The Uncomfortable Truth About Police Promotion: How Facial Traits Predict Who Gets Promoted

Estimated reading time: 11 minutes

Looks Shouldn’t Matter. But They Do.

There’s a reason certain people get called “a natural leader” before they’ve led anything. They walk into a room, and something clicks. People move out of their way, defer to their opinions, assume they’re in charge. It’s not always about what they’ve done, sometimes it’s just the face. We’ve all met that person. Some of us have quietly wondered if we are that person or painfully suspected we’re not.

Outside of policing, the research on this is well established. Studies have documented a so-called “beauty premium“: a tendency for people perceived as attractive or leader-like to earn more, advance faster, and be rated as more competent and trustworthy by others. It shows up in corporate hiring, salary negotiations, and political elections. Fine. That’s the private sector. That’s politics. Those processes are notoriously subjective.

But law enforcement is different, right? Promotions to sergeant, lieutenant, and beyond aren’t handed out based on looks. Agencies spend serious time and money on written exams, assessment centers, interview panels, performance reviews, and structured evaluation criteria. These processes exist precisely to make sure that the best officer gets the badge with extra brass on it.

Right?

Well… A peer-reviewed study published by researchers at the University of South Carolina just put that assumption through its paces and the results are going to make you think twice about your next promotion board. Researchers took archival photographs of police recruits from decades ago and showed them to college students with zero law enforcement background. No resume. No test scores. No interview. Just a face. What those students were able to predict about those recruits’ careers is striking.

Here’s what the research actually shows and what it means for your agency.

The Study: Faces, Dyads, and Police Promotion

The researchers, Ian Adams, Scott Mourtgos, Christopher Simon, and Nicholas Lovrich, built their study around a straightforward but revealing design. They pulled 98 archival academy photographs of police recruits from a capital city police agency. These photos were taken between 1988 and 2008. Nothing in the images identified the people as police officers such as uniforms, insignia, or badges. Just headshots.

They then recruited 507 adult college students across the United States to serve as respondents. These participants had no connection to law enforcement. The study ran in two phases.

Phase 1: Rating the Face

In the first phase, respondents were each shown a random selection of those headshots which included anywhere from 2 to 63 photographs. For each photo, they were asked to rate the pictured person on six facial trait dimensions using a 7-point scale:

• Attractiveness
• Trustworthiness
• Distinctiveness
• Dominance
• Masculinity
• Secureness

They were also asked to rate the person’s overall leadership ability. Again, no context, no background, no explanation. Just a face and a scale.

Here’s what it looked like in practice: a headshot of a young recruit, just the face and the top of a shirt collar, with nothing to identify who they are or what they do. A respondent pulls up that photo and rates how trustworthy, dominant, or leader-like the person appears. That’s the entire task. Multiply that across 507 respondents and you get roughly 15,000 individual observations.

Phase 2: The Head-to-Head

In the second phase, the same respondents were shown pairs of photographs with two recruits side by side and asked a single question: Which of these two people would make a better leader?

Here’s the critical design detail: every pair included either an officer who was never promoted alongside one who eventually made lieutenant, or a sergeant alongside a lieutenant. Respondents saw up to 24 of these side-by-side comparisons. In total, 5,739 paired comparisons were analyzed.

Their picks were then cross-referenced with the actual promotional records of those officers. The recorded careers that unfolded over the 20-plus years after those academy photos were taken.

The respondents had no idea this is what was happening. They thought they were rating leadership potential. What they were actually doing was predicting police careers.

The Findings: Uncomfortable, But Hard to Dismiss

Let’s start with the headline number: an AUC score of 0.70.

AUC stands for Area Under the Curve: a statistical measure of predictive accuracy. An AUC of 0.50 means a model performs no better than flipping a coin, essentially chance. An AUC of 1.0 means perfect prediction. The respondents in this study scored 0.70 across all promotion comparisons. When the researchers narrowed it to the sharpest contrast, unpromoted officers versus those who made lieutenant, the AUC was 0.67. For sergeant-versus-lieutenant comparisons, it was 0.70 again.

In plain terms: participants correctly identified the more-promoted officer about 70% of the time, based solely on a photograph taken when that officer was a recruit.

Put another way: the more leader-like a respondent thought someone looked, the more likely they were to correctly pick that person as the higher-ranking officer. Across all promotional comparisons, each notch up on the leadership rating scale made a correct guess 27% more likely. That number climbed to 34% when comparing sergeants against lieutenants, the sharpest test of whether facial impressions track not just promotion, but level of promotion.

The study also tested what actually drives those leadership perceptions. Researchers built four separate statistical models to explain why respondents rated some faces as more leadership-ready than others:

• Model 1: Perceived facial traits (attractiveness, trustworthiness, dominance, etc.)
• Model 2: Other observable traits from the photo (glasses, hair, facial hair, perceived gender, perceived race)
• Model 3: Respondent demographics (age, gender, etc.)
• Model 4: Respondent personality traits (openness, conscientiousness, extraversion, agreeableness, anxiety)

Model 1 – facial traits outperformed all the others. By a substantial margin. In other words, it wasn’t the respondent’s personality that drove the predictions, and it wasn’t surface-level observable features like whether someone wore glasses. It was the facial traits themselves. So which specific traits mattered most? Trustworthiness had the strongest positive effect on perceived leadership. Distinctiveness actually had a negative effect. Dominance, interestingly, contributed almost nothing.

