Predicting political elections from rapid
and unreflective face judgments
Charles C. Ballew II* and Alexander Todorov*†‡
*Department of Psychology and†Woodrow Wilson School of Public and International Affairs, Princeton University, Princeton, NJ 08540
Edited by Edward E. Smith, Columbia University, New York, NY, and approved September 25, 2007 (received for review June 10, 2007)
the facial appearance of candidates predicted the outcomes of
gubernatorial elections, the most important elections in the United
States next to the presidential elections. In all experiments, par-
ticipants were presented with the faces of the winner and the
runner-up and asked to decide who is more competent. To ensure
that competence judgments were based solely on facial appear-
ance and not on prior person knowledge, judgments for races in
which the participant recognized any of the faces were excluded
from all analyses. Predictions were as accurate after a 100-ms
exposure to the faces of the winner and the runner-up as exposure
after 250 ms and unlimited time exposure (Experiment 1). Asking
participants to deliberate and make a good judgment dramatically
increased the response times and reduced the predictive accuracy
of judgments relative to both judgments made after 250 ms of
exposure to the faces and judgments made within a response
deadline of 2 s (Experiment 2). Finally, competence judgments
collected before the elections in 2006 predicted 68.6% of the
gubernatorial races and 72.4% of the Senate races (Experiment 3).
These effects were independent of the incumbency status of the
candidates. The findings suggest that rapid, unreflective judg-
ments of competence from faces can affect voting decisions.
face perception ? social judgments ? voting decisions
United States. American states are significant political and eco-
nomic entities, with some being larger and economically more
powerful than many foreign countries. In addition to wielding
to ascend to the presidency. For example, 17 of 43 presidents have
and potential of a governorship comes at a cost. In 1998, the 36
gubernatorial races averaged $14.1 million in expenses each (1). By
comparison, the average Senate campaign cost was $3.3 million in
Despite the significance of gubernatorial races, we show that
rapid, unreflective judgments of competence based solely on facial
appearance and made after as little as 100 ms of exposure to the
our prior work on forecasting the outcomes of Senate elections (3),
we showed that people believe that competence is the most
important attribute for a politician and that trait inferences of
competence from faces but not other traits (e.g., trustworthiness,
attractiveness, likeability, etc.) predict election outcomes. We ar-
gued that these inferences are rapid, intuitive, and unreflective, but
we did not provide direct evidence for this assumption.
The first objective of the current research was to provide such
evidence. The second objective was to replicate the Senate findings
for gubernatorial races, which are arguably more important. The
third objective was to test whether the effect of competence
incumbency status of the candidates. In the Senate and House of
Representatives elections, there are no terms limits, and incum-
bents have overwhelming odds of being reelected (4). In contrast,
ith the exception of the president, state governors are
arguably among the most powerful elected officials in the
many states have term limits for governors, and, correspondingly,
there are fewer incumbents in gubernatorial races.
Faces are a rich source of social information, and trait judgments
from faces can be made after minimal time exposure (5). For
example, we have shown that 100 ms of exposure to a face is
sufficient for people to make a variety of trait judgments, including
competence, and that additional time only increases confidence in
judgments (6). In Experiment 1, we tested whether competence
judgments made after 100 ms of exposure to the faces of the
candidates predict the outcomes of gubernatorial elections better
than chance and whether additional time exposure (250 ms and
unlimited time) improves the accuracy of prediction.
In our previous work (3) and Experiment 1, participants were
asked to rely on their ‘‘gut’’ feeling or first impression when making
on judgments. Deliberating about judgments that are unreflective
and not easy to articulate can interfere with the quality (7) and
consistency (8) of the judgments. If trait judgments from faces are
unreflective, instructions to deliberate and make a good judgment
should not improve the predictive accuracy of judgments. In
Experiment 2, we tested whether deliberation judgments are less
accurate in predicting the election outcomes than judgments made
after 250 ms of exposure to the faces and judgments made under
response deadline of 2 s, forcing participants to rely on quick
In Experiment 3, we collected competence judgments for both
gubernatorial and Senate races in 2006 before the actual election.
We tested whether these judgments based solely on facial appear-
ance would predict the election outcomes better than chance, as we
did in our previous work on predicting the Senate elections
prospectively in 2004 (3).
