[Proposal] In Awe of Evil Kings: Appraisal of awe in non-virtuous leaders promotes tolerance of moral violations


We are awe geeks collaborating across 3 universities: Sean Goldy (University of California Irvine), Yachen Li (University of Toronto), and Ji Song (University of Melbourne). Thanks in advance for reading!

Background and Significance

“I could stand in the middle of Fifth Avenue and shoot somebody, and I wouldn’t lose any voters, OK?” (Trump, 2016)

Why is it that we let some leaders get away with bad behaviour (Marti, 2019), when we normally hold our leaders to a higher moral standard (Walter & Redlawsk, 2019)? The emotion of awe may hold the key. People can feel awe when they experience something vast (e.g., tremendous skill) that challenges their current understanding of the world, prompting deference, social connection, and reduced self-importance. Given that awe can serve as a catalyst for the formation and strengthening of social collectives (Bai et al., 2017; Keltner & Haidt, 2003), leaders who inspire awe may more readily amass followers than others. However, such social influence can have downsides. Is there something about being awe-struck by these kinds of leaders that we also judge them by a different moral standard than more ordinary folks? It may be that there is a dark-side to awe-inspiring leadership, and its consequences for moral accountability, that has yet to be fully examined empirically.

Awe-inspiring leadership may cause this kind of ethical fading, where unethical behaviours are overlooked or minimized, because of the way awe prioritizes our binding to the broader social collective and submission to the social hierarchy. Keltner and Haidt (2003) have suggested that, deep in our evolutionary past, a primordial awe hardwired an instinct for submission towards powerful individuals, and in this way awe played the social function of maintaining our social hierarchies. Indeed, studies have found that awe is a self-transcendent emotion, which shifts our focus from narrow self-interests to the interests of the broader collective (Bai et al., 2017; Stellar et al., 2017). And the evolutionary drive to maintain the social hierarchy around awe-inspiring leaders may have been so powerful that, as an evolutionary by-product, it also led to a greater tolerance for the moral violations of those awe-inspiring leaders. In this way, by reducing the perceived need to challenge awe-inspiring leaders on the basis of ethical violations, it may have further stabilized group cohesion around those leaders.

While the relationship between awe-inspiring leadership and tolerance for ethical violations remains to be directly examined; the more general leadership literature has shown that people are more tolerant of moral violations from leaders who are perceived to be initially virtuous (Wang & Chan, 2019), charismatic (Conger & Kanungo, 1998), or transformational (Bass & Steidlmeier, 1999). These leadership traits can be thought of as lying on the spectrum towards awe-inspiring leadership (Sy, Horton, & Riggio, 2018), and we expect the tolerance for ethical violations to only get stronger as these traits approach awe-inspiring levels.

We believe these questions are pertinent to our times, and that our initial research project will be an important step towards understanding leadership in our current socio-political climate.


Specifically, our study aims to answer the following questions:

(H1) People will show (a) greater tolerance for moral violations and (b) a greater willingness to follow awe-inspiring leaders, compared to more typical leaders.

(H2) To what extent are the relationships in (H1) moderated by the characteristics of the people who feel awe towards these kinds of leaders (social dominance orientation, dark triad, need for belonging, and anomie)?

(H3) To what extent do the relationships in (H1) and (H2) generalize from awe-inspiring leaders to other kinds of awe-inspiring public figures (e.g. athletes, artists)?


We propose first a set of pilot studies to develop and validate 4 leadership vignettes, followed by 2 online experiments to examine our hypotheses.

Pilot studies

Building on the theoretical and qualitative research around ethical (vs. unethical) charismatic leadership (Conger & Kanungo, 1998) and moral character in authentic (vs. inauthentic) transformational leadership (Bass & Steidlmier, 1999), we will first develop four leadership vignettes. These vignettes will be used in our experiments to induce the perception of an (1) awe-inspiring political leader, (2) typical political leader, (3) awe-inspiring public figure (non-political), or (4) typical public figure (non-political).

We will primarily aim to develop vignettes that can create differences in perceptions of awe while controlling for confounding variables (e.g., status, physical attractiveness, power).

Study 1

Study 1 examines whether people are more tolerant of moral violations from awe-inspiring leaders relative to more ordinary leaders. Participants will first complete a set of questionnaires measuring their social dominance orientation, dark triad, need for belonging, and enlightened compassion. Participants will then be randomly assigned to read about either an awe-inspiring or typical political leader, using the respective vignettes validated in the pilot studies. Next, participants will indicate their willingness to follow this leader and how wrong it would be for the leader to engage in a series of hypothetical moral foundation violations (Walter & Redlawsk, 2019). This study will allow us to test whether people are indeed more tolerant of transgressions by awe-inspiring leaders, in addition to being more willing to follow them, relative to more typical political leaders.

Study 2

Study 2 will adopt the same design as Study 1, and test whether the aforementioned effects extend to other awe-inspiring public figures such as artists, athletes and social media influencers. Participants will read vignettes in which an awe-inspiring public figure, or a non-awe-inspiring public figure committing the same moral violations as in Study 1. Next, participants will indicate their willingness to follow this public figure and how wrong it would be for them to engage in a series of hypothetical moral violations.

Sample size and costs

Prior to running the proposed studies, we will pilot our vignettes for each of the two studies to ensure they elicit the anticipated level of awe for each hypothetical leader and public figures. We plan to recruit 120 participants for each pilot study, with a total of 240 participants. For a 10-minute pilot study, we will pay £1.40 each, which comes to £224 in total.

We aim for a sample size of 714 for Study 1 (N = 357 per condition), and a sample size of 879 (293 per condition) for Study 2. The minimum total sample size will be 1593. This will allow us to detect a small effect size with between-subject designs (Cohen’s d = 0.21, 80%, error probability of 0.05). To account for roughly 10% of careless responses, we will recruit 1770 participants in total. For each 20-minute experiment, we will pay £2.60 each, which comes to £6136. The estimated total cost of the proposed project will be £6360.

Pre-registration and Open science

Our hypotheses will be pre-registered at AsPredicted.org (links to follow). All study materials (e.g. vignettes, R code) and datasets will be made available at OSF.io (we will update the community when we are finished!).