This replicates findings from decades of prior research. A 1984 study found similar results among military academy cadets. Swiss researchers showed undergraduates pairs of French parliamentary candidates and asked who looked more competent — they picked the actual winner 72% of the time. Children given the same task got it right 71% of the time. The pattern, across cultures and contexts, is remarkably consistent: humans use faces to make fast, automatic leadership judgments and those judgments predict real outcomes.

Why This Should Concern Every Agency Leader

Consider two officers at your agency: Officer Chen and Officer Ramirez. Chen has the kind of face that people consistently rate as trustworthy, secure, and leadership-ready. Ramirez doesn’t. Even if Ramirez is the stronger candidate on every measurable dimension such as exam scores, performance reviews, time-in-service, research suggests Chen may still have a meaningful edge at the promotion board. Not because of anything he’s done. Because of his face.

That’s the mechanism at work here. The researchers describe it as ‘System 1’ thinking: fast, automatic, largely unconscious judgment. The brain sizes up a face in milliseconds. Before anyone in an interview panel can articulate a coherent reason, a gut-level impression has already formed. Every subsequent interaction then gets filtered through that lens.

The stakes extend well beyond any individual career. Leaders shape use-of-force culture, resource allocation, and community engagement strategy and when officers sense those leaders were chosen unfairly, the consequences are measurable. Voluntary resignations in law enforcement rose 280% in 2020.

Research on organizational justice shows that officers who perceive promotions as inequitable report lower job satisfaction, generate more complaints, and disengage from discretionary effort.

One nuance worth noting: the researchers found that respondents rated women and non-white officers as having higher leadership potential, and those perceptions accurately predicted promotional success. This doesn’t mean bias in policing is solved. Women, particularly non-white women, still face documented delays in promotion. But it does suggest that facial characteristics may not be the primary mechanism driving those particular inequities.

The study also can’t tell us exactly why facial traits predict promotion. It’s possible that appearance is correlated with behaviors that genuinely contribute to professional success. An officer perceived as trustworthy may actually be more effective at building victim rapport or gaining suspect cooperation, outcomes that earn recognition and advance careers. The face may be a signal, not the cause.

What Agencies Should Do: Guarding Against the Face That Fits

The researchers aren’t calling for agencies to scrap their promotion processes; they’re calling for vigilance within them. Their core recommendation is straightforward: persist in using multiple forms of testing and stay open to new evaluation methods that can better identify the best candidates.

The findings also suggest value in simply understanding this bias exists. Awareness matters. When promotion panels know that facial impressions form fast and stick hard, they’re better positioned to interrogate their own judgments and lean more heavily on the objective data in front of them.

Two places to start. First, look at how much unstructured face-to-face evaluation your process relies on relative to scored, documented performance. Interviews and oral boards aren’t going away, nor should they, but the less your final decision rests on panel impressions alone, the less vulnerable it is to the bias this research documents. Second, consider structured panel training. Not a general diversity seminar, but specific instruction on how fast impressions form, how persistent they are, and how to actively counter them by returning to scored criteria.

Resources

• IACP Testing & Assessment Centers — The IACP designs and administers custom promotional examinations and assessment centers for agencies of all sizes, from first-line supervisor through chief. A useful option if your agency wants an externally run, structured, and legally defensible process. International Association of Chiefs of Police

• Police Chief Magazine — “A Promotional Process for the Smaller Police Agency” — A practical guide specifically for smaller agencies navigating written exams, oral boards, and assessment center testing. A Promotional Process For The Smaller Police Agency – Police Chief Magazine

Final Thoughts

There’s an uncomfortable irony buried in this research. Agencies spend enormous resources (time, money, staff hours, consultant fees) building promotion systems designed to be fair and rigorous. And then the human brain, doing what it evolved to do, quietly undermines a portion of that work by sizing up a face in the first fraction of a second of an interview.

That doesn’t mean the systems are worthless. It means they require active, ongoing vigilance. It means promotion panels need training. It means data needs to be collected and reviewed. It means the question ‘does this person look like a leader?’ needs to be replaced, relentlessly, with ‘does this person’s documented performance demonstrate the capacity to lead?’

The stakes are high enough to get this right. Leaders shape agencies. They shape the culture officers work in, the way communities are policed, and the organizational justice that determines whether your best people stay or walk. The officers you promote will make thousands of decisions that ripple outward for years.

Don’t let a face make that call for you.

Enjoyed this post?

Get more research translated into plain language delivered straight to your inbox.

Subscribe below and let me know which topics you want to hear more about. It only takes a minute.

If you found this helpful, share it with your team or on your socials. It helps more people make sense of research that matters.

Carlee Ruiz

Carlee Ruiz is the founder of PoliceResearchHub, a platform that turns academic research into practical insights for law enforcement. She brings seven years of experience as a crime analyst and justice researcher, with a focus on evidence based policing, gun violence prevention, and public safety strategy. Carlee currently works as a researcher at National Policing Institute, where she supports evaluations of law enforcement programs and criminal justice initiatives across the country. Through PoliceResearchHub, she aims to make research easier to understand and more useful for the people doing the work on the ground.