’s rhetoric was significantly more negative, hostile, and aggressive than Clinton’s. In addition, quantitative analyses show a direct relationship between followers’ attitudes toward gender and attributions of charisma and effectiveness to Trump. Finally, followers’ perceived threat of social groups mediated relationships between follower self-esteem, romance of leadership, gender attitudes, and attributions of charisma and effectiveness to Trump. Our results shed further light on the important role that follower characteristics and perceptions play in predicting social constructions of leadership and attributions of charisma. Implications for both organizations and scholars of leadership are discussed.

Keywords

Charisma followership social construction attributions gender presidential leadership follower-centric approaches

“Let me tell you. If Jesus Christ gets down off the cross and told me Trump is with Russia, I would tell him, ‘Hold on a second, I need to check with the President [to see] if it’s true.’”

(Mark Lee, US Citizen and Trump Supporter)

What makes citizens devoutly follow a political leader whom they have never met? This question has long been pondered by political scientists, philosophers, and leadership researchers, yet situations like the 2016 US presidential election remind us that we still lack clear answers. In the leadership literature, scholars point to the importance of followers’ social constructions of leadership (Meindl, 1990, 1995); suggesting that ‘leadership is in the eye of the beholder’ (Avolio et al., 1999; Conger and Kanungo, 1987; Meindl, 1995). Indeed, several decades of research supports the notion that followers’ construct the leadership relationship (Meindl, 1995), attribute leadership qualities to the leader (Conger and Kanungo, 1987; Weber, 1947), and are largely responsible for empowering and enabling leaders’ ascension to power (Howell and Shamir, 2005). However, what role did followers play in the 2016 US presidential election? Although this question is multifaceted in nature, our study attempts to unravel one part of this question by examining follower characteristics that affect attributions of charisma and effectiveness to Donald Trump.

Research suggests that followers play a vital role in the charismatic relationship, and are largely responsible for constructing and attributing charisma to the leader (Conger and Kanungo, 1987; Meindl, 1990, 1995). Followers observe and interpret the leader’s messages and behavior (Awamleh and Gardner, 1999; Conger, 1991), and subsequently make attributions regarding the leader’s level of charisma and effectiveness (Damen et al., 2008). When followers perceive a leader as charismatic, they identify with the leader, become committed to the leader’s vision, and work diligently to turn the vision into reality (Conger and Kanungo, 1998; Kark et al., 2003), in some cases involving personal sacrifice and risk-taking (Bligh and Robinson, 2010). Yet not all followers will attribute charisma to the same leader (Smircich and Morgan, 1982). Charismatic attributions are dependent on the values and vision espoused by the leader, and whether those values resonate with the followers. Followers must perceive that the leader espouses values similar to their own, satisfies some need or deficiency, engages followers’ self-concepts and fulfills a sense of identity (Shamir et al., 1993), and restores order in times of uncertainty (Thoroughgood et al., 2012). Thus, with any single leader, there may be an immense divide between those followers who see the leader as charismatic and those who do not.

In the 2016 presidential election, there was indeed a strong divide between followers who saw Trump as charismatic and those who did not. In fact, the election was “widely heralded as the most divisive presidential contest in the history of the republic” (Vardy, 2016), and post-election media articles lamented how to “heal” the country through helping Trump and Clinton’s followers find a way to talk to one another again (Itkowitz, 2016). Given that the majority of followers never had direct contact with Trump, perceptions of charisma depended largely on his rhetoric. By all accounts, Trump’s rhetoric was indeed unconventional, bombastic, and even degrading to many social groups (i.e., women, immigrants, Muslims, those with physical disabilities, etc.; Wang et al., 2016). But what characterizes the followers who identified with Trump and his vision for improving the US? Previous research suggests that studying characteristics of followers, rather than the leader, provides insight on whether or not a leader is seen as charismatic (Howell and Shamir, 2005; Yukl, 1999).

In the present study, we draw from charismatic leadership theory to identify the characteristics and values that influenced followers’ attributions of charisma and effectiveness to Donald Trump. For example, given Trump’s rhetoric and behavior towards women, we investigate the extent to which followers’ gender attitudes affect attributions of charisma to Trump. Trump also espoused values of US nationalism and independence while denouncing immigrants and refugees. Thus, we investigate followers’ values of communion and collective well-being as predictors of charismatic attributions. Charismatic leadership theory also suggests that certain personal traits may impact how strongly a follower identifies with the leader (Felfe and Schyns, 2006). For example, previous research has linked self-esteem to attributions of charisma, suggesting that individuals with lower self-esteem seek leaders who can provide meaning and stability (Kark et al., 2003; Thoroughgood et al., 2012). Likewise, research suggests that one’s propensity to romanticize leaders affects the extent to which a follower makes charismatic attributions (Meindl, 1995).

Whereas our study explores personal values and characteristics that are theoretically linked to charismatic attributions, we would be amiss if we ignored the important, and often overlooked, role of context in influencing the charismatic relationship (Howell and Shamir, 2005). In light of the challenges facing the US during the presidential campaign, we posit that followers with certain characteristics and values are more likely to perceive threat in the environment (i.e., perceived crisis), and that perceptions of threat mediate the relationship between personal characteristics and attributions of charisma and effectiveness. This is directly in line with charismatic leadership theory suggesting that crisis facilitates and elevates followers’ perceptions of charisma (Bligh et al., 2004; Conger and Kanungo, 1987; Weber, 1947).

Crisis is defined as a situation with little structure that is perceived as threatening and lacks a clear resolution. Although perceptions of crisis are largely subjective, and individuals may differ in the extent to which they deem any situation a crisis, our study examines one specific area of uncertainty that was salient in the 2016 election—the perceived threat of social groups. For example, terrorist events throughout the world and the Syrian refugee crisis experienced prior to the election may have all contributed to followers feeling destabilized and vulnerable. Research suggests that political leaders—and especially those who are considered ‘outsiders’ (Bligh et al., 2005)—may capitalize on instability by blaming incumbent leaders for the crisis (Boin et al., 2010) and offering novel solutions to problems (Williams et al., 2012). Certain followers may have been more susceptible to crisis rhetoric and media coverage than others—further fueling the perception that America was being threatened by certain social groups (i.e., extremists). Thus, we examine the followers’ perceived threat of social groups as a mediating mechanism through which followers attributed charisma to Donald Trump.

Our study examines these relationships against the backdrop of Trump’s campaign rhetoric. The distance that exists between political leaders and followers intensifies the importance of the leader’s rhetoric and language in influencing follower beliefs and perceptions (Conger, 1991; Shamir, 1995; Yagil, 1998). The rhetoric used by Donald Trump in the 2016 US presidential election was uncommon and even peculiar to many Americans (Goldhill, 2017). Yet, while many Americans were alienated by Trump’s rhetoric, others gravitated toward him as a leader who self-identified as a political outsider, and as such could draw on his prior business experiences to improve the state of the nation. Research on charismatic leadership has produced mixed results with regard to leadership rhetoric and attributions of charisma (Damen et al., 2008). While some studies find only positive, inspirational rhetoric as a predictor of charismatic attributions (Awamleh and Gardner, 1999), others suggest that all emotion-laden rhetoric that produces arousal (whether positive or negative) will impact follower perceptions of charisma (Askanasy and Dasborough, 2003; Pastor et al., 2007). Our study evaluates the rhetoric of both Donald Trump and Hillary Clinton, and tests for significant differences in both positively and negatively valenced rhetoric. We use this analysis of key themes underlying the candidates’ rhetoric as a backdrop for examining the types of followers who gravitated toward Trump and perceived him as charismatic.

This study contributes to both theory and research on follower-centric perspectives of leadership. Most notably, the number of empirical studies examining Trump’s rise to power are relatively small, and there is much that can and should be learned about how this particular leader gained resounding support from followers. Moreover, those studies that have examined Trump’s leadership largely do so from the perspective of the leader—choosing to focus on Trump’s rhetoric, delivery style, behavior, or dishonesty (Ahmadian et al., 2017). Our study adds to this literature by providing the unique perspective of the followers who helped him gain power. Finally, our study adds to both theory and research on charismatic leadership by studying followers’ charismatic attributions within a specific and very real context. More specifically, our study identifies individual level variables that were particularly important in understanding followers’ susceptibility to Trump’s rhetoric, vision, and values in an attempt to identify the followers with whom his message resonated. Applying a follower-centric approach to the study of Trump’s leadership allows for greater understanding of how this particular leader was capable of influencing so many Americans, and helps us to unravel the complex interplay among leaders, followers, and contextual factors in predicting a leader’s rise to power.

Follower attributions of charisma

Conger and Kanungo’s (1987) behavioral theory of charismatic leadership provides a foundation to understand both the rhetoric of Donald Trump and the devotion of followers. According to the theory, a leader will be seen as charismatic by followers when that leader: (1) presents a vision of the future that is starkly different from the status quo, (2) uses innovative and unconventional methods to achieve desired change, (3) provides a realistic view of resources and constraints available to bring about desired change, and (4) uses sophisticated methods of articulation and impression management that inspire followers to pursue the vision. When these conditions are met, followers attribute charisma to the leader and subsequently see the leader as highly effective, inspirational, and visionary. The unique rhetoric used by Donald Trump was indeed unconventional, and espoused a vision that was different from the status quo; however, not all followers saw him as equally charismatic. Whereas some followers were strongly devoted to Trump, others saw him as offensive and even appalling. To better understand Trump’s rhetoric, and the followers who were attracted to it, our paper investigates the content of Trump’s speeches as well as characteristics of followers who endorsed him. In the sections below, we present theory on presidential rhetoric as an important antecedent of charismatic attributions, and build on this theory to present hypotheses regarding follower characteristics and beliefs that influence attributions of charisma and effectiveness.

