Liberals give dominant leaders lower odds of success than conservatives do, study finds

Political liberals frequently express shock when forceful, domineering figures win major elections or succeed in the business world. A new series of studies published in Political Psychology suggests this surprise may reflect a psychological tendency that goes beyond party loyalty. Across several studies, liberals tended to see dominant leaders as less fair and caring than conservatives did, liked them less, and gave them lower odds of success, a pattern the authors suggest could make liberals more complacent at the ballot box.

When people rise to positions of power in workplaces or governments, they generally rely on two broad strategies: prestige or dominance. Prestige-oriented leaders earn respect by sharing their expertise, cooperating with others, and treating group members fairly. Dominance-oriented leaders, by contrast, pursue status through assertiveness, intimidation, and the forceful exercise of authority to get their way. Evidence from a 2012 study of group hierarchies indicates that while people generally dislike dominant individuals, intimidation and force remain just as effective as earned respect for gaining actual influence over a group’s decisions.

Forceful leaders often surge in popularity when the public feels anxious or threatened. For instance, a 2017 study of 69 countries co-authored by one of the new paper’s researchers, Hemant Kakkar, found that during periods of economic instability, such as rising unemployment or poverty, people feel a lack of personal control and become more inclined to support domineering leaders over respected, collaborative ones. Personal worldviews also shape how people react to aggressive authority. Other research on support for harsh leaders indicates that people who view the social world as a ruthless competition tend to judge harsh, antagonistic leaders as competent and effective, whereas those with a cooperative worldview see the same behavior as ineffective.

These worldview differences often map onto the political spectrum. In a 2019 survey experiment, right-wing voters showed a stronger preference for dominant political candidates than left-wing voters did, largely because conservatives viewed forceful candidates as more competent at handling conflict.

Psychologists also explain ideological divides through moral foundations theory, a framework proposing that people base their moral judgments on distinct sets of core values. In a study of moral priorities, liberals ranked individual-welfare values like fairness and protecting others from harm at the top, whereas conservatives placed greater weight on group-binding values such as loyalty, respect for authority, and purity (a concern with avoiding things seen as degrading or contaminating).

Free daily newsletter

While past work explains why conservatives often gravitate toward strongmen, less is known about how liberals forecast the trajectories of such figures. To explore this question, Danbee Chon, an assistant professor of management at the University of South Florida’s Muma College of Business, teamed up with Hemant Kakkar, the Dean’s Chair and an associate professor of management and organization at the National University of Singapore’s NUS Business School. They noticed that liberal commentators and citizens repeatedly predicted failure for forceful politicians like Donald Trump and business executives the authors describe as dominant, such as Elon Musk and Jeff Bezos.

“The study was broadly motivated by a real-world phenomenon where we observed that political liberals tend to underestimate leaders who come across as dominant, assertive, or at times authoritarian,” Chon and Kakkar told PsyPost. “The most striking example of this for us was the 2016 US election, where many people assumed that Trump would lose the election, and even on the eve of the election, outlets like The New York Times gave Clinton an 85% chance of winning. However, this was not just limited to the 2016 election. Similar observations were made in the 2024 election and even outside the US, where the Remain campaign in the UK underestimated the appeal and effectiveness of the Brexit campaign, particularly the influence of figures such as Nigel Farage and Boris Johnson.”

These recurring miscalculations prompted the researchers to test, across five main studies and four supplemental ones, whether a deeper psychological pattern was at play. Chon and Kakkar suspected that when a leader uses coercion and authority, liberals see a violation of their core values of fairness and harm avoidance, causing them to dislike the leader. Because people often let their emotional impressions cloud their predictions of the future, a tendency psychologists call the affect heuristic, this moral dislike could lead liberals to lowball a dominant leader’s chances.

To test this idea outside of partisan politics, the researchers first examined how people evaluate their actual bosses in the workplace in Study 1. They recruited 177 full-time employees in the United States, with an average age of 47 and nearly 23 years of work experience, using a two-part survey spaced one week apart. In the first survey, employees rated their immediate supervisor’s dominance and prestige behaviors on a 1-to-7 scale and reported their political views on a single 7-point scale (known as a Likert scale) from very liberal to very conservative.

