Evolutionary psychology’s “macho” face ratio theory has a major flaw

For years, evolutionary psychologists and biologists have investigated the idea that the shape of a man’s face can predict his behavior. A specific measurement known as the facial width-to-height ratio has garnered attention as a potential biological billboard for aggression and dominance. A new comprehensive analysis, however, challenges the validity of this metric.

The research suggests that this specific ratio is not a reliable marker of sexual difference. Instead, the study points toward a simpler measurement that may hold the key to understanding facial evolution. These findings were published in the journal Evolution and Human Behavior.

The human face is a complex landscape that conveys biological information to others. We instinctively look at faces to judge health, age, and emotion. Beyond these immediate signals, researchers have hypothesized that facial structure reveals deeper evolutionary traits. The primary metric used to test this is the facial width-to-height ratio, often abbreviated as fWHR. To get this number, a researcher measures the distance between the cheekbones and divides it by the distance between the brow and the upper lip.

The prevailing theory has been that men with wider, shorter faces possess higher levels of testosterone and are more formidable. Previous studies have linked a high ratio in men to aggressive behavior in sports and financial success in business. The underlying assumption is that this facial structure evolved because it signaled a competitive advantage to potential mates or rivals. This concept relies on the existence of sexual dimorphism, which is the condition where the two sexes of the same species exhibit different characteristics.

Despite the popularity of this theory, the scientific evidence has been inconsistent. Some studies find a strong link between the ratio and masculine traits, while others find no connection at all. A major issue in past research is the inconsistent definition of the ratio itself. Different scientists measure the height of the face using different landmarks, such as the eyelids, the brow, or the hairline. Furthermore, many studies fail to account for the overall size of the person.

Free daily newsletter

To address these inconsistencies, a team of researchers led by Alex L. Jones from the School of Psychology at Swansea University conducted a rigorous re-examination of the evidence. The team included Tobias L. Kordsmeyer, Robin S.S. Kramer, Julia Stern, and Lars Penke. They aimed to apply a more sophisticated statistical approach to determine if the facial width-to-height ratio is truly a sexually dimorphic trait. They also sought to determine if simple facial width might be a more accurate signal of biological differences than the ratio.

The researchers utilized a statistical method known as Bayesian inference. This approach differs from traditional statistics by incorporating prior knowledge into the analysis. It allows researchers to estimate the probability of a hypothesis being true given the available data. This contrasts with standard methods that often focus solely on whether a specific result is statistically significant. The team argues that Bayesian models are better suited for understanding subtle biological patterns because they can simulate data and quantify uncertainty.

In their first study, the group analyzed facial photographs of 1,949 individuals drawn from nine different datasets. The sample included 818 men and 1,131 women from various Western countries. The researchers used computer software to automatically place landmarks on the facial images. This ensured that the measurements were consistent across all photographs. They calculated the width-to-height ratio using five different common definitions of facial height to see if the measurement method mattered.

Crucially, the team controlled for body size in their statistical model. They adjusted the data for both height and weight. This is a vital step because men are generally larger than women. Without this control, a feature might appear to be a specific facial signal when it is actually just a byproduct of having a larger body. The researchers also defined a “region of practical equivalence.” This is a statistical tool used to determine if a difference is large enough to matter in the real world.

The results of this first analysis contradicted the popular evolutionary theory. When controlling for height and weight, the researchers found that men did not have a larger width-to-height ratio than women. In fact, the model showed a small tendency for women to have a larger ratio. However, this difference was so minute that it fell within the region of practical equivalence. This means the difference was effectively zero for any practical purpose.

The study also revealed that the ratio is heavily influenced by general body geometry. The researchers found that as a person’s height increases, their facial width-to-height ratio tends to decrease. Conversely, as body weight increases, the ratio tends to increase. This suggests that previous findings linking the ratio to aggression might have actually been detecting differences in body mass index rather than specific facial architecture. The researchers argue that the ratio is not a standalone signal of masculinity.

Following these results, the team conducted a second study focusing solely on the width of the face. This measurement is known technically as bizygomatic width. It is the distance between the two zygions, or the most outer points of the cheekbones. The researchers hypothesized that raw width might be the sexually selected trait that earlier scientists were trying to capture with the ratio.

For this second analysis, they examined the same large dataset of photographs. They also analyzed a smaller subset of 305 individuals for whom they had detailed measurements of upper body size. This included shoulder width, chest girth, and arm girth. This allowed them to test if facial width is connected to muscularity and physical strength, which are key components of evolutionary dominance.

The findings for facial width were starkly different from those for the ratio. The Bayesian analysis showed a very high probability that men have wider faces than women. This held true even when the researchers adjusted for height and weight. The difference was substantial, amounting to roughly half a standard deviation.

When the researchers looked at the smaller group and controlled for upper body size, the distinction became even clearer. The model indicated that men have almost a two standard deviation greater face width than women. The analysis suggested that an individual man has a 99.9 percent probability of having a wider face than a woman of similar body composition. This indicates that facial width is a robust, sexually dimorphic trait.

The authors propose that the evolutionary signal is driven by the lateral growth of the cheekbones. During puberty, male faces tend to grow wider, a process likely driven by testosterone. This growth trajectory aligns with the development of other skeletal features associated with physical formidability. The study implies that the horizontal width of the face is a reliable indicator of physical size and strength.

There are caveats to this research. The study relied on static two-dimensional photographs. This method cannot capture the dynamic nature of facial expressions or the three-dimensional structure of the skull as effectively as medical imaging. Additionally, the samples were primarily from Western populations. It is possible that facial metrics vary across different ethnic groups and environments. Future research would need to verify these findings in more diverse global populations.

The researchers also noted that facial perception is complex. While physical measurements provide hard data, human social interaction relies on how these features are perceived. It remains to be seen if the human brain specifically attends to raw width when making judgments about dominance or threat. The current study focuses on the physical reality of the face rather than the psychological processing of it.

This research represents a methodological correction for the field of evolutionary psychology. By using advanced Bayesian statistics and proper body size controls, the authors have dismantled a widely held belief about the facial width-to-height ratio. They argue that the ratio is likely a statistical artifact rather than a meaningful biological signal.

The shift in focus toward bizygomatic width offers a clearer path for future investigation. If facial width is the true signal of formidability, previous studies on aggression and leadership may need to be re-evaluated. The authors suggest that researchers should move away from the ratio and focus on simple width in future work. This simplification may lead to more consistent and replicable results in the study of human evolution.

The study, “Updating evidence on facial metrics: A Bayesian perspective on sexual dimorphism in facial width-to-height ratio and bizygomatic width,” was authored by Alex L. Jones, Tobias L. Kordsmeyer, Robin S.S. Kramer, Julia Stern, and Lars Penke.