Participants were presented with the faces of the winner and the
runner-up for 89 gubernatorial races and asked to judge which
1). In the third condition the faces were presented until the
participant responded, with no time constraints. For each race,
participants made three consecutive judgments: a binary choice of
who was more competent, a continuous judgment (on a nine-point
scale) of how much more competent the chosen person was§, and
a recognition judgment. If participants recognized either of the
faces for a given race, their responses for that race were not
Author contributions: C.C.B. and A.T. designed research; C.C.B. performed research; A.T.
analyzed data; and A.T. wrote the paper.
The authors declare no conflict of interest.
This article is a PNAS Direct Submission.
‡To whom correspondence should be addressed. E-mail: firstname.lastname@example.org.
§This measure did not contribute any additional information over the information gained
from the binary competence judgments. Details are provided in SI Text.
This article contains supporting information online at www.pnas.org/cgi/content/full/
© 2007 by The National Academy of Sciences of the USA
November 13, 2007 ?
vol. 104 ?
no. 46 www.pnas.org?cgi?doi?10.1073?pnas.0705435104
analyzed. Thus, all judgments of competence were based only on
facial appearance and not on other knowledge.
Results. Analysis at the level of participants. Participants in all three
conditions were more likely to choose the winner than the run-
of 0.50. The judgments in the different conditions did not differ
(Fig. 2A) (F ? 1). Across conditions, the mean judgment was 0.57
[SE ? 0.006; t(119) ? 11.22, P ? 0.001, d ? 2.05]. Although the
unlimited time condition were twice as long as the response times
in the 100-ms and 250-ms conditions (Fig. 3A). The response times
in the latter two conditions did not differ from each other and were
significantly shorter than the response times in the unlimited time
condition [t(117) ? 13.16, P ? 0.001, and t(44.84) ? 10.22, P ?
0.001 (not assuming equal variance), respectively].
Analysis at the level of the races. Aggregating across participants, the
percentage of correctly predicted races (i.e., races for which ?50%
of participants judged the winner as more competent) was higher
in the 250-ms condition than in the other two conditions (Table 1),
although this difference was not significant. Aggregating across the
three experimental conditions, the binary competence judgments
predicted 64.0% of the outcomes of the gubernatorial races, which
was significantly higher than chance [?2(1) ? 7.02, P ? 0.008].
We also tested whether the difference in competence between
the two candidates was linearly related to the difference in votes
between them. As shown in Table 1, in all conditions the average
competence of the candidate correlated positively and significantly
with the proportion of votes won by this candidate. The more
competent the candidate was perceived to be relative to the other
candidate, the higher the proportion of votes for this candidate.
Averaging across the three experimental conditions, the mean
competence judgments for the candidates correlated 0.27 (P ?
0.011), with the proportion of votes [supporting information (SI)
Fig. 5]. Thus, snap judgments of competence from facial appear-
ance accounted for 7.2% of the variance of vote share.
Discussion. These findings suggest that simple, fast, binary judg-
ments of competence aggregated across a relatively small sample
elections. Judgments made after as little as 100 ms of exposure to
the faces of the candidates predicted the election outcomes better
than chance. Additional time exposure to the faces did not improve
these predictions, although the response times for the judgments
substantially increased when the time exposure was unconstrained.
To our knowledge, this study is the first demonstration that
judgments made after minimal time exposure to the faces of the
candidates predict election outcomes. In our previous work (3), the
minimum time exposure to the faces was 1 s. Clearly, much less
exposure is needed for these judgments. The current findings are
consistent with the ideas that trait judgments from faces can be
characterized as rapid, unreflective, intuitive (‘‘system 1’’) judg-
ments (e.g., refs. 9 and 10) and that, because of these properties,
their influence on voting decisions may not be easily recognized by
a good judgment (rather than relying on a gut feeling or first
impression) affect competence judgments. Participants were ran-
domly assigned to one of three conditions: a deliberation condition
in which they were asked to think carefully about their choice and
had to decide within 2 s, and a 250-ms replication condition.
