Does the shape of a world leader’s face predict their likelihood of starting a war? New research suggests that leaders with wider facial proportions tend to be more prone to initiating international military conflicts, especially during their first years in power. The findings, published in Politics and the Life Sciences, provide evidence that subtle biological cues might offer a window into the psychological tendencies that shape global politics.
The facial width-to-height ratio is a physical measurement of the human face. It is calculated by dividing the distance between the outer cheekbones by the distance from the top of the upper lip to the middle of the brow. Over the years, evolutionary psychologists have proposed that this specific facial metric serves as a biological signal of a person’s underlying predisposition toward dominance and physical formidability. In ancestral environments, wider faces may have reliably signaled fighting ability, teaching humans to subconsciously read these proportions as indicators of aggression.
Earlier research discovered that wider facial proportions correlate with aggressive behavior in everyday situations. For example, a 2008 study found that men with larger facial width-to-height ratios tended to act more aggressively in laboratory experiments and professional hockey games. Later research showed this pattern held for political leaders making military decisions. A 2022 study found that United States presidents with wider faces were more likely to use military force abroad.
Citizens also seem to respond to these cues. As indicated by a study covered by PsyPost in 2025, people who perceive an ongoing threat of conflict between groups are more likely to prefer political leaders who look and act dominant.
The new research expands this line of inquiry globally. “One of the real world problems that interests me is why some leaders are more conflict-prone than others,” Ross Miller, an associate professor of political science at the University of Nebraska-Lincoln, told PsyPost. “Even within the same states, leaders that confront similar security issues respond differently to threats. Why?”
While traditional political science often points to factors like a country’s economic power or strategic rivalries to explain wars, decisions to deploy military force fall to individual leaders. Past attempts to measure a leader’s aggressive tendencies usually relied on analyzing their speeches or biographies, which takes immense time and limits the number of leaders scientists can study. By using the facial width-to-height ratio, Miller sought to evaluate the aggressive predispositions of over a thousand world leaders to see if physical traits could predict international conflict across different cultures and eras.
“Traditional approaches emphasize important contextual factors – like power,” Miller explained. “For example, consider a traditional approach to this question: Why did Russia invade Ukraine? Because they are significantly more powerful than Ukraine, and figured they could force Ukraine to change its policy of joining NATO.” While these structural explanations are valid, he noted, his approach highlights that “it is also useful to take into account leaders’ qualities.”
To investigate this, Miller analyzed a massive dataset of 1,761 leaders from 174 countries over a 63-year period, from 1947 to 2010. This sample included nearly every major world leader from the post-World War II era. The idea to apply this metric to global conflict actually emerged from a parenting moment when Miller issued an empty threat to his daughter about throwing away her toys.
“That was a big threat, which I immediately regretted, because my daughter simply studied my face, checking to see if the resolve was there,” he recalled. “In class an idea formed: is there a literature on faces and conflict.” This realization led him to previous research on facial proportions and aggression, prompting him to spend his time during the COVID-19 pandemic compiling photos of world leaders.
To measure facial proportions, the researcher collected high-quality photographs of these leaders where they were facing the camera directly with a neutral expression. Using digital imaging software, Miller measured the width and height of each face to calculate the exact ratio. These facial measurements were then compared to a historical database of militarized interstate disputes.
This database records instances where leaders threatened, displayed, or actually used military force against another country. To ensure the results were not simply capturing other political dynamics, the analyses controlled for a variety of alternative factors. These included the country’s military power, geographic proximity to adversaries, whether the state was a democracy, and the leader’s age and gender.
The results indicated that leaders with higher facial width-to-height ratios were more likely to initiate international conflicts than leaders with narrower faces. Specifically, this effect was highly dependent on how long the leader had been in power. In the first year of their tenure, leaders with wider faces were roughly three times more likely to initiate a conflict compared to their peers with narrower faces.
“In my reading, the key takeaway from this work is that there seems to be an association between the [facial width-to-height ratio] of elected leaders and, particularly in their early tenure, the likelihood of initiating an international conflict,” said Alex Jones, an associate professor at Swansea University who was not involved in the study.
“Conflict is a rare event in IR [international relations], so the statistical, observation models often have difficulty making strong predictions because of that rarity,” Miller explained. “My results indicate that high-FWHR leaders are over three times as likely as low-FWHR leaders to initiate conflict. But we have to place that in context that we are starting from a low baseline – a low probability event. Tripling a small probability still leaves a pretty small probability. Just because a high-FWHR leader is in place does not mean automatic conflict initiation.”
While a relative threefold increase sounds substantial, the absolute number of conflicts initiated remains very small. Because full-scale military actions are so uncommon, minor shifts in probability can appear magnified when expressed as relative percentages.
Jones agreed that the actual increase in risk is slight. “The graphs highlight a change in probability of something like .02,” he told PsyPost, noting that in extremely large samples, almost all comparisons become statistically significant. “I’d assume that such a small effect probably doesn’t mean much in practical terms.”