Rhetorical influence on charismatic attributions

The field of rhetorical studies evaluates an actor’s choices about framing, evidence, organization, style, and phrasing in order to achieve desired goals within a specific context. In a campaign year, the goals of a presidential candidate typically focus on gaining support, influencing voters, and working to ensure that voters understand campaign objectives. Audiences exposed to the rhetoric engage in perceptual sense-making, interpret the speech’s meaning, and construct an understanding of both the leader and his/her initiatives (Zarefsky, 2004). When a candidate makes choices about framing and phrasing, they are actively shaping reality for constituents and making salient certain characteristics of the social context (Mayseless and Popper, 2007). Thus, candidates who use phrases of hardship and aggression create characterizations of social reality as more pessimistic and negative. In contrast, candidates who use hopeful, empowering phrases create a view of social reality that is more optimistic and inspirational in nature.

Charismatic leadership theory posits that the rhetoric of the leader is directly related to follower perceptions of both the leader and the situation (Conger, 1991; Conger and Kanungo, 1998; Gardner and Avolio, 1998). Leaders use language to arouse the needs of followers (Yukl, 1999), identify weaknesses in current systems, and create an appealing vision for the future (Emrich et al., 2001). Although researchers have suggested that both the content and delivery of a leader’s speech are related to follower attributions of charisma (Coombs and Holladay, 1994; Holladay and Coombs, 1993), in this study, we are primarily interested in the content or phrasing used to influence followers.

Conger and Kanungo (1987) posit that charismatic leaders use language that demonstrates contempt of the status quo, personalized power, personal cost or risk, and is unconventional or counter-normative. Although these are all elements of charismatic language, the actual words and phrases a leader uses to communicate these elements may be drastically different. For example, research suggests that Donald Trump used more first-person pronouns in the primary election, denoting a more personalized (as opposed to socialized) style of communication (Ahmadian et al., 2017). Others have focused on Trump’s language intended to “demonize” certain social groups (Connolly, 2017), or position the opposing party and sitting president as “outsiders” and “anti-Christian” (Whisker, 2012), denoting displeasure with the status quo and counter-normative behavior. But was Trump’s language more hostile and aggressive than Hillary Clinton’s? Addressing this question could have important implications for the type of followers that gravitated toward Donald Trump, as well as the perceptions followers held about the state of the nation. Specifically, followers gravitate toward leaders who speak publicly about “sentiments that followers feel privately but are unable or unwilling to express themselves” (Bligh et al., 2004: 214). Thus, understanding the content of the leaders’ speeches, as well as how this content differed in terms of pessimism, aggression, and hostility allows us to better understand the followers who attributed charisma to Donald Trump and supported him in the general election.

Follower characteristics and charismatic attributions

Theory and research on charismatic leadership suggests that individual characteristics of followers influence the type of leader that followers see as charismatic or transformational (Felfe and Schyns, 2006; Schyns et al., 2007). For example, Howell and Shamir (2005) posit that it is the followers’ self-concept, rather than specific identifiable actions by the leader, that lead to attributions of charisma. Self-concept clarity is defined as the belief or mental image one holds of herself, as well as the stability of this mental image (Campbell, 1990). Followers with weak self-concept clarity lack a consistent image of the self, require greater direction from others, and are more susceptible to influence from strong and powerful leaders. Conversely, individuals with strong self-concept clarity are more consistent in their beliefs and behaviors, and more attracted to leaders who clearly link their vision to social values and identities. In other words, self-concept clarity provides a foundation for studying individual characteristics of followers that may drive charismatic attributions of a leader who espouses more personalized versus socialized rhetoric. Thus, in light of the importance that follower characteristics play in influencing attributions of charisma, in this study we specifically examine followers’ self-esteem, romance of leadership, attitudes toward gender, and communal values in an attempt to understand which followers resonated most strongly with Trump’s leadership persona.

Self-esteem is considered the “evaluative component of the self-concept” and describes the extent to which people hold a positive or negative view of the self (Howell and Shamir, 2005: 101). When followers have low self-esteem, they are expected to have a greater need for, and dependence on, leaders (Kark et al., 2003). For example, Thoroughgood et al. (2012) suggest that followers with low self-concept clarity and low core self-evaluations seek out leaders who can provide clarity, direction, and lift their self-image. As such, followers with low self-esteem may be more likely to attribute charisma to a leader who espouses personal strength, determination, and confidence. Research also suggests that followers with low self-esteem are prone to feelings of instability (Campbell et al., 1996), and a leader who hones in on uncertainty or threat and attempts to explain it may be perceived as more charismatic (Howell and Shamir, 2005). The uncertainty espoused by the leader may also be used as a sense-making tool, whereby followers with low self-esteem may explain their lack of internal self-concept by the chaos and crisis happening externally around them. Thus, in an uncertain context, followers may be more likely to attribute charisma to a leader who explains the uncertainty and helps them make sense of it.

Hypothesis 1: Followers’ self-esteem will be negatively associated with (a) attributions of charisma and (b) effectiveness to Donald Trump.

Romance of leadership conveys a strong belief or assumption among followers that leaders are of utmost importance in determining organizational success or failure (Meindl and Ehrlich, 1987). Followers with a highly romanticized view of leadership see leaders as sole determinants of organizational outcomes (Meindl, 1995), and have a tendency to attribute outcomes to leaders while discounting contextual or other factors (Meindl et al., 1985). Meindl and Ehrlich (1987: 92) suggest that romanticizing leaders provides a way for individuals to “cope and come to grips with the cognitive and moral complexities of understanding organized activity.” In other words, followers are susceptible to romanticizing leaders in part because they want to believe that leaders have agency, control outcomes, and can resolve at least some of the ambiguities inherent in modern life.

Researchers have studied the individual propensity to romanticize leaders as a trait-like characteristic that influences the extent to which followers attribute outcomes to leadership (rather than to equally plausible factors such as contextual factors or ongoing processes), as well as the extent to which followers are likely to view a given leader as more transformational or charismatic (Bligh et al., 2004; Schyns et al., 2007). Among these studies, a meta-analysis by Schyns et al. (2007) reveals a strong and robust relationship between a follower’s tendency to romanticize leaders and perceptions of transformational leadership. The authors concluded that followers’ “perceptions of leaders comprise more than merely the actual behavior the leader exhibits” (p.505).

Additional research has found that followers who romanticize leaders are more likely to see fresh, new leaders (i.e., “outsiders”) as more charismatic than incumbent leaders (Bligh et al., 2005). In their study of potential voters in the 2003 California Recall Election, Bligh et al. (2005) found that followers with higher romance of leadership scores rated the two outside challengers as more highly charismatic than the incumbent Gubernatorial candidate, whereas followers with lower romance of leadership scores did not differ in their attributions of charisma to each leader. These findings suggest that leaders who portray themselves as outside challengers to the status quo may benefit disproportionately from the romance of leadership effect, in that followers are more likely to rate these leaders as charismatic.

The notion that outside candidates may benefit from the romance of leadership effect is directly relevant to Trump’s candidacy in the 2016 presidential election. Throughout the election, Trump portrayed himself as both an outsider to the Washington “elite” (e.g., “Trump’s war against elites and expertise”; LA Times, 17 July 2017; “Trump: Power belongs to the people, not the Washington elite” Griffiths, 20 January 2017), and as the clear and concise answer to both the immigration crisis (i.e., “build a wall”) and the economic crisis (i.e., import tariffs). For followers with a stronger propensity to romanticize leaders, Trump’s newness to the political arena may have made him a more attractive candidate, elevating followers’ beliefs that he could solve the nation’s problems (Williams et al., 2009). Conversely, followers with a lower tendency to romanticize leaders may not have been positively influenced by Trump’s novelty, or his espoused notion that the country could only be healed by an outsider.

Hypothesis 2: Followers’ romance of leadership will be positively associated with (a) attributions of charisma and (b) effectiveness to Donald Trump.

In addition to follower’s personal characteristics that drive attributions of charisma, this study also investigates whether attitudes related to gender roles and communal values are antecedents to attributed charisma and effectiveness. Gender research suggests that individuals hold strong beliefs about men and women as encompassing distinct social categories (Ashmore et al., 1995). From a social relations perspective, each social category is perceived differently, largely based on gender stereotypes. Research suggests that individuals differ in their beliefs regarding men and women in both interpersonal and societal roles. Whereas one person may hold a very traditional view of a woman as “mother” and “wife,” others may hold differing views of a woman as “business owner” or “politician.” The beliefs that individuals hold regarding roles of men and women influence their attitudes toward how men and women should behave and the positions they are eligible to occupy.

