NEW YORK, NY — At a moment when diversity efforts are in retreat in organizations across corporate America, a new study from researchers at Columbia Business School reveals that Americans unconsciously associate leadership with looking white and male—even when shown faces that are ambiguous in gender and race. The findings offer a new explanation for why women and minorities remain underrepresented in leadership roles.
Published in PLOS ONE, the study, “The Look of a Leader” co-authored by Columbia Business School Professors Modupe Akinola and Michael Slepian, and Columbia Business School Ph.D. graduate Katherine Qianwen Sun used a psychological technique known as reverse correlation to uncover how people visually imagine leaders. Nearly 4,000 people took part in the research, which encompassed 15 studies and included five core experiments that tested visual biases and their intersection with demographic cues. Researchers found that when participants were asked to select which face looked more like a leader, they consistently chose faces that appeared more white and more male—even when the base image was intentionally race- and gender-neutral.
“When people imagine a leader, their mental image tends to skew white and male, regardless of their own background or the context,” said Professor Akinola, the Barbara and David Zalaznick Professor of Business at Columbia Business School. “This has implications for how hiring, promotion, and leadership potential are perceived in organizations. Companies need to be aware that bias can be visual, not just verbal or behavioral.”
To measure visual leadership biases and how they intersect with race and gender, Akinola, Slepian and Sun conducted five unique studies. In Study 1, participants were shown a series of race- and gender-ambiguous composite faces, each overlaid with random visual noise. They were asked to select which of two versions looked more like a “leader” or a “follower.” By averaging participants’ selections, the researchers generated composite images that represented the mental template of a leader versus a follower. The reverse correlation method was repeated in Study 2 using a clearly male—but race-ambiguous—base image, and in Study 3 using a clearly female—but race-ambiguous—base image. In Studies 4 and 5, participants were shown faces that were gender-ambiguous but clearly white for Study 4 or clearly Black for Study 5. In each case, participants judged which faces appeared more “leader-like,” allowing researchers to isolate how race and gender shape perceptions of leadership.
Across all five experiments, participants consistently associated leadership with faces that appeared more white and more male—regardless of the base image. These biases were strongest when no demographic cues were explicitly provided, revealing how deeply ingrained these visual stereotypes are.
Additional Findings:
- Leadership Standards Shift Based on Identity: Results suggested that female or Black leaders may need to appear both dominant and positive to fit a leadership mold, while white male leaders are afforded more flexibility in their appearance.
- Dominance as Leadership: Both white and Black leaders were envisioned to be more masculine and dominant.
- Male Participants Amplified Bias: Men were more likely than women to imagine leaders as happier, more trustworthy, and warmer than followers, indicating differing expectations by participant gender.
“These findings help explain why some individuals face higher barriers to leadership,” said Katherine Qianwen Sun, Columbia Business School Ph.D. graduate and UCLA-Anderson Post-Doctoral Scholar. “Our work adds empirical evidence to the idea that social category cues like race and gender still actively shape how people envision and evaluate leaders.”