We included a response deadline condition in which the faces
were presented until the participants responded, but they had to
respond within 2 s. As shown in Experiment 1 (Fig. 3A), this time
was longer than the average response time for the 100- and 250-ms
conditions but substantially shorter than the average response time
for the unlimited time condition. Thus, the response deadline
procedure should force participants to rely on quick judgments. If,
as we argue, trait judgments from faces are rapid and unreflective,
participants’ judgments in this condition should predict the out-
comes of the elections better than chance. However, the judgments
in the deliberation condition should be less predictive of the
election outcomes than the judgments in the 250-ms and response
are of the same gender and ethnicity are more interesting because
differences in judgments of competence cannot be attributed to
differences in gender and ethnicity. Moreover, the salience of the
latter factors can activate gender and race stereotypes and, accord-
ingly, change participants’ responses. In fact, when the analysis was
limited to the 55 gubernatorial races in which the winner and the
runner-up were of the same gender and ethnicity, the predictions
improved, just as they did in our previous work on Senate elections
(3). Averaging across the three conditions, the percentage of
correctly predicted races was 69.1% [?2(1) ? 8.02, P ? 0.005], and
the linear correlation between the perceived competence of the
candidates and the vote share was 0.32 (P ? 0.017). Thus, in
Experiment 2, participants made judgments only for the 55 races in
which the candidates were of the same sex and ethnicity.
Results. Analysis at the level of participants. As in Experiment 1,
participants in all three conditions were more likely to choose the
winner than the runner-up as more competent (P ? 0.001).
However, the effect was smaller in the deliberation condition than
in the 250-ms and response deadline conditions (Fig. 2B) [F(2,
107) ? 3.51, P ? 0.033 for the omnibus test]. Follow-up contrast
condition. Participants decided who was more competent.
An example of an experimental trial in the 250-ms presentation
100 ms250 msUnlimited
Proportion of correctly
250 ms Response
winner was judged as more competent than the runner-up. (A) As a function
of time exposure to faces in Experiment 1. (B) As a function of experimental
and deliberation. Error bars show the SEM.
Proportion of correctly predicted gubernatorial races in which the
Ballew and TodorovPNAS ?
November 13, 2007 ?
vol. 104 ?
no. 46 ?
tests showed that, although the judgments in the latter two condi-
tions were not significantly different (t ? 1), they were significantly
better than the judgments in the deliberation condition [t(107) ?
2.65, P ? 0.009, d ? 0.51].
The response times in the deliberation condition were substan-
tially longer than the response times in the 250-ms and response
t(39.69) ? 7.90, P ? 0.001 (not assuming equal variance), respec-
tively]. The response times in the latter two conditions were not
significantly different (t ? 1).
Analysis at the level of the races. Aggregating across participants, the
judgments in the 250-ms and response deadline conditions pre-
dicted a higher percentage of races than the judgments in the
deliberation condition (Table 1), although these differences were
not significant. The percentage of correctly predicted races in the
deliberation condition was not significantly different from chance.
Aggregating across the 250-ms and the response deadline condi-
tions, the binary competence judgments predicted 70.9% of the
gubernatorial races, which was significantly higher than chance
better than the predictions in each of the conditions, 250 ms and
response deadline (see Table 1), demonstrating that aggregating
across more judges improves the prediction (see the supporting
online material for ref. 3 for bootstrapping simulations).
As shown in Table 1, in all conditions the average competence of
by the candidate, although the only correlation that reached sig-
nificance was in the 250-ms condition. Aggregating across the
250-ms and response deadline conditions, the correlation between
competence judgments and vote share was 0.32 (P ? 0.018). Thus,
rapid, unreflective judgments of competence from facial appear-
ance accounted for 10.2% of the variance of vote share.
Deliberation judgments and unreflective judgments—judgments
aggregated across the 250 ms and response deadline conditions—
were highly correlated (r ? 0.78, P ? 0.001). This shared variance
is consistent with the possibility that deliberation judgments were
anchored on rapid, immediate impressions from the faces. If this is
the case, removing the shared variance between deliberation and
unreflective judgments should not affect the correlation with the
vote share for unreflective judgments, but it should affect this
correlation for deliberation judgments. Partial correlation analysis
confirmed this hypothesis. The partial correlation between unre-
flective judgments and vote share controlling for deliberation
judgments was 0.34 (P ? 0.011) (Fig. 4A). In contrast, the partial
correlation between deliberation judgments and vote share con-
trolling for unreflective judgments was ?0.19 (P ? 0.18) (Fig. 4B).
Analysis across both experiments. Althoughtheresponsetimesforthe
judgments in the unlimited time (Experiment 1) and deliberation
(Experiment 2) conditions were almost identical (Fig. 3) (see also
affected in the latter condition (Fig. 2). This finding suggests that
intuitive judgments are affected by the deliberative mind set rather
necessarily lead to changes in judgments, although it may lead to
increased confidence in judgments (6). Although judgments in the
(r ? 0.82, P ? 0.001), perhaps reflecting controlled processing, this
shared variance should not predict the outcome of the races to the
extent that these predictions are based on rapid, intuitive judg-
ments. Because Experiment 2 included a subset of the races used
Table 1. Percentage of correctly predicted gubernatorial races and correlations between
perceived competence of candidates and their vote share as a function of experimental
conditions in Experiments 1–3
100-ms exposure to faces
250-ms exposure to faces
Response deadline (2 s)
The percentages indicate the races in which the candidate who was perceived as more competent by the
†, P ? 0.05.