Followers’ attitudes toward gender should influence the type of leader that they gravitate toward and whether they make charismatic attributions. Followers with a traditional view of men being very powerful and women very nurturing may be more likely to attribute charisma to a leader who presents a strong, masculine image. Indeed, Ashmore et al. (1995) suggest that individuals with conservative and traditional values are more likely to hold negative attitudes about female political leaders. Eagly and Karau (2002) suggest that these negative views of women in leadership positions result from the incongruence between traditional attitudes toward gender roles and seeing a woman in a leadership position or engaging in leadership behavior. In the 2016 US general election, Donald Trump not only presented himself as a strong and masculine candidate, but he actively degraded women including his opponent (Wang et al., 2016). These actions may have resonated with followers who held more traditional conservative beliefs and attitudes about gender, thus producing a stronger connection to the leader and a greater likelihood for attributed charisma.

Hypothesis 3: Followers’ negative attitudes toward female political leadership will be positively associated with (a) attributions of charisma and (b) effectiveness to Donald Trump.

Trapnell and Paulhus (2012) define communal values as those that promote collaborative partnership and relationships. Research suggests that individuals with communal values look to others for solidarity, friendliness, and intimacy (Locke, 2003). Individuals with stronger communal values also place greater importance on collective relationships when identifying goals, engaging interpersonally, and evaluating the behaviors of others (Kluckhohn, 1951; Schwartz and Bilsky, 1987). Conversely, those with weaker communal values focus less on the collective and are not as concerned about partnerships and relationships with others. In his campaign for president, Donald Trump actively degraded many social groups including racial minorities, those with disabilities, and immigrants. Statements he made on the campaign trail are in direct opposition to the value of communality, which emphasizes benevolence and universalism (Trapnell and Paulhus, 2012). Trump was also found to engage in speech that was highly narcissistic (Ahmadian et al., 2017); a trait that is in direct opposition to communal values. Followers who identified with Trump, and subsequently saw him as more charismatic, likely held values that were less communal in nature. Given theory to suggest that followers attribute charisma to leaders who embody their own values and attitudes (Shamir et al., 1993), we hypothesize the following:

Hypothesis 4: Followers’ communal values will be negatively associated with (a) attributions of charisma and (b) effectiveness to Donald Trump.

Perceptions of crisis as a mediator

The earliest conceptualizations of charismatic leadership identified the important role that crisis plays in attributing charismatic qualities to a leader (Weber, 1947). Since this time, however, the debate over whether or not crisis is a required element of charisma has intensified (Davis and Gardner, 2012). Whereas some view crisis as a contextual factor that facilitates charisma (Boal and Bryson, 1988), others have found evidence of charisma arising outside of crisis situations (Pillai and Meindl, 1998). In the present study, we are not interested in objectively measuring crisis, but rather assessing followers’ perceptions of crisis as a mediator between their personal characteristics and attributions of charisma and effectiveness.

A crisis allows opportunities for challenging and shifting the status quo (Boin et al., 2010). Trump’s rhetoric regarding crisis and instability in the US, and around the world, provided an opportunity to criticize the sitting president for allowing the crisis to occur, as well as present novel solutions to problems (Whisker, 2012). Research suggests that leaders who promote crisis and create stress among followers are subsequently seen as more charismatic (Halverson et al., 2004). In this way, perceptions of crisis are also socially constructed and fueled by leader rhetoric and media coverage. Research has found that challengers in political elections can capitalize on a crisis situation and are often seen as more charismatic than incumbent leaders (Bligh et al., 2005; Williams et al., 2009). According to Williams et al., this may occur because “the existence of a crisis implies ineffective leadership” from the incumbent (2012: 325). The opportunities that leaders have to hone in on the crisis, blame existing leaders for its occurrence (Bolin et al., 2010), and present a novel vision for how to rectify the situation are prime inputs for charismatic attributions (Conger and Kanungo, 1987). However, in order for Trump to capitalize on instability in the national context, his followers had to first agree with his assessment of the environment as threatening and insecure. Following theory advanced by Howell and Shamir (2005), we argue that certain followers are more likely to perceive instability and threat in their environment, and subsequently attribute charisma to a leader who espouses a strong vision.

For example, research suggests that individuals who are low in self-esteem are more likely to experience helplessness and psychological distress than those who are high in self-esteem (Baumeister, 1997). These feelings of helplessness may be exacerbated by leaders who voice concern over external threats and environmental instability, and individuals may be more likely to perceive threat in the environment as a way of explaining the instability inherent in their own self-concepts. A leader who enables and fuels such perceptions may subsequently be seen as more charismatic.

Followers who have a tendency to romanticize leaders may also be more prone to perceiving threat in the environment, especially when leaders emphasize such threats in their rhetoric (Hamblin, 1958). Given that these followers look to leaders to shape their reality (Mayseless and Popper, 2007) and explain complex situations (Meindl and Ehrlich, 1987), individuals who are high on romance of leadership may be more easily influenced by the leader’s rhetoric, thus aligning their perceptions of the environment with the more negative rhetoric espoused by the leader. These followers may experience greater arousal from the leader’s messages regarding uncertainty, and subsequently find the leader more attractive and charismatic (Pastor et al., 2007).

Furthermore, the values that followers embody may also affect perceptions of threat and crisis, and ultimately make leaders more appealing. In the context of the 2016 presidential election, followers with negative gender attitudes may have seen a female candidate as a threat to their traditional values, and aligned with the stronger male candidate who embodied their attitudes toward women in leadership roles (Ashmore et al., 1995). Likewise, followers with weaker communal values may view dissimilar social groups as destabilizing to society. Research suggests that those who are low in communal values see dissimilar others as holding harmful intentions and, as a result, they are more cautious and critical in nature (Wojciszke et al., 2011). Therefore, lower levels of communal values may lead some followers to see outside social groups as a threat, and subsequently look to a leader who not only embodies similar values, but presents a vision of the future that remedies these perceived threats.

Hypothesis 5: Perceived threat of social groups will mediate the relationship between follower characteristics and (a) attributions of charisma and (b) effectiveness to Donald Trump.

Methods

Our study utilizes both qualitative and quantitative analyses. To address the question of whether Trump and Clinton differed in their rhetoric, we utilized qualitative content analysis to examine the differences in speech content. To test our hypotheses above, we utilized quantitative survey methodology to collect data on followers’ personal characteristics, perceptions of perceived threat of social groups, and attributions of charisma and effectiveness to Donald Trump. Both methodologies are discussed in greater detail below.

Qualitative methodology

We collected transcribed speeches from The American Presidency Project hosted by the University of California, Santa Barbara. The website holds transcriptions of every speech made by presidential candidates dating back to 1960. A total of 14 transcribed speeches were collected for Donald Trump and Hillary Clinton (7 speeches per candidate). In order to achieve consistency across candidate speeches, we collected the Nomination Acceptance speech from each candidate, transcriptions from the three general election debates, and three additional candidate speeches that occurred at universities in the same state at around the same time. Text from each of the speeches was cleaned (e.g., removal of non-speech text, names and introductions provided by hosts or moderators, etc.) and entered into a software program designed to conduct content analysis.

We analyzed speech content using DICTION 7.1.3 (Hart, 2000a, 2000b). DICTION is a computerized content analysis program specifically designed for political discourse, although it has also been used in analysis of shareholder reports (Yuthas et al., 2002) strategic management (Short and Palmer, 2008), and accounting (Sydserff and Weetman, 2002). According to Hart (1984, 1987, 2001), DICTION was specifically designed for political dialogue and was a natural choice given its focus on political discourse and careful attention to linguistic theory (see Neuendorf, 2017 for a recent overview of computerized content analysis programs). Thus, it seemed particularly appropriate for the analysis of the two 2016 US presidential candidates (Bligh et al., 2004; Bligh and Robinson, 2010).

Quantitative methodology

Participants and procedure

A total of 440 participants were recruited from MTurk (a crowd sourcing internet marketplace; N = 203) as well as graduate and undergraduate business courses in a public state university in California (N = 237). All data collection was completed in the week immediately preceding the 2016 US Presidential Election on 8 November, 2016, and participants were invited to fill out an online survey assessing their perceptions of leader and follower characteristics and attributions regarding the upcoming election. Participants were between the ages of 18 and 69 (M = 29.53, SD = 10.61, Mdn = 25), and 53% were female. Political affiliation was divided between Democrat (40.2%), Republican (23.6%), Independent (30.5%), and ‘Other’ (5.7%). Participants rated their political ideology as 3.84 (SD = 1.30) on a scale of 1 = very conservative to 6 = very liberal, indicating that they were somewhat liberal. Political affiliation and ideology of this sample were compared well with that of 24,558 voters across the exit polls conducted by CNN and published on 23 November 2016.

Measures

Unless otherwise specified, study variables used a 7-point Likert scale ranging from 1 (Strongly disagree) to 7 (Strongly agree).

Follower self-esteem was measured with Judge et al.’s (2002) 7-item subscale (α = .95) from the Core Self-Evaluation scale. Sample items from this scale include “I feel that I have a number of good qualities” and “I have a positive attitude toward myself.”

