Whose Face will be First?


N E W S  R E L E A S E


For Immediate Release

October 29, 2008



Caltech-Led Researchers Find Negative Cues from Appearance Alone Matter for Real Elections



PASADENA, Calif.– Brain-imaging studies reveal that voting decisions

are more associated with the brain’s response to negative aspects of

a politician’s appearance than to positive ones, says a team of

researchers from the California Institute of Technology (Caltech),

Scripps College, Princeton University, and the University of Iowa.

This appears to be particularly true when voters have little or no

information about a politician aside from their physical appearance.


The research was published online in the journal Social Cognitive and

Affective Neuroscience (http://scan.oxfordjournals.org) on October 28.


Deciding whom to trust, whom to fear, and indeed for whom to vote in

an election depends, in part, on quick, implicit judgments about

people’s faces. Although this general finding has been scientifically

documented, the detailed mechanisms have remained obscure. To probe

how a politician’s appearance might influence voting decisions,

Michael Spezio, an assistant professor of psychology at Scripps

College and visiting associate at Caltech, and Antonio Rangel, an

associate professor of economics at Caltech, examined brain

activation in subjects looking at the faces of real politicians.


Using a functional magnetic resonance imaging (fMRI) scanner at the

Caltech Brain Imaging Center, the researchers obtained

high-resolution images of brain activation as volunteers made

decisions about politicians based solely on their pictures.


The researchers conducted two independent studies using different

groups of volunteers viewing the images of different politicians.

Volunteers were shown pairs of photos, each with a politician coupled

with their opponent in a real election in 2002, 2004, or 2006.

Importantly, none of the study subjects were familiar with the

politicians whose images they viewed.


In some experiments, the volunteers had to make character-trait

judgments about the politicians–for example, which of the two

politicians in the pair looked more competent to hold congressional

office, or which looked more likely to physically threaten the

volunteer. In other experiments, volunteers were asked to cast their

vote for one politician in the pair; once again, their decisions were

based only on the politicians’ appearances.


The results correlated with actual election outcomes. For example,

politicians who were thought to look the most physically threatening

in the experiment were more likely to have actually lost their

elections in real life. The correlation held true even when

volunteers saw the politicians’ pictures for less than one tenth of a



Importantly, the pictures of politicians who lost elections, both in

the lab and in the real world, were associated with greater

activation in key brain areas known to be important for processing

emotion. This was true when volunteers simply voted and also when

they closely examined the politicians’ pictures for character traits.

The studies suggest that negative evaluations based only on a

politician’s appearance have some effect on real election

outcomes–and, specifically, may influence which candidate will lose

an election. This influence appears to be more uniform  than the

influence exerted by positive evaluations based on appearance.


This finding fits with prior studies in cognitive neuroscience as

well as in political theory.


“The results from our two studies suggest that intangibles like a

candidate’s appearance may work preferentially, or more uniformly,

via negative motives, and by means of brain processing contributing

to such negative evaluations,” says Michael Spezio, the lead author

on the study.


“It’s important to note that the brain region most closely associated

with seeing pictures of election losers, known as the insula, is

known to be important in processing both negative and positive

emotional evaluations. Its increased activation in response to the

appearance of election losers is consistent with its association with

negative emotional evaluations in several domains, including the

sight of someone who looks disgusted or untrustworthy,” Spezio says.


“Candidates try to evoke emotional reactions when they campaign for

office, and this research gives us a new perspective on how much

emotions might matter, and how they might matter, in terms of how

voters view candidates,” says study coauthor R. Michael Alvarez, a

professor of political science at Caltech and codirector of the

Caltech/MIT Voting Technology Project.


One surprise in the study is that negative evaluations, such as the

perception that a candidate is threatening, influence election loss

significantly more than positive evaluations like attractiveness

influence election success.


“While these findings are certainly very provocative, it is important

to note their limitations,” says study senior author Ralph Adolphs,

Bren Professor of Psychology and Neuroscience and professor of

biology at Caltech, and director of the Caltech Brain Imaging Center.


In particular, Adolphs says, the observed effects, while

statistically significant, were rather small. “There is no doubt that

many, many sources of information come into play when we make

important and complex decisions, such as will happen in the upcoming

elections. We are not claiming that how the candidates look is all

there is to the story of how voters make up their minds–or that this

is even the biggest part of the story. However, we do think it has

some effect–and, moreover, that this effect may be largest when

voters know little else about a candidate.”


Adds Spezio, “Given the size of the effects we see, we are likely

detecting the influence of voters who have little or no information

about a candidate’s views or life story, for example, or who choose

not to pay attention to that information. Our finding is consistent

with literature showing that humans prioritize negative information

about outgroups”–groups of individuals who are perceived to not

belong to one’s own group, as defined by characteristics such as

profession, age, gender, social community, and shared values, but to

an outside group. “A voter who knows nothing about a candidate will

likely put that candidate into a default outgroup position. From

there, negative attributions are expected to get the primary weight

in decisionmaking. And that is precisely what we see,” he says.


“Earlier behavioral studies showed that rapid, effortless inferences

from facial appearance predict the outcomes of political elections,”

says study coauthor Alex Todorov, an assistant professor of

psychology and public affairs at Princeton University. In 2005,

Todorov published the first study to show that voter decisions are

significantly associated with character-trait judgments that are

based entirely on the visual appearance of political candidates.


“However,” Todorov adds, “these studies did not show how these

inferential processes could play out at the level of individual

voters. Two types of evidence will be critical to delineate the

causal effects of appearance on electoral success: work by political

scientists studying real voting decisions and work by cognitive

neuroscientists studying the proximal mechanisms of the effects of

inferences on decisions. The fMRI studies are an important step in

the latter direction.”


The coauthors of the study, titled, “A neural basis for the effect of

candidate appearance on election outcomes,” are John O’Doherty,

associate professor of psychology at Caltech, Kyle Mattes of the

University of Iowa, and Hackjin Kim of Korea University.


The work was supported by the Gordon and Betty Moore Foundation, the

National Science Foundation, and the National Institutes of Health.