Skin Tone as a Visual Cue in Political Campaigns: Evidence from Elections in Mexico

Research has long established that physical appearance and phenotypical traits influence how people evaluate political candidates. Skin color is one of the visual cues the electorate perceives during campaigns and media coverage, and a possible source of prejudice and bias. Here we use computational visual analysis to investigate the association between skin tone and electoral success during the 2024 Mexican Federal Elections. Our analyses incorporate important correlates not considered in prior research, most prominently news coverage, political ads, and social media use. We also account for geographic differences in average skin tone, explicitly modeling a source of variation that is obscured when skin tone is analyzed disregarding contextual baselines. We use fixed effects generalized linear models to control for state and party variation as well as candidate-level demographics, political experience, incumbency status, and media presence (news coverage, political ads, and social media use). Consistent with past work, we find that as candidates’ skin tone becomes darker, their predicted vote share declines. However, news coverage, political ads, and social media use are, together, more important predictors of electoral success than skin tone, indicating that this phenotypical trait is not as prominent a visual cue as earlier research suggested.