Romance of Leadership was measured with Meindl and Ehrlich’s (1988) 11-item Version C of the Romance of Leadership Scale (α = .85). The scale measures one’s tendency to view leadership as important in determining organizational outcomes. A sample item from this scale is “Sooner or later, bad leadership at the top will show up in decreased performance.”

Attitudes Toward Gender was measured with the Female Political Leadership Subscale of Ashmore et al.’s (1995) Gender Attitude Inventory (α = .92). This subscale measures participants’ attitudes specifically toward female political leaders. A sample item is “Women, as well as men, should be found in top political offices.”

Communal Values was measured with Helgeson and Palladino’s (2012) 9-item unmitigated communion scale (α = .82). Sample items from this scale include “I’m only happy when others are happy” and “I often worry about others’ problems.”

Perceived threat of social groups was measured with three items written specifically for this study (α = .72). Drawing from literature on perceived environmental threats (Quillian, 1995), we asked participants the extent to which they agreed or disagreed with the following items: “Terrorism is a serious threat to our country,” “Illegal immigration is a serious threat to our country,” and “Foreign business competition is a serious threat to our country.” Factor analysis of the three items demonstrated that the last item did not load with the others and significantly reduced the Cronbach’s alpha reliability. Thus, we used only the first two items in the final test of the model.

Attributed Charisma was measured with three subscales from Bass and Avolio’s (1995) short form of the Multi-factor Leadership Questionnaire (MLQ-5X Short form). The subscales used to assess attributed charisma included Attributed Charisma (AC), Idealized Influence (II), and Inspirational Motivation (IM); whereas the subscale for Leader Effectiveness (EF) was used to measure expected effectiveness. This use of the MLQ-5X was based on the findings provided by Bligh et al.’s (2005) work, where they found that AC, II, and IM loaded on one factor and effectiveness loaded on another in a factor analysis (also see Awamleh and Gardner, 1999). An exploratory factor analysis (EFA) was conducted and the same two-factor solution was produced.

Each of the AC (α = .85), II (α = .84), and IM (α = .86) subscales contain four items. Chronbach’s alpha for the combined measure was α = .85. The EF subscale contains two items (α = .97). A sample item of AC is “Instills pride in being associated with him”. A sample item of II is “Talks about his most important values and beliefs”. A sample item of IM is “Talks optimistically about the future”. A sample item of effectiveness is “Overall, would lead the US effectively”. A higher score indicates that the respondents perceive Trump as highly charismatic and highly effective as a leader.

Control variables

Previous research suggests that both demographic variables and general political knowledge are related to one’s political preference for certain leaders (Bligh et al., 2005). Thus, we controlled for a number of variables when testing our study hypotheses. In addition to controlling for gender and age, we also controlled for participants’ general political knowledge. We measured political knowledge with five items adapted from Kathlene (1989): (1) Did you watch the 9 October Presidential Debate? (2) What political party currently has the most members in the US Senate? (3) How long is the term of office for a US Senator? (4) How many US Senators are there? and (5) What percent of US Senators are women? Scores for each participant were computed by assigning one point for every question answered correctly. This produced an overall political knowledge score for each participant that ranged from 0 (not very knowledgeable) to 5 (very knowledgeable).

Quantitative analytic strategy

All hypotheses were tested with regression-based analyses using the PROCESS macro with bootstrapping (Preacher and Hayes, 2008). PROCESS allows the researchers to test for direct and indirect effects while controlling for covariates. The program also produces bootstrapping results on the indirect effects based on a resampling technique. Each model tested included the designated x, y, and m variables identified in the hypothesis, and all other independent and control variables were entered as covariates. This allows for a more robust test of each independent variable while controlling for other predictor and control variables that may share variance with the mediator or outcomes. The PROCESS macro is preferable over other methods of testing mediation (i.e., Baron and Kenny, 1986) because it does not assume that the distribution of the indirect effect is normal and it allows for tests of indirect effects even when the direct effect between x and y is weak or non-existent (Rucker et al., 2011).

Results

Results from the qualitative content analysis of campaign speeches are listed in Table 1. The DICTION program analyzes a passage using 33 different pre-defined dictionaries, containing over 10,000 search words. These dictionaries contain individual words only. The program generates scores for each 500-word passage for ease of comparison. In addition, statistical weighting procedures are used to handle homographs (Hart, 2000a), and words which are not included in the existing dictionaries are also excluded from the analysis. We removed three short speech segments with less than 95 words, giving us an overall sample of 153,500-word segments (70 for Clinton and 83 for Trump).

Table 1.

Results of qualitative speech content analysis.

	Sample quotes	Mean for Trump speeches	Mean for Clinton speeches	t
Construct dictionaries
Optimism	“Let's look to the future with courage and confidence” (Clinton)	47.92	50.52	5.27**
Communality	“I refuse to believe we can't find common ground here. We have to heal the divides in our country, not just on guns, but on race, immigration and more.” (Clinton)	49.33	50.01	2.55*
Activity	“When innocent people suffer, because our political system lacks the will, or the courage, or the basic decency to enforce our laws … I am not able to look the other way, and I won't look the other way.” (Trump)	52.88	50.83	−4.84**
Certainty	“Nobody knows the system better than me, which is why I alone can fix it.” (Trump)	50.55	47.93	−6.68**
Individual dictionaries
Inspiration	“We are clear-eyed about what our country is up against. But we are not afraid. We will rise to the challenge just as we always have.” (Clinton)	2.87	4.68	4.24**
Familiarity	“Millions of people across our country are standing up and saying: We believe in an America that is great because it is good.” (Clinton)	2.58	3.79	6.87**
Cooperation	“Our country needs your ideas, energy and passion. That is the only way we can turn our progressive platform into real change for America” (Clinton)	2.58	3.79	2.63*
Passivity	“This is one of those make-or-break moments for the United States. And it truly is in your hands, as it should be.” (Clinton)	3.36	4.51	2.66**
Blame	“So if you want to hear the corporate spin, the carefully-crafted lies, and the media myths the Democrats are holding their convention next week. Go there.” (Trump)	4.56	2.11	−6.11**
Denial	“We're six days away from the change you've been waiting for your entire life. It'll never happen again. It'll never, ever happen. It's not going to happen in four years.” (Trump)	10.45	7.41	−4.08**
Hardship	“Americans watching this address tonight have seen the recent images of violence in our streets and the chaos in our communities.” (Trump)	5.56	4.42	−2.03*
Aggression	“We are going to defeat the barbarians of ISIS, and we are going to defeat them fast.” (Trump)	5.49	4.46	−1.95*

N = 151. *p < .05. **p < .01, two-tailed test.

Based on the DICTION master constructs, we found significant differences in rhetoric between the two candidates. Specifically, Clinton used significantly more Optimism, which is defined as language endorsing a person, group, concept or event, and highlighting their positive entailments. In addition, Clinton used significantly more Commonality, defined as language highlighting the agreed-upon values of a group and “rejecting idiosyncratic modes of engagement.”

In contrast, Trump utilized significantly more language in the Certainty dictionary, which includes language indicating resoluteness, inflexibility, and completeness and a tendency to speak ex cathedra (Latin for “from the chair,” meaning to speak from the seat of authority or with authority). He also used significantly more Activity than Clinton did, which is defined as language featuring movement, change, the implementation of ideas and the avoidance of inertia. The two candidates did not differ in their levels of Realism, a measure of language describing tangible, immediate, recognizable matters that affect people’s everyday lives (see sample quotes in Table 1).

Finally, we examined the individual dictionaries for significant differences between the two candidates. Based on comparative analyses, Trump uses significantly more Blame, Denial, Hardship, and Aggression, while Clinton’s campaign rhetoric included significantly more Inspiration, Familiarity, Cooperation, Passivity, and Human Interest. These results provide evidence that there are significant differences between the rhetoric used by Donald Trump and Hillary Clinton, and that these differences can largely be classified by negatively and positively valanced rhetoric used by each candidate, respectively.

An analysis of our quantitative data suggests that some, but not all, of our study hypotheses were supported. Means, standard deviations, and intercorrelations of the key study variables are presented in Table 2. Prior to testing our hypotheses, we subjected the measurement model to confirmatory factor analysis (CFA) using EQS 6.1 (Bentler, 2005).

Table 2.

Means, standard deviations, and inter-correlations of study variables.

	1	2	3	4	5	6	7	8	9	10
1. Age	–
2. Sex	−.08	–
3. Political knowledge	.15*	.15*	–
4. Follower self-esteem	.04	−.01	.12*	(.95)
5. Romance of leadership	.09	.12*	.19*	.24**	(.85)
6. Gender attitudes	.07	−.17*	.16*	.03	.19**	(.92)
7. Communal values	−.17*	−.18*	−.05	.01	.14**	.03	(.82)
8. Prcd threat of social grps	.01	−.11*	−.02	.17**	.27**	−.15**	.06	(.72)
9. Trump charisma	.05	.05	−.08	.10**	.09	−.27*	.03	.43**	(.85)
10. Trump effectiveness	.01	.09*	−.07	.09	.04	−.38**	−.01	.48**	.80**	(.98)
Mean	.47	29.5	3.15	5.51	5.60	6.0	4.17	4.73	3.77	2.88
SD	.50	10.61	1.48	1.21	.93	1.13	.97	1.59	1.44	2.00

SD: standard deviation.