One week later, the employees rated how much their boss valued fairness and avoiding harm, how warm and likable their boss was, and how likely their boss would be to win a workplace promotion against three other candidates. The researchers also accounted for employee demographics, job tenure, and the supervisor’s prestige score.

The results showed a consistent ideological split when bosses behaved dominantly. As a supervisor’s dominance score rose, liberal employees rated their boss as less committed to fairness and harm avoidance, less likable, and less likely to be promoted than conservative employees did. When supervisors scored low on dominance, the differences between liberal and conservative employees on fairness, likability, and promotion odds were not statistically significant.

Statistical modeling traced a chain: liberals saw dominant bosses as less fair and compassionate, this was associated with liking them less, and lower liking was in turn associated with lower expectations of promotion.

Because workplace surveys cannot establish cause and effect, the authors next conducted a pair of controlled experiments (Studies 2A and 2B) testing the two steps of this pathway in a political context. In Study 2A, 281 online participants read a fictional scenario about a small-town mayoral candidate named Riley, with no political party mentioned. Half of the participants read that Riley was a dominant leader who was assertive and seized control of situations, while the other half read that Riley was a prestigious leader who shared expertise and earned admiration, though both versions described Riley as competent. Liberals who read about the dominant version of Riley rated the candidate as less caring, less fair, and less likable than conservatives did, whereas ideological differences for the prestigious candidate were not statistically significant.

In Study 2B, involving 589 participants, the researchers tested whether this reduced likability was linked to predictions of electoral victory, even when small sums of real money were at stake. Participants viewed a simulated social media post from a mayoral candidate featuring quotes that projected either dominance or prestige, and then rated the candidate’s likability and chances of winning. The researchers also gave everyone a 10-cent bonus and allowed them to bet any portion of it on the candidate winning the election, promising to double whatever amount they wagered if the candidate succeeded.

Compared to conservatives, liberals rated the dominant candidate as less likable, estimated their chances of winning as lower, and bet less of their bonus money on the candidate’s victory, while no meaningful ideological gap appeared for the prestigious candidate. Mediation analyses showed that lower likability statistically accounted for part of liberals’ lower success estimates and smaller bets on the dominant contender.

“The critical point for an average person is to understand that our underlying values can color the way we see someone as a prospective leader,” the researchers explained. “If our values are in conflict, we tend to dislike the individual, leading to underestimation. Likability thus acts as a halo and pushes us into wishful thinking that such leaders may not be successful, while that may not be the case as we saw with the election examples above.”

The halo effect is the tendency for an overall impression of someone to spill over into unrelated judgments about them. In one of their supplemental studies, the researchers also found that wishful thinking helped explain the statistical link between disliking a dominant leader and giving them lower odds of success.

One possible alternative explanation was that liberal participants simply assumed any domineering politician must be a conservative, and bet against them out of partisan rivalry. To rule this out, the authors ran Study 3 with 1,042 participants who evaluated the same social media campaign posts in an explicitly nonpartisan city government election. Even with party politics removed, liberals still disliked the dominant candidate more than conservatives did and gave the candidate lower odds of winning. The direct ideological gap in betting amounts fell just short of statistical significance, though the indirect pathway through reduced likability remained reliable.

The researchers also found a statistical pathway running from dislike to lower success estimates to complacency, which they measured as the priority participants placed on voting and their intention to vote in the election. Liberals and conservatives did not differ directly on these voting measures, however.

“Once liberals underestimated such leaders’ success, it affected their tendency to take actions such as prioritizing voting or exercising their civic duties because in their minds, such leaders would be unsuccessful, so what’s the point,” Chon and Kakkar noted. “Critically, such complacency on their part, if it translates into lower vote participation, could increase the chances of these individuals succeeding even more.”

In Study 4, Chon and Kakkar examined how stubborn this expectation gap remained when people were given explicit polling information. A group of 788 participants first viewed campaign ads for both a dominant candidate and a prestigious candidate running against each other in the same race. After rating both contenders, participants were divided into three groups and shown local news exit polls indicating that the dominant candidate had either a 45.13%, 50.13%, or 55.13% chance of winning the election. Participants then decided how much of a 10-cent bonus to bet on the dominant candidate winning.