100 ms250 msUnlimited
Response times (ms)
in Experiment 2: 250-ms exposure to faces, response deadline of 2 s, and
deliberation. Error bars show the SEM.
Response times for competence judgments. (A) As a function of time
-0.3 -0.2 -0.10
Two-party vote share for candidate
-0.3 -0.2-0.100.1 0.20.3
Competence of candidate (non-shared
variance with deliberation judgments)
Competence of candidate (non-shared
variance with unreflective judgments)
nonshared variance of unreflective judgments of competence of the candi-
dates (the x axis plots the regression residuals of unreflective judgments
regressed on deliberation judgments) (A) and nonshared variance of deliber-
ation judgments of competence of the candidates (the x axis plots the regres-
(B). Each point represents a gubernatorial race. The line represents the best
fitting linear curve.
Scatter plots of the two-party vote share for the candidates and
www.pnas.org?cgi?doi?10.1073?pnas.0705435104Ballew and Todorov
in Experiment 1 and the analysis was conducted at the level of the
races, we could test this hypothesis. Controlling for the shared
variance did not affect the correlation between vote share and
judgments in the unlimited time condition. The partial correlation
(0.28, P ? 0.05) was practically the same as the zero order
correlation (0.27, P ? 0.05). Thus, what predicted the outcomes of
the races in the unlimited time condition in Experiment 1 was the
nonshared variance with deliberation judgments. Finally, the pre-
dictive accuracy of the judgments in the unlimited time condition
was eliminated when the analysis controlled for unreflective judg-
ments. The correlation between vote share and the corrected time
unconstrained judgments was reduced from 0.28 to ?0.02 (see also
SI Text and SI Table 3).
Discussion. As in Experiment 1, judgments made after 250 ms of
exposure to the faces of the candidates predicted the outcomes of
that were made within a response deadline of 2 s, forcing partici-
pants to rely on rapid, unreflective judgments. The judgments of
participants who were asked to deliberate and make a good
judgment were less accurate in predicting the election outcomes
and substantially slower than the judgments of participants in the
other two conditions.
Deliberation judgments shared a substantial amount of variance
with unreflective judgments. Removing this variance did not affect
the relation between vote share and unreflective judgments, but it
did affect the relation between vote share and deliberation judg-
ments. Specifically, whereas the corrected unreflective judgments
predicted vote share, the corrected deliberation judgments did not
predict vote share. If anything, the correlation between corrected
deliberation judgments and vote share was negative. These findings
are consistent with the hypothesis that deliberation judgments are
anchored on rapid, automatic trait impressions from faces and that
any positive relation between deliberation judgments and vote
share can be accounted for by the shared variance of deliberation
judgments with these automatic impressions. That is, what predicts
the outcomes of elections is the automatic component of trait
judgments. Deliberation instructions add noise to automatic trait
judgments and, consequently, reduce the accuracy of prediction.
This hypothesis is also consistent with the analyses across Experi-
ments 1 and 2. What predicted the outcomes of the races in the
unlimited time condition in Experiment 1 was not the variance that
was shared with deliberation judgments, but the variance that was
shared with rapid, intuitive judgments.
In this experiment, we collected competence judgments 2 weeks
judgments can predict elections prospectively. Participants were
presented with the pictures of the Democratic and Republican
candidates for each gubernatorial race and asked to choose the
the 2006 Senate races in this experiment.
Participants were more likely to choose the winner than the
SE ? 0.01; t(63) ? 3.94, P ? 0.001, d ? 0.99] races. Aggregating
across participants, the judgments predicted 68.6% of the guber-
natorial races [?2(1) ? 4.83, P ? 0.028] against the chance predic-
tion of 50%, and 72.4% of the Senate races [?2(1) ? 5.83,
P ? 0.016].
The correlation between the perceived competence of the can-
didates and their vote share was 0.47 (P ? 0.011) for the Senate
races and 0.29 (P ? 0.09) for the gubernatorial races. Although the
latter correlation was not significant, it was comparable in size to
the correlations obtained in the other experiments (see Table 1).