N = 440. Gender was coded as 0 = Female, 1 = Male. *p < .05. **p < .01, Two-tailed test.

Variables included in the CFA included follower self-esteem, romance of leadership, gender attitudes, communal values, and attributed charisma. Neither perceived threat of social groups nor effectiveness was included in the CFA because each scale had only two items each. The 5-factor model produced adequate fit with the data (χ² (849) = 1794.33, p < .001, CFI = .90, TLI = .90, RMSEA = .05). We tested four alternative models by combining constructs together under single factors to test for a better fitting model (Anderson and Gerbing, 1988). None of the alternative models demonstrated superior fit with the data (see results in Table 3). Thus, our CFA results provide some evidence that the relationships between study variables were not contaminated by single-source bias (Richardson et al., 2009).

Table 3.

Tests of baseline and alternative measurement models.

Model	Description	χ ²	df	Δχ²	CFI	TLI	RMSEA	SRMR
Baseline	Hypothesized 5-factor model	1794.33	849		.90	.90	.05	.06
Alternative
Model 1	4-Factor model with follower self-esteem and romance of leadership combined	2563.45	853	−769.12*	.85	.81	.07	.08
Model 2	4-Factor model with romance of leadership and attributed charisma combined	2610.38	853	−816.05*	.81	.80	.07	.08
Model 3	4-Factor model with romance of leadership and gender attitudes combined	2604.36	853	−810.03*	.81	.80	.07	.08
Model 4	1-Factor model with all variables combined	8287.57	859	−6493.24*	.20	.16	.14	.16

χ²: chi-square; df: degree of freedom; CFI: comparative fit index; TLI: Tucker–Lewis index; RMSEA: root mean square error of approximation; SRMR: standardized root mean square residual.

N = 440; *p < .05.

Hypothesis 1 proposed that follower self-esteem would be directly related to follower attributions of (a) charisma and (b) effectiveness to Trump. As shown in Table 4, no support was found for a direct effect between follower self-esteem and attributed charisma or effectiveness after controlling for demographic and other individual characteristics in the model.

Table 4.

Effects of follower characteristics on attributions of charisma and effectiveness.

	Direct effects				Indirect effects through PTSG
			95% CI				95% CI
Outcome variables	Effect	SE	LL	UL	Effect	SE	LL	UL
	Direct and indirect effects of romance of leadership
Trump attributed charisma	.04	.07	−.10	.18	.17**	.04	.09	.25
Trump effectiveness	−.034	.09	−.22	.15	.27**	.06	.15	.41
	Direct and indirect effects of follower self-esteem
Trump attributed charisma	.05	.05	−.05	.15	.05*	.02	.002	.11
Trump effectiveness	.06	.06	−.08	.19	.08*	.04	.002	.16
	Direct and indirect effects of attitudes toward gender
Trump attributed charisma	−.26**	.05	−.36	−.15	−.11*	.03	−.16	−.06
Trump effectiveness	−.50**	.07	−.65	−.36	−.17**	.04	−.25	−.10
	Direct and indirect effects of communal values
Trump attributed charisma	.05	.06	−.07	.17	.001	.03	−.06	.06
Trump effectiveness	−.02	.08	−.18	.15	.002	.05	−.10	.10

PTSG: perceived threat of social groups; SE: standard error; CI: confidence interval; LL: lower limit; UL: upper limit.

N = 440. *p < .05; **p < .01.

Hypothesis 2 proposed that romance of leadership would be positively related to follower attributions of (a) charisma and (b) effectiveness to Trump. Again, no support was found for a direct effect between romance of leadership and attributed charisma or effectiveness after controlling for other variables.

Hypothesis 3 posited that positive attitudes toward gender would be negatively associated with attributions of follower (a) charisma and (b) effectiveness to Trump. Support was found for a direct effect between gender attitudes and attributed charisma (−.26, 95% CI [−.36, −.15]) and effectiveness (effect = −.50, 95% CI [−.65, −.36]). Thus, both Hypotheses 3a and 3b were supported.

Hypothesis 4 stated that communal values would be negatively associated with follower attributions of (a) charisma and (b) effectiveness to Trump. This hypothesis was not supported as there was no association between communal values and attributed charisma or effectiveness.

Finally, Hypothesis 5 stated that perceived threat of social groups would mediate the relationship between all four follower characteristics and follower attributions of (a) charisma and (b) effectiveness to Trump. Interestingly, there was strong support for the mediated effects in our model, despite the fact that many of the direct effects were non-significant. Specifically, perceived threat of social groups significantly mediated the relationships between romance of leadership and both attributed charisma (indirect effect = .17, 95% CI [.09, .25]) and effectiveness (indirect effect = .27, 95% CI [.15, .41]). We also found significant mediation between gender attitudes and both attributed charisma (indirect effect = −.11, 95% CI [−.16, −.06]) and effectiveness (indirect effect = −.17, 95% CI [−.25, −.10]). Finally, perceived threat of social groups mediated the relationship between follower self-esteem and both attributed charisma (indirect effect = .05, 95% CI [.002, .11]) and effectiveness (indirect effect = .08, 95% CI [.002, .16]), however the direction of the effect was opposite of what we hypothesized. Specifically, followers with higher self-esteem perceived greater threat from social groups and subsequently attributed more charisma and effectiveness to Trump. No support was found for mediation in the relationships between communal values and either attributed charisma or effectiveness.

Discussion

The 2016 US presidential campaign left citizens of the US and around the world bewildered and shocked by the rhetoric and behavior of Donald Trump. Trump’s rhetoric and behavior influenced his followers’ distrust in both political candidates and media outlets, leading Dictionary.com to add the term “Fake News” for the first time (Steinmetz, 2017). Some have posited that this election brought in the “Post-Truth Era” where the line between truth and lies becomes blurred, and emotions or personal belief carry more weight than objective facts (Davis, 2017). In a post-truth era, the rise of any leader may then depend more on followers’ emotions, beliefs, and personal characteristics than the objective qualities of the leader. This idea is found in both the popular press and scholarly journals, as evidenced by a recent article in Leadership that captures the widespread astonishment with the election outcome: “How could so many Americans have voted for the inexperienced and unpredictable Donald Trump?” (Ladkin, 2017). Our research addresses this question by studying follower characteristics and beliefs related to attributions of charisma and effectiveness against a backdrop of Trump’s rhetorical content.

The results of our qualitative analysis provide empirical evidence that Trump’s rhetoric contained significantly more blame, denial, and aggression than Clinton’s. Although this finding may simply confirm what many already knew about Trump (Connolly, 2017), what is interesting is that Trump also used more rhetoric of activity and certainty. While it appears that Trump’s rhetoric focused intensely and aggressively on the problems with the status quo, he also spoke about specific actions that would be taken to change it. According to the arousal-attraction hypothesis of charismatic leadership (Pastor et al., 2007), the aggressive and blaming language used by Trump may have sparked emotions among followers, thereby increasing perceptions of attractiveness and making them more attentive to the unconventional solutions he advocated (i.e., “building a wall”). Both the dissatisfaction with the status quo (blaming) and the arousal of his speech (aggression) are directly related to charismatic leadership perceptions according to Conger and Kanungo (1987). Trump’s rhetoric also aligned with another important element of charismatic leadership: It was unconventional and stirred controversy. Specifically, Trump actively assigned blame for the problems inherent in the status quo by calling out certain social groups (i.e., Democrats, immigrants, Muslims, etc., Wang et al., 2016), a tactic that many presidential candidates avoid for fear of losing voter support. Yet, Trump’s rhetoric seemed to attract voters who were “tired of hearing the party line” and wanted a “real candidate.”

The results of our quantitative analysis shed some light on who these voters were. Although our study does not show much support for direct effects of follower characteristics on attributions of charisma and effectiveness, it does provide evidence for the important role that perceived threat of social groups plays as a mediator. Among our respondents, it appears that followers who attributed charisma and effectiveness to Trump did so because they perceived immigration and terrorism as major threats to American life. This finding has been supported by recent research demonstrating that white working-class voters overwhelmingly supported Trump due to their perceived threat of cultural displacement (Cox et al., 2017). According to the study, these voters felt that foreign newcomers were threatening American culture, and were heavily in favor of protecting America from foreign influence. Our study supports and extends these findings by suggesting that voters who felt this way, and subsequently attributed charisma and effectiveness to Trump, were higher in self-esteem, higher in romance of leadership, and held negative gender attitudes.

Previous research has shown that individuals with lower self-esteem are more influenced by the rhetoric of political leaders (Campbell et al., 1996; Thoroughgood et al., 2012), and more likely to be affected by perceived instability and crisis (Yukl, 1999). The findings of this study are counter to what has been demonstrated in the past. Specifically, followers in our study who had higher self-esteem were more likely to perceive threat from social groups and subsequently attribute charisma and effectiveness to Trump. This finding is likely due to the specific context surrounding the election and our use of perceived threat of social groups (rather than general crisis perceptions) as a mediator. Research suggests that Trump voters were primarily white working-class Americans who felt that their dominant status was being threatened by immigrants and foreigners (Schildkraut, 2018). Our results suggest that the more positive these individuals felt about themselves and their status, the more they perceived threat by outsiders. Trump was able to capitalize on white voters by pandering to their higher self-esteem, degrading outsiders and blaming them for the troubles in America. For voters with higher self-esteem, Trump’s rhetoric was more appealing, perhaps sparking more emotion, and leading them to stronger attributions of charisma and effectiveness.