When comparing the two candidates side by side, liberals again judged the dominant contender as less likable and less likely to win than conservatives did. In this head-to-head format, liberals also liked the prestigious candidate more than conservatives did and expected them to be more successful, which the authors note they had not consistently observed in their other studies and suggest may have resulted from the forced comparison.

When shown the exit polls, liberals wagered less money on the dominant candidate than conservatives did not only when the candidate trailed at 45.13%, but even when the poll showed the candidate slightly ahead at 50.13%. Only when the exit poll gave the dominant candidate a more substantial lead of 55.13% did liberals raise their bets enough that the overall betting difference between liberals and conservatives was no longer statistically significant.

Even when the dominant candidate held that 55.13% lead, statistical mediation models indicated that liberals’ personal dislike of the candidate was still linked to smaller bets. “Even in that condition, despite the lack of a direct effect, we still observed an indirect effect on liberals’ tendency to underestimate the success of such leaders,” the authors said. “This, for us, was quite a surprising finding.”

An internal meta-analysis, which statistically combined the results of their studies, found that the gap between liberals and conservatives was modest in size but highly consistent from study to study. Supplemental studies, including an analysis of World Values Survey data from multiple countries, pointed in the same direction.

“For us the most surprising finding has been the robustness of this effect,” Chon and Kakkar told PsyPost. “This effect replicated in different settings, such as political or election scenarios beyond partisanship, but also among employees rating their leaders whose political affiliation should be irrelevant in such a context.”

These results are in line with earlier research on preferences for authoritarian-looking leaders, which linked conservative ideology and dogmatic thinking to a preference for dominant facial features, and linked liberal views to greater trust in information sources. That earlier work measured personal voting preferences based on appearance, whereas the new study focuses on how much people like dominant leaders and how likely they think those leaders are to succeed.

As with all research, there are some caveats to consider when interpreting these findings. “One important caveat to mention would be that our study was not able to truly capture the actual level of underestimation,” Chon and Kakkar explained. “For that to happen, we would have required participants’ estimated odds of success versus a leader’s actual success rate. Hence, what we capture is a tendency to underestimate rather than an actual demonstration of underestimation bias since we don’t have an objective comparison.”

Because the studies measured relative differences between liberals and conservatives rather than tracking actual election or promotion results, the data cannot establish whose predictions are objectively more accurate in the real world. In the exit poll experiment, for instance, conservatives’ bets stayed about the same whether the candidate’s odds were 45% or 55%. The authors suggest this may reflect conservatives overweighting a dominant leader’s chances, so the gap between the groups could reflect conservative overconfidence as well as liberal skepticism.

There are a few other things to keep in mind regarding how these psychological steps unfold. The researchers measured whether participants believed a leader valued fairness and harm avoidance, rather than directly measuring the participants’ own moral values. In addition, because the links between moral impressions, likability, and success predictions were analyzed statistically rather than manipulated one by one, people might form their emotional dislike and their doubts about a leader’s viability at the exact same time.

Several items on the complacency measure also referred to voting for the candidate in question, so the data do not show whether opponents of a dominant leader become less likely to vote against them. Also, the political experiments used fictional candidates and 10-cent bets, whereas real elections involve party cues, media coverage, and higher stakes that could widen or narrow the gap.

Conservatives’ evaluations of prestige-oriented leaders also varied across some of the experiments, suggesting that other unmeasured factors may shape how people react to collaborative leaders. Future studies could directly compare predictions against real-world election outcomes and test whether lower expectations actually reduce turnout among people who oppose dominant leaders.

“The long-term goal here is to understand how political ideology and values color our judgment and support for certain kinds of leaders and how that affects our psychology and behaviors,” the researchers added. “This study was very much focused on dominant leaders, and ideology was mostly measured using a single Likert-scale item. We hope to consider various facets of political ideology based on economic, social, and other factors in studying this broader question rather than classifying people broadly in terms of liberals and conservatives.”

The study, “A blind spot for dominance? Why liberals underestimate dominant leaders more than conservatives,” was authored by Danbee Chon and Hemant Kakkar.