magazine issue 9 2008 / Issue 8

The Surprising Effect of Facial Appearance on Political Decision-Making

• written by Theresa DiDonato
• edited by Yudit Namer

Bailenson, Iyengar, Yee, and Collins (in press) used a creative method to test this idea. Building on their previous work (Bailenson, Garland, Iyengar, & Yee, 2006), they examined the influence of facial similarity on voting behavior by actuallymanipulating the degree of facial similarity between participants and candidates. How? By digitally morphing images of participants’ faces with photographs of current candidates! Essentially, they screened out participants with glasses and facial hair, and used only high-quality photographs. They then used a computer program to morph, or blend, participants’ faces into the faces of real-life United States’ politicians, such as Hillary Clinton. Participants were unaware of the image modifications.

The researchers conducted three experiments in which they showed participants candidate photos that had been morphed with themselves (self-morph) and/or with a random other participant (other-morph). In some cases, these images were of widely-known politicians, like John Edwards or Rudy Giuliani, while other times they were of unfamiliar candidates. Participants rated each candidate on a set of ten positive personality traits (i.e. moral, intelligent, and friendly), reported their party affiliation (Democrat or Republican), and indicated the strength of that affiliation. In the final experiment, participants also saw a brief description of the candidates’ positions on issues like the Iraq War along with their picture.

The findings revealed a fascinating effect of facial similarity on voting decisions. With only a week until election day, people actually preferred candidates whose faces were manipulated to appear like their own. A more nuanced analysis of this effect showed that only judgments about unfamiliar candidates, not well-known candidates, were affected by the similarity manipulation. Overall, participants generally preferred familiar faces, but among unfamiliar candidates, the one who looked most like the participant got his or her vote. Applying this research, we can imagine, for example, that were the French president Nicolas Sarkozy to run for re-election in 2012 against a relatively unknown candidate, people would use the facial similarity cue in their impression of the opponent, but not in their judgment of the well-known Sarkozy. While typically only recognizable political figures compete for national offices, many lesser-known people fight for and win powerful government jobs, such as president of regional council (France), minister-president (Germany), or governor (United States). The implications of this research are particularly charged for local or state-wide races like these, in which voters may be unacquainted with candidates’ policy positions, and may turn to visual information as a cue to guide their decision.

Bailenson and colleagues’ (in press) second experiment provided some evidence that independents and people who only weakly-identify with their political parties were susceptible to the influence of visual cues even in judging famous politicians. In other words, while die-hard Republicans favored Guiliani regardless of the similarity morph, less-partisan people relied on facial similarity to decide their preference. Interestingly, strongly-identified Democrats or Republicans used their party as a cue for voting only when candidates were described as sharing their perspectives on issues, like the Iraq War or the outsourcing of American jobs. This places policy information at the forefront of a voter’s decision. At the same time, partisan voting was limited to well-known candidates; decisions made about unfamiliar candidates still depended, at least in part, upon facial similarity.