Followers who were higher in romance of leadership also perceived greater threat from social groups and subsequently attributed more charisma and effectiveness to Trump. Research suggests that those who romanticize leaders are more likely to see the leader as having the greatest source of power, knowledge, and expertise (Uhl-Bien and Pillai, 2007). It is possible that these followers were also more strongly influenced by Trump’s blaming rhetoric, seeing certain social groups as the source of the country’s problems, and perceiving Trump as the candidate with the best solutions. This notion is central to attribution theory, and the romance of leadership; that “leaders try to manage the interpretations of important constituencies and stakeholders through the manipulation of retrospective, causal accounts” (Meindl and Ehrlich, 1987: 93). Thus, followers who attributed charisma and effectiveness to Trump may have been specifically influenced by the rhetoric he used to blame and denigrate social groups.

Our results demonstrate both strong direct and indirect effects between followers’ less favorable attitudes towards women politicians and attributions of both charisma and effectiveness to Trump. The direct effect between gender attitudes and attributed effectiveness may be due to the negative and misogynistic rhetoric used by Trump during the campaign (Ladkin, 2017). For those who supported Trump, this might be explained by the notion that followers are more attracted to leaders who say publically what they feel and believe privately (Bligh et al., 2004). The indirect effect is a bit more interesting, however. The social relations perspective sees each gender as comprising a different social category, with different associated attributes and expected behaviors (Ashmore et al., 1995). Thus, followers who perceive the social category of women as comprising traditional, nurturing, and motherly roles may see the rise of a female candidate as a threat. The dissonance felt by these followers who faced Hillary Clinton as the first highly viable female candidate for the US presidency (Eagly and Karau, 2002) may have incited greater perceived threat and greater attributions of charisma and effectiveness to Donald Trump.

Our results did not support the hypothesis that communal values related to attributions of charisma and effectiveness. However, a recent review on authoritarianism suggests that authoritarian followers (i.e., those who ascribe to autocratic and agentic leaders) are not necessarily less communal in nature. Indeed, Harms et al. (2017) suggest that authoritarian followers are highly trustworthy and dedicated to helping others. The difference is that these individuals seek communion with similar others who share their particular beliefs and values (Altemeyer, 1998), which Harms notes may be different from the values of society at large. Although our lack of evidence allows us to merely speculate on the role that communal values play in attributions of charisma and effectiveness to Donald Trump, emerging theory and research on authoritarianism sparks interesting questions that can be pursued in future research.

Another fruitful area to explore in explaining these results is the Social Identity Theory approach to leadership (SIT; Hogg et al., 2012).¹ According to Hogg et al., followers are more likely to attribute charisma and effectiveness to leaders whom they consider more representative of the group. As group membership becomes increasingly salient and important to members of the group, as it arguably does in contested elections with strong party systems in place, members identify more strongly with the group (Republican vs. Democrat). As a result, effective leadership rests increasingly on the leader being considered by followers to possess prototypical properties of the group. According to Hogg et al. (2012: 267), “the clear implication of this attribution process is that followers construct a charismatic leadership personality for group prototypical leaders.” While we did not directly assess followers’ group identities or ratings of leader prototypicality in this sample, SIT may help explain why self-esteem was positively related to attributions of charisma and effectiveness in our results. According to SIT, the higher a follower’s self-esteem, the more likely he or she will be to endorse (and perhaps also passionately defend) the group that he or she identifies with, and therefore more likely to attribute greater charisma and effectiveness to that group’s prototypical leader. Future research may want to more directly assess these relationships.

In addition, the work of Tee et al. (2013) may be relevant here, particularly in understanding the role of perceived threats to social groups. In their theoretical model, Tee et al. (2013: 906) argue that leaders who are viewed as positive influences on group identity arouse positive emotions from followers and motivate them to support the leader's influence on the group. Applied to the current context, this model suggests that followers who perceive direct threats to the group’s identity, and thus one’s own identity, may be more motivated to support a prototypical leader that can save them from the threat (i.e., “the charismatic savior”). This line of reasoning resonates with existing research on the role of crisis perceptions in fostering charismatic attributions, as followers who feel threatened are more likely to romanticize a leader who can step in and resolve the crisis and corresponding uncertainty (e.g., Bligh et al., 2004; Pillai and Meindl, 1998). While our findings may provide some indirect empirical support for Tee et al.’s theory, clearly additional research is needed to explore the mechanisms and boundary effects of these relationships.

Implications and future directions

Our study has implications for both leadership scholars and those following political leadership more generally. In the US, political candidates for state governor, as well as seats in the Senate and House of Representatives have taken a “page out of Trump’s playbook” and are using similar campaign tactics and unconventional rhetoric to appeal to followers, and this is likely occurring elsewhere globally as well. In this way, the “post-truth” era of politics may be intensifying because, among certain groups of followers, the use of emotion over rational truth seems to be effective. For example, followers who are higher in self-esteem, higher in romance of leadership, and hold less favorable attitudes toward women politicians may be susceptible to a candidate who evokes emotion by blaming social groups for discontent with the status quo. Followers who fit this profile will likely see that the perceived threat calls for a strong leader, perhaps an outsider, with a bold vision and unconventional rhetoric.

Our study also has implications for the continued study of charismatic leadership, and specifically those interested in understanding followers’ perceptions of such leaders. It has been noted that certain followers are more susceptible to the influence and rhetoric of leaders (Howell and Shamir, 2005). On the other hand, these same followers empower the leader and motivate the leader to engage in certain behaviors and rhetoric. Thus, in a reciprocal and symbiotic manner, as the leader influences followers, the followers also influence and reward the leader. The idiom of leadership being in the “eye of the beholder” is evident both in the results of our study and the continued follower support for President Trump. Our results support the importance of continuing to study follower characteristics and beliefs as an important enabler of leadership attributions and leaders themselves. One important question involves the role that emotional contagion (Hatfield et al., 1993) played among followers during the Trump campaign. If charisma is a consequence of attraction and arousal (Pastor et al., 2007), to what extent were followers’ reactions to Trump largely emotional (as opposed to rational) in nature, and was this emotion amplified by other Trump followers? Additionally, how could changes in context alter this type of situation and outcome in the future? We know that the context of the 2016 presidential campaign was turbulent; however, would the same rhetorical tactics of blame and aggression work in other social and political situations? Future research may address some of these questions either through continued study of the 2016 US presidential election, or subsequent political campaigns that attempt to apply Trump’s rhetorical strategy under different circumstances.

Finally, our study has implications for research and theory on follower characteristics as antecedents to charismatic attributions. Self-concept theory of charisma provides a foundation from which to study the individual characteristics of followers as antecedents to charismatic attributions. In this particular study, we chose individual characteristics that were thought to most strongly influence attributions of charisma to Donald Trump. However, our lack of direct effects between personal characteristics and attributions of charisma and effectiveness raises a number of questions. First, it is possible that follower characteristics play a relatively less important role in directly affecting attributions of charisma and effectiveness than context or group membership. Specifically, it is possible that personal characteristics such as self-esteem or communal values are less important in determining attributions of charisma than social and political context, or party identification and group affiliation. This raises questions regarding the type of self-concept voters had, and whether self-concept affected charismatic attributions to Trump above and beyond followers’ party allegiance (Howell and Shamir, 2005). Future research should explore these within-group effects to see if self-concept or self-esteem is influential in determining the type of leader that followers find more prototypical within the same group or party.

Second, our findings speak directly to the importance of crisis perceptions in influencing charismatic attributions. Our findings demonstrate that personal characteristics influence perceptions of context (i.e., crisis), and it is perceptions of crisis that drive charismatic attributions. Thus, in this particular study, our results support original theorists who hypothesized that crisis was essential to charisma (e.g., Boal and Bryson, 1988; Weber, 1947). Furthermore, our results point to crisis as more influential than personal characteristics in understanding attributions of charisma and effectiveness. Future research may want to examine how personal characteristics relate to differing perceptions of crisis, or how different types of crises affect charismatic attributions. Therefore, we call on future research to continue studying both the importance of self-concept clarity and the role of crisis in examinations of charismatic leadership.

Limitations

Our study has several limitations that should be noted and addressed in future research. First, our quantitative data come from a single source (followers) measured in a single time period. Although our CFA results suggest that variables were not conflated by response bias, future research should attempt to collect data at different time periods or with a variety of raters. In addition, the qualitative portion of our study only assessed the content of the candidates’ speeches. Since content and delivery go hand in hand, future research may attempt to understand the full context of candidate rhetoric by additionally assessing the delivery tactics Trump used in his speeches.