This surge in aggression early in a leader’s time in office aligns with the idea that new leaders often use military force to establish their reputation and show adversaries that they will not back down. Once their aggressive tendencies become known to the international community, the need to prove themselves diminishes. Opposing countries may also actively avoid provoking leaders they know to be highly aggressive. As a result, the influence of facial proportions on conflict initiation faded the longer a leader remained in power.
“I was surprised that the latitude to initiate conflicts that leaders enjoy early in their tenure declines so rapidly,” Miller said. “By year three in power there is not much difference between low- and high-FWHR leaders. I expected it to diminish as other states adapt, but not that quickly.”
The findings are in tension with research covered by PsyPost in 2026, which questioned the validity of the facial width-to-height ratio as an evolutionary marker of dominance. That earlier study found that, when adjusting for body size, men in the general public did not actually have larger facial width-to-height ratios than women. However, it is worth noting that the earlier research focused on anatomical differences between sexes in the general population, whereas the new study specifically measures behavioral outcomes and conflict decisions among elite political leaders.
Jones also pointed out that “the assumption that [facial width-to-height ratio] is related to testosterone… has been tested a few times with not much support.” He warned against assuming a direct biological pipeline. “I think the clear painting of the results and a first reading is that the causal chain goes something like ‘hormones -> appearance -> behaviour -> conflict’,” he said. “I don’t think this provides good evidence for that.”
As with all research, there are some caveats to consider. First, the facial width-to-height ratio is just one single physical characteristic. It does not account for other facial cues that people use to judge aggression, such as the width of the chin or the shape of the eyebrows. Second, measuring facial proportions from photographs can sometimes be inexact, though the researcher took steps to standardize the measurements by using multiple photos when possible.
This measurement challenge is significant, according to Jones. “[Facial width-to-height ratio] is a notoriously tricky measure,” he noted. It is “affected by all sorts of simple things like distance to the camera, head orientation, and so on.” He explained that his own 2025 research found the metric is highly sensitive to the vagaries of photography. “With a ‘wild’ dataset like this, it’s surely unstandardised and injects a lot of noise,” Jones added.
There is also a possibility that countries facing severe security threats purposefully select leaders who appear more dominant and aggressive. If a nation is already on the brink of war, voters or ruling elites might naturally elevate a leader with a wider face. This makes it difficult to completely separate the leader’s inherent aggression from the country’s existing path toward conflict.
“One reason I find these results interesting is that they complement work showing that people become more attracted to dominant-looking leaders when they perceive greater intergroup conflict,” Adi Wiezel, an assistant professor of psychology at Elon University, told PsyPost.
“Taken together, these findings raise interesting questions about the relationship between conflict, whom people select as leaders, and how those leaders subsequently behave,” continued Wiezel, who directs the Wiezel Behavior, Opinions, and Relations between Groups (BORG) Lab and was not involved in the research.
This potential bias in impressions means that a finding like this could emerge even if there is no inherent biological link between a leader’s appearance and their actual behavior. If a restless, conflict-prone nation intentionally chooses an intimidating leader, the environment itself may be the true cause of any subsequent wars.
“Whether that leader initiates action is due to a lot of factors, and if there’s no link between behaviour and appearance, then this finding might emerge as a function of the bias in impressions,” Jones said.
Wiezel cautioned against drawing definitive causal conclusions from observational data. “Facial appearance can shape how other people perceive and respond to us, and those reactions can themselves influence behavior and opportunities over time,” she explained. “The study identifies a provocative association across a very large historical sample, but it cannot isolate which of these possible pathways produced it.”
The dataset also lacked controls for the race or ethnicity of the leaders, as comprehensive historical data on these factors across 174 countries does not currently exist. Different populations can have varying average facial structures, which could influence the broader patterns. Future research might explore how these facial metrics interact with a leader’s personal history, such as whether they experienced childhood trauma or served in the military.
Wiezel would like to see future research examine how a leader’s age and gender affect this relationship. While the new study found no difference between men and women, she noted that “the relatively small number of conflicts initiated by women limits how definitive that conclusion can be.” Age could also be an important factor, she added, because “some of the theory linking facial characteristics to aggression emphasizes physical formidability and intrasexual competition. This raises the possibility that such processes may matter more among younger men.”
For his part, Jones suggested using a bottom-up cognitive psychology approach that decomposes facial images into various sources of variability. “I’d like to see a more holistic approach to examining facial appearance and its link with behaviour undertaken,” he said, which “opens up novel theory driven approaches.”
Additionally, scientists hope to examine whether wider-faced leaders are less likely to engage in international cooperation and peace-building, providing a more complete picture of their foreign policy behaviors.
“I am working on another paper on U.S. interventions in conflicts guided by FWHR,” Miller said.
The study, “Looks that kill: Facial metrics and international conflict 1947–2010,” was authored by Ross Miller.