The small number of races in this experiment makes the rejection
of the null hypothesis more difficult. Because in this experiment we
used the same procedures as those used in the unlimited time
condition in Experiment 1, we combined the mean judgments for
from Experiment 1. For these 124 races, the correlation between
the perceived competence of candidates and their vote share was
0.27 and highly significant (P ? 0.003).
Replicating our prior findings of prospectively predicting the
based solely on the facial appearance of the candidates and col-
both gubernatorial and Senate elections.
Incumbency Status and Competence Judgments
Gubernatorial races are not only more important than House and
Senate races, but also more interesting with respect to addressing
the effects of incumbency and perceived competence of candidates
on predictions of the election outcomes. It is a well known fact that
incumbents have a distinctive advantage in American politics (4,
11). For example, in the House races studied by Todorov et al. (3),
incumbents won in 89% of the races. In the Senate races, incum-
bents won in 74% of the races. If incumbents appear more
competent than challengers and participants are choosing the
incumbent more frequently than the challenger, the predictive
effect of competence judgments may be explained as an artifact of
incumbency status. That is, according to this account, competence
judgments will predict the winner only in races in which the
the test of this hypothesis because many states have term limits for
governors and, correspondingly, there are fewer incumbents in
this hypothesis. Competence judgments were independent of in-
cumbency status in predicting the outcome of the elections.
Because the races used in Experiment 1 included the races used
in Experiment 2, we report the analysis only for Experiment 1. For
simplicity of presentation, for Experiment 1, we used the compe-
tence judgments aggregated across the three experimental condi-
tions. In all of these conditions, participants were instructed to rely
on their gut feeling when making the judgments, and the results
were identical when the analysis was performed separately for each
condition. The test for dependence of judgments and incumbency
status was not significant [?2(2) ? 1.95, P ? 0.38] (see Table 2 for
the relevant proportions). Collapsing across the races in which the
was perceived as more competent won in 62.7% of the races. The
was 65.9% [?2(1) ? 1, P ? 0.77]. For Experiment 3, as in the case
of Experiment 1, the test for dependence of judgments and incum-
bency status was not significant [?2(2) ? 1, P ? 0.65] (Table 2).
Combining the races from both experiments (n ? 124 races) to
increase statistical power did not change the results. The candidate
who was perceived as more competent won in 67.7% of the races
in which the incumbent won and in 62.9% of the races in which the
incumbent lost or there was no incumbent [?2(1) ? 1, P ? 0.57].
Thus, incumbency status and perceived competence were indepen-
dent predictors of the election outcomes.
Table 2. Percentage of correctly predicted gubernatorial races
by competence judgments as a function of incumbency status
ExperimentIncumbent wonIncumbent lostNo incumbent
65.8% (n ? 38)
70.8% (n ? 24)
85.7% (n ? 7)
100% (n ? 1)
59.1% (n ? 44)
60.0% (n ? 10)
For Experiment 1, the competence judgments were aggregated across the
three experimental conditions.
Ballew and TodorovPNAS ?
November 13, 2007 ?
vol. 104 ?
no. 46 ?
Extending our prior work on forecasting the outcomes of Senate
elections (3), we have shown that rapid, unreflective judgments of
competence based solely on facial appearance predicted the out-
comes of gubernatorial elections. Even after 100 ms of exposure to
the faces of the winner and the runner-up, participants were more
likely to choose the winner as more competent. In addition to
showing that people rapidly extract trait information from faces (5,
6), we also show that instructions to deliberate and make a good
judgment led to less accurate predictions of the election outcomes.
These findings are consistent with research showing that deliber-
ation can interfere with the quality of unreflective judgments (7)
and even with judgments that can be characterized on simple
quality of jams and making trait judgments from faces are quite
different, but both processes rely on mechanisms that are most
describing a face can interfere with face recognition (14, 15), and
thinking about the reasons for liking faces can reduce the consis-
tency of liking judgments (8).