Although our study measured the follower characteristics that were thought to be most salient in the context of the 2016 US presidential election, there are a number of other individual characteristics that could impact attributions of charisma and effectiveness. For example, our study did not measure followers’ levels of authoritarian personality or narcissism, both of which may have been influential in perceptions of threat and attributions of charisma (Galvin et al., 2010; Harms et al., 2017). Our study also did not measure specific types of media or entertainment that followers were exposed to. For example, Gabriel et al. (2018) found that exposure to Trump though his previous appearances on the reality television show, The Apprentice, and through other media, predicted the formation of parasocial bonds with Trump, which predicted voters believing Trump’s promises, disregarding his unpopular statements, and having generally more positive evaluations of him.

Finally, our study did not measure perceptions of general crisis, but instead used a proxy measure (perceived threat of social groups) that was deemed appropriate for this particular context. Given the history and debate that has surrounded the role of crisis in charismatic attributions (Davis and Gardner, 2012), the use of a proxy measure may not lend much new insight on this debate. Future research may assess crisis from a more generalized perspective, or study follower perceptions of instability as a means of addressing the role crisis plays in charismatic attributions.

Conclusion

Followers’ role in the charismatic leadership relationship has been the subject of academic theory and empirical research for over three decades. Yet, the nature of the 2016 US presidential election provided a unique and fascinating context in which to study this multifaceted process. Perhaps more importantly, it underscored the critical importance for both followers and societies to understand these dynamics in a world characterized by global interconnections and local communities, social media, and the widespread disruption of truth and fact. By examining leader rhetoric, the relationships between follower characteristics and attributions of charisma and effectiveness, and the important role of perceived threat as a mediator, we move closer to answering elusive questions like, “How did we get here?” and “Who are we willing to follow?” Ultimately, these answers may simply reflect Thomas Jefferson’s cautionary warning, “the government you elect is the government you deserve.”

Footnotes

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.

Note

We would like to thank the anonymous reviewer for this suggestion.

Author biographies

Melissa K Carsten is a professor of Management at Winthrop University in Rock Hill, SC. Dr Carsten conducts research on followership, follower role orientations, and follower-centric approaches to leadership. She is specifically interested in and how followers interpret their role and contribute to the leadership process in organizations. Her research has been published in Leadership, The Leadership Quarterly, Journal of Organizational Behavior, and Journal of Occupational and Organizational Psychology among others. Dr Carsten serves on the Editorial Boards of The Leadership Quarterly, Journal of Organizational Behavior, and Group and Organization Management.

Michelle C Bligh is dean of the School of Social Science, Policy and Evaluation, and Professor of Organizational Behavior at Claremont Graduate University. Her research interests focus on organizational culture and the role of leaders in influencing and changing corporate cultures, particularly in post-merger organizations. In addition, her research interests include charismatic leadership, interpersonal trust, and political and executive leadership. Her work has been published in Journal of Applied Psychology, Leadership, Employee Relations, The Leadership Quarterly, Applied Psychology: An International Review, Group and Organization Management, Journal of Managerial Psychology, and The Journal of Business Ethics, and she was awarded the 2003 Sage Outstanding Paper Award for Research Methods.

Jeffrey C Kohles is a professor of Management and Organizational Behavior in the College of Business Administration, as well as Director and founding member of the Center for Leadership Innovation and Mentorship Building (CLIMB), at California State University San Marcos. Prior to his current position he was a Research Fellow at the Center for International Leadership in Buffalo, New York, as well as a Research Associate at the Kravis Leadership Institute in Claremont, California. His research interests include leadership, organizational and political communication, and the implementation of organization-level vision and strategy at the individual level.

Vienne Wing-Yan Lau is a PhD student in Organizational Behavior at Claremont Graduate University, where she received her MA degree in Psychology. Her research interests include leadership and followership, gender, and creativity.

References

Ahmadian

Azarshahi

Paulhus

(2017) Explaining Donald Trump via communication style: Grandiosity, informality, and dynamism. Personality and Individual Differences 107: 49–53.

Altemeyer

(1998) The other “authoritarian personality”. Advances in Experimental Social Psychology 30: 47–92.

Anderson

Gerbing

(1988) Structural equation modeling in practice: A review and recommended two-step approach. Psychological Bulletin 103(3): 411.

Ashkanasy

Dasborough

(2003) Emotional awareness and emotional intelligence in leadership teaching. Journal of Education for Business 79(1): 18–22.

Ashmore

Del Boca

Bilder

(1995) Construction and validation of the Gender Attitude Inventory, a structured inventory to assess multiple dimensions of gender attitudes. Sex Roles 32(11–12): 753–785.

Avolio

Bass

Jung

(1999) Re‐examining the components of transformational and transactional leadership using the multifactor leadership. Journal of Occupational and Organizational Psychology 72(4): 441–462.

Awamleh

Gardner

(1999) Perceptions of leader charisma and effectiveness: The effects of vision content, delivery, and organizational performance. The Leadership Quarterly 10: 345–373.

Baron

Kenny

(1986) The moderator–mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology 51(6): 1173.

Baumeister RF (1997) Esteem threat, self-regulatory breakdown, and emotional distress as factors in self-defeating behavior. Review of General Psychology 1(2): 145.

10.

Bass

Avolio

(1995) MLQ Multifactor Leadership Questionnaire. Redwood City, Ca: Mind Garden.

11.

Bentler

(2005) EQS 6 Structural Equations Program Manual. Encino, CA: Multivariate Software.

12.

Bligh

Kohles

Meindl

(2004) Charisma under crisis: Presidential leadership, rhetoric, and media responses before and after the September 11th terrorist attacks. The Leadership Quarterly 15(2): 211–239.

13.

Bligh

Kohles

Pillai

(2005) Crisis and charisma in the California recall election. Leadership 1(3): 323–352.

14.

Bligh

Robinson

(2010) Was Gandhi “Charismatic”? Exploring the rhetorical leadership of Mahatma Gandhi. The Leadership Quarterly 21: 844–855.

15.

Boal

Bryson

(1988) Charismatic Leadership: A Phenomenological and Structural Approach. In: Hunt JG, Baliga BR, Dachler HP and Schriesheim CA (eds) International leadership symposia series. Emerging leadership vistas (pp. 11-28). Lexington, MA, England: Lexington Books/D. C. Heath and Com

16.

Boin A, Hart PT, McConnell A, et al. (2010) Leadership style, crisis response and blame management: The case of Hurricane Katrina. Public Administration 88(3): 706–723.

17.

Campbell

(1990) Self-esteem and clarity of the self-concept. Journal of Personality and Social Psychology 59(3): 538.

18.

Campbell

Trapnell

Heine

et al . (1996) Self-concept clarity: Measurement, personality correlates, and cultural boundaries. Journal of Personality and Social Psychology 70(1): 141.

19.

Conger

(1991) Inspiring others: The language of leadership. Academy of Management Perspectives 5(1): 31–45.

20.

Conger

Kanungo

(1987) Toward a behavioral theory of charismatic leadership in organizational settings. Academy of Management Review 12: 637–647.

21.

Conger

Kanungo

(eds.) (1998) Charismatic Leadership in Organizations. Thousand Oaks, CA: Sage.

22.

Connolly

(2017) Trump, the working class, and fascist rhetoric. Theory & Event 20: 23–37.

23.

Coombs

Holladay

(1994) Speaking of visions and visions being spoken. Management Communication Quarterly 8(2): 165–189.

24.

Cox

Lienesch

Jones

(2017) Beyond economics: fears of cultural displacement pushed the white working class to trump. PRRI, May, 9.

25.

Damen

Van Knippenberg

(2008) Leader affective displays and attributions of charisma: The role of arousal. Journal of Applied Social Psychology 38(10): 2594–2614.

26.

Davis

(2017) Post-truth: Why We have Reached Peak Bullshit and What We can do about It. London: Little Brown Book.

27.

Davis

Gardner

(2012) Charisma under crisis revisited: Presidential leadership, perceived leader effectiveness, and contextual influences. The Leadership Quarterly 23(5): 918–933.

28.

Eagly

Karau

(2002) Role congruity theory of prejudice toward female leaders. Psychological Review 109(3): 573.

29.

Emrich

Brower

Feldman

et al . (2001) Images in words: Presidential rhetoric, charisma, and greatness. Administrative Science Quarterly 46(3): 527–557.

30.

Felfe

Schyns

(2006) Personality and the perception of transformational leadership: The impact of extraversion, neuroticism, personal need for structure, and occupational self-efficacy. Journal of Applied Social Psychology 36(3): 708–741.

31.

Gabriel

Paravati

Green

et al . (2018) From apprentice to president: The role of parasocial connection in the election of Donald Trump. Social Psychological and Personality Science. Epub ahead of print 2018. DOI: 1948550617722835

32.

Galvin

Waldman

Balthazard

(2010) Visionary communication qualities as mediators of the relationship between narcissism and attributions of leader charisma. Personnel Psychology 63(3): 509–537.

33.

Gardner

Avolio

(1998) The charismatic relationship: A dramaturgical perspective. Academy of Management Review 23(1): 32–58.

34.

Goldhill O (2017) The world is relying on a flawed psychological test to fight racism. Web blog post Retrieved October 23rd.

35.

Halverson SK, Holladay CL, Kazama SM, et al. (2004) Self-sacrificial behavior in crisis situations: The competing roles of behavioral and situational factors. The Leadership Quarterly 15(2): 263–275.