The current findings contribute to a growing body of evidence
that the outcomes of important elections can be predicted from
person judgments (refs. 3 and 16, and D. Benjamin and J. Shapiro,
personal communication). In the research of Benjamin and Sha-
piro, participants predicted the outcomes of gubernatorial races
after observing 10 s of gubernatorial debates. When the sound of
the debate was off or muffled, these judgments predicted the
predictor of the vote share after controlling for incumbency,
campaign spending, and a number of economic indicators. Inter-
estingly, when the sound was on, predictions were at chance,
suggesting that the useful information in terms of prediction was
nonverbal and that inferring the party affiliation of the candidates
and policy positions led to worse predictions. These findings are
consistent with a large body of evidence in social psychology that
‘‘thin slices’’ of nonverbal behaviors can provide sufficient infor-
mation for accurate social judgments (17–23). The current findings
show that in the case of elections, even 100 ms of exposure to the
the election outcomes.
A recent study suggests that presidential elections can be pre-
dicted by face judgments too. Using a morphing technique, Little et
al. (ref. 16, study 1) created faces based on the shape differences
between the candidates for the highest posts in the United States,
United Kingdom, Australia, and New Zealand. These novel pairs
of faces, although derived from the politicians’ faces, are not
recognizable by participants. Remarkably, participants were more
procedure. We have shown that simulated voting decisions are
competence judgments and cast hypothetical votes for faces. Most
likely, when faced with a voting choice between two faces, partic-
ipants make a rapid judgment of competence and base their voting
decision on this judgment.
How do effects of facial appearance play out in the real world?
Certainly, having a competent face is not sufficient for electoral
success. If a politician does not have the backing of one of the two
major parties in the United States, his or her face would not make
the candidates represented these parties. Having the support of a
major party, a politician with competent appearance can have
higher chances of electoral success. However, competence as
assessed in our studies is always relative. Thus, in some races a
politician may appear more competent relative to the challenger,
and in others they may appear less competent. Finally, there are
multiple routes through which competent appearance can affect
electoral outcomes. For example, party leaders can promote com-
petent-appearing candidates for key positions even though these
decisions, competent appearance most likely would not affect
strongly identified partisans but can affect voters who are not
strongly identified with a party. In many cases, these are precisely
the voters who can swing an election. Appearance can also affect
deter undecided voters, who have a mild preference for the
decision processes will be critical to delineate the causal influences
of appearance on electoral success. Our findings suggest that, in
many cases, the effects of appearance on voting decisions may be
subtle and not easily recognized by voters (cf. ref. 24).
We focused on judgments of competence because of our prior
work, which showed that people believe that competence is one of
the most important traits for a politician and that competence
judgments predict election outcomes (3). However, the context of
election can change the relative importance of traits and, conse-
quently, voters’ preferences. For example, Littleet al. (ref. 16, study
2), using the morphing procedure described above, showed that
of war’’ context but preferred the morphed John Kerry face ‘‘in a
time of peace’’ context. The former face was perceived as more
masculine and dominant but less intelligent and forgiving. This
finding suggests that ‘‘fitting the face to the context’’ may be a more
important factor in elections than having a competent appearance.
Experiment 1. Participants. One hundred and twenty Princeton Uni-
for $5. Participants were randomly assigned to one of six experi-
mental conditions: 3 (presentation time: 100 ms vs. 250 ms vs.
unlimited time) ? 2 (counterbalancing of the position of the
images: left vs. right). In prior bootstrapping simulations (see
supporting online material for ref. 3), we have shown that reliable
estimates of the perceived competence of the candidates can be
obtained from a sample of ?40 participants. Thus, we recruited 40
participants for each of the presentation time conditions.
Gubernatorial races. Using the Almanac of American Politics (25), we
compiled a list of all gubernatorial races from 1995 to 2002,
Pictures of the winner and the runner-up were collected from
various Internet sources (e.g., CNN, Wikipedia, and local media
sources). Seven races were unusable because standardized pictures
of both major candidates could not be found. For the remaining 89
races, the image of each politician was cropped to 150 ? 215 pixels,
race pair was combined into a single image with 30 pixels of white
the other half. The position of the images was counterbalanced
across participants. In Experiment 2, we used only those races in
which the candidates were of the same sex and ethnicity (n ? 55).
In Experiment 3, we used the same procedure to standardize the
images of the candidates in the 2006 election.
Procedures. The instructions in all conditions emphasized that
candidates were mentioned at any point.
The order of the 89 races was randomized for each participant.
For each race, participants made three consecutive judgments: a
binary competence judgment, a nine-point scale competence judg-
ment for the person selected as more competent, and a recognition
judgment. The intertrial interval was 1 s. Each trial started with a
fixation cross (?) presented for 500 ms. The race pair image was
www.pnas.org?cgi?doi?10.1073?pnas.0705435104 Ballew and Todorov