36.

Hamblin RL (1958) Leadership and crises. Sociometry 21(4): 322–335.

37.

Harms

Wood

Landay

et al . (2017) Autocratic leaders and authoritarian followers revisited: A review and agenda for the future. The Leadership Quarterly 29(1): 105–122.

38.

Hart

(1984) Verbal Style and the Presidency: A Computer-based Analysis. Orlando, FL: Academic Press.

39.

Hart

(1987) The Sound of Leadership: Presidential Communication in the Modern Age. Chicago, IL: University of Chicago Press.

40.

Hart

(2000a) DICTION 5.0: The Text-analysis Program. Thousand Oaks, CA: Scolari/Sage.

41.

Hart

(2000b) Campaign Talk: Why Elections are Good for Us. Princeton, NJ: Princeton University Press.

42.

Hart

(2001) Redeveloping DICTION: Theoretical considerations. In: West

(ed.) Theory, Method, and Practice of Computer Content Analysis. New York, NY: Ablex, pp.26–55.

43.

Hatfield

Cacioppo

Rapson

(1993) Emotional contagion. Current Directions in Psychological Science 2(3): 96–100.

44.

Helgeson

Palladino

(2012) Agentic and communal traits and health: Adolescents with and without diabetes. Personality and Social Psychology Bulletin 38(4): 415–428.

45.

Hogg

van Knippenberg

Rast

DE, III

(2012) The social identity theory of leadership: Theoretical origins, research findings, and conceptual developments. European Review of Social Psychology 23(1): 258–304.

46.

Holladay SJ and Coombs WT (1993) Communicating visions: An exploration of the role of delivery in the creation of leader charisma. Management Communication Quarterly 6(4): 405–427.

47.

Howell

Shamir

(2005) The role of followers in the charismatic leadership process: Relationships and their consequences. Academy of Management Review 30(1): 96–112.

48.

Itkowitz

(2016) This election has divided the country. Getting Clinton and Trump voters talking is one way to heal. The Washington Post, 14 November. Available at https://www.washingtonpost.com/news/ (accessed 24 April 2018).

49.

Judge

Erez

Bono

et al . (2002) Are measures of self-esteem, neuroticism, locus of control, and generalized self-efficacy indicators of a common core construct?. Journal of Personality and Social Psychology 83(3): 693.

50.

Kark R, Shamir B and Chen G (2003) The two faces of transformational leadership: Empowerment and dependency. Journal of applied psychology 88(2): 246–255.

51.

Kathlene

(1989) Uncovering the political impacts of gender: An exploratory study. Western Political Quarterly 42(2): 397–421.

52.

Kluckhohn

(1951) Values and value-orientations in the theory of action: An exploration in definition and classification. In Parsons T and Shils E (eds) Toward a general theory of action (pp. 388-433). Cambridge. MA: Harvard University Press.

53.

Ladkin

(2017) How did that happen? Making sense of the 2016 US presidential election through the lens of the leadership moment. Leadership 13(4): 313–412.

54.

Locke

(2003) Status and solidarity in social comparison: Agentic and communal values and vertical and horizontal directions. Journal of Personality and Social Psychology 84(3): 619.

55.

Meindl

(1990) On leadership-an alternative to the conventional wisdom. Research in Organizational Behavior 12: 159–203.

56.

Meindl

(1995) The romance of leadership as a follower-centric theory: A social constructionist approach. The Leadership Quarterly 6(3): 329–341.

57.

Meindl

Ehrlich

(1987) The romance of leadership and the evaluation of organizational performance. Academy of Management Journal 30(1): 91–109.

58.

Meindl

Ehrlich

(1988) Developing a Romance of Leadership Scale. Proceedings of the Eastern Academy of Management 30: 133–135.

59.

Meindl

Ehrlich

Dukerich

(1985) The romance of leadership. Administrative Science Quarterly 30: 78–102.

60.

Neuendorf

(2017) The Content Analysis Guidebook. 2nd ed. Los Angeles, CA: Sage.

61.

Pastor

Mayo

Shamir

(2007) Adding fuel to fire: the impact of followers' arousal on ratings of charisma. Journal of Applied Psychology 92(6): 1584.

62.

Pillai

Meindl

(1998) Context and charisma: A “meso” level examination of the relationship of organic structure, collectivism, and crisis to charismatic leadership. Journal of Management 24(5): 643–671.

63.

Popper M and Mayseless O (2007) The building blocks of leader development: A psychological conceptual framework. Leadership & Organization Development Journal 28(7): 664–684.

64.

Preacher KJ and Hayes AF (2008) Asymptotic and resampling strategies for assessing and comparing indirect effects in multiple mediator models. Behavior research methods 40(3): 879–891.

65.

Quillian

(1995) Prejudice as a response to perceived group threat: Population composition and anti-immigrant and racial prejudice in Europe. American Sociological Review 60(4): 586–611.

66.

Richardson

Simmering

Sturman

(2009) A tale of three perspectives: Examining post hoc statistical techniques for detection and correction of common method variance. Organizational Research Methods 12(4): 762–800.

67.

Rucker

Preacher

Tormala

et al . (2011) Mediation analysis in social psychology: Current practices and new recommendations. Social and Personality Psychology Compass 5(6): 359–371.

68.

Schildkraut

(2018) Young people in the American electorate. Journal of Race, Ethnicity and Politics 1: 4.

69.

Schwartz

Bilsky

(1987) Toward a universal psychological structure of human values. Journal of Personality and Social Psychology 53(3): 550.

70.

Schyns

Felfe

Blank

(2007) Is charisma hyper-romanticism? Empirical evidence from new data and a meta-analysis. Applied Psychology 56(4): 505–527.

71.

Shamir B, House RJ and Arthur MB (1993) The motivational effects of charismatic leadership: A self-concept based theory. Organization science 4(4): 577–594.

72.

Shamir B (1995) Social distance and charisma: Theoretical notes and an exploratory study. The Leadership Quarterly 6(1): 19–47.

73.

Short

Palmer

(2008) The application of DICTION to content analysis research in strategic management. Organizational Research Methods 11: 727–752.

74.

Steinmetz K (2017) The dictionary is adding an entry for ‘fake news.’ Time Magazine Online, September 27, 2017.

75.

Sydserff

Weetman

(2002) Developments in content analysis: A transitivity index and DICTION scores. Accounting, Auditing & Accountability Journal 15(4): 523–545.

76.

Tee

Paulsen

Ashkanasy

(2013) Revisiting followership through a social identity perspective: The role of collective follower emotion and action. The Leadership Quarterly 24(6): 902–918.

77.

Thoroughgood CN, Padilla A, Hunter ST, et al. (2012) The susceptible circle: A taxonomy of followers associated with destructive leadership. The Leadership Quarterly 23(5): 897–917.

78.

Trapnell

Paulhus

(2012) Agentic and communal values: Their scope and measurement. Journal of Personality Assessment 94(1): 39–52.

79.

Uhl-Bien M and Pillai R (2007) The romance of leadership and the social construction of followership. In: Shamir B, Pillai R, Bligh MC and Uhl-Bien M (eds) Follower-centered perspectives on leadership: A tribute to the memory of James R. Meindl. Greenwich, CT: Information Age Publishing, pp. 187–210.

80.

Vardy

(2016) The most divisive election in American history. Stock Investor, 8 November. Available at: https://www.stockinvestor.com/23720/divisive-election-american-history/ (accessed 24 April 2018).

81.

Wang

Luo

Niemi

et al . (2016) Catching fire via “Likes”: Inferring topic preferences of Trump followers on Twitter. Proceedings of the Tenth International AAAI Conference on Web and Social Media (ICWSM 2016), pp.719–722. May 17-20, 2016, Cologne Germany

82.

Weber

(1947) The Theory of Social and Economic Organization (Henderson AM and Parsons T, Trans.). New York, NY: Oxford University Press.

83.

Whisker

(2012) Apocalyptic rhetoric on the American religious right: Quasi-charisma and anti-charisma. Max Weber Studies 12: 159–184.

84.

Williams

Pillai

Deptula

et al . (2012) The effects of crisis, cynicism about change, and value congruence on perceptions of authentic leadership and attributed charisma in the 2008 presidential election. The Leadership Quarterly 23(3): 324–341.

85.

Williams

Pillai

Lowe

et al . (2009) Crisis, charisma, values, and voting behavior in the 2004 presidential election. The Leadership Quarterly 20: 70–86.

86.

Wojciszke B, Baryla W, Parzuchowski M, et al. (2011) Self-esteem is dominated by agentic over communal information. European Journal of Social Psychology 41(5): 617–627.

87.

Yagil

(1998) Charismatic leadership and organizational hierarchy: Attribution of charisma to close and distant leaders. The Leadership Quarterly 9(2): 161–176.

88.

Yukl G (1999) An evaluative essay on current conceptions of effective leadership. European Journal of Work and Organizational Psychology 8(1): 33–48.

89.

Yuthas K, Rogers R and Dillard JF (2002) Communicative action and corporate annual reports. Journal of Business Ethics 41(1–2): 141–157

90.

Zarefsky

(2004) Presidential rhetoric and the power of definition. Presidential Studies Quarterly 34(3): 607–619.