ArticlePDF Available

What Goes Up Apparently Needn't Come Down: Asymmetric Predictions of Ascent and Descent in Rankings

Authors:

Abstract and Figures

In eight studies, we document an upward mobility bias, or a tendency to predict that a rise in rankings is more likely than a decline. This asymmetry was observed in predictions of classroom performance, NBA and NFL standings, business school rankings, and employee performance rankings. The bias was found for entities people care about and want to see improve their standing, as well as entities in which people are not invested. It appears to result from people's tendency to give considerable weight to a focal agent's intentions and motivation, but to give less weight to the intentions of competitors and other factors that would thwart the focal agent's improvement. We show that this bias is most pronounced for implicit incremental theorists, who believe that performance is malleable (and hence assign more weight to intentions and effort). We discuss implications of this asymmetry for decision making and for an understanding of the underdog bias. Copyright © 2015 John Wiley & Sons, Ltd.
Content may be subject to copyright.
What Goes Up Apparently Neednt Come Down: Asymmetric Predictions of
Ascent and Descent in Rankings
SHAI DAVIDAI
1
* and THOMAS GILOVICH
1
1
Department of Psychology, Cornell University, Ithaca, NY, USA
ABSTRACT
In eight studies, we document an upward mobility bias, or a tendency to predict that a rise in rankings is more likely than a decline. This asym-
metry was observed in predictions of classroom performance, NBA and NFL standings, business school rankings, and employee performance
rankings. The bias was found for entities people care about and want to see improve their standing, as well as entities in which people are not
invested. It appears to result from peoples tendency to give considerable weight to a focal agents intentions and motivation, but to give less
weight to the intentions of competitors and other factors that would thwart the focal agents improvement. We show that this bias is most pro-
nounced for implicit incremental theorists, who believe that performance is malleable (and hence assign more weight to intentions and effort).
We discuss implications of this asymmetry for decision making and for an understanding of the underdog bias. Copyright © 2015 John Wiley
& Sons, Ltd.
key words upward mobility bias; focalism; relative performance; prediction
In an often-told tale, two campers see a bear running toward
them and realize theres no way they can outrun it. One ne-
vertheless proceeds to swap his hiking boots for a swifter
pair of sneakers. The other nds this odd and remarks, What
are you doing? You cant outrun a bear.”“I dont have to,
he replies, I only have to outrun you.
As this tale highlights, success in competitive settings is
measured in relative terms. Financial success is commonly
expressed in relative terms (e.g. the 1%;middle class),
students are frequently graded on a curve,sport teams ad-
vance to the playoffs based on how their record compares to
others, and workers are named employee of the month
whether absolute performance that month was strong or
weak. Predicting performance in competitive settings there-
fore involves an extra layer of complexity, as it requires
assessing not just the attributes of a target person, team, or
organization, but the attributes of its competitors as well.
How, and how well, do people make such predictions?
Predicting relative and absolute performance
Attribution theorists have explored the naïve psychological
processes underlying peoples understanding of behavior.
Nearly all of them have highlighted the distinction between
personal and impersonal causal factors. Heider (1958) sug-
gested that three factors affect a persons understanding of
an agents behavior: the agents motivation (i.e. the extent
to which she is trying to perform the behavior), her ability
(i.e. the extent to which she is able to perform the behavior),
and external circumstances beyond the agents control (i.e.
the extent to which environmental forces prevent, hinder, fa-
cilitate, or virtually guarantee the behavior). To the extent
that an individual is believed to be both motivated and able
to overcome external obstacles, she would be expected to
succeed at the task at hand. In contrast, to the extent that an
individuals motivation or ability is seen as insufcient to
overcome any inhibitory forces, she is considered likely to
fail. A student believed to be both motivated and intellectu-
ally able is expected to succeed in her courses. A student
known to be insufciently motivated or intellectually lacking
is expected to do poorly. When predicting achievement in
absolute settingswhere performance is measured accord-
ing to objective criteriathe only information necessary is
whether the persons motivation and ability surpass any
inhibiting external forces.
In competitive settings, prediction is more complex. Since
a targets ranking is, by denition, relative (the rise of one
person, team, or company necessarily entails the decline of
another), accurate predictions of performance must take into
account each competitors motivation and ability in addition
to those of the focal target.
Imagine a course in which students are graded on a curve.
To simplify matters, imagine that only two students are en-
rolled in the course, one of whom will receive an A while
the other will have to settle for a B. How would the professor
go about predicting the performance of Student X, when X
nervously asks her to do so during ofce hours? First, as
attribution theory suggests, the professor will need to infer
Student Xs motivation and ability from her past behavior
(e.g. her attendance record), accomplishments (e.g. her
GPA), and so on. Second, the professor will need to assess
the attributes of the other student, Student Yher motiva-
tion and ability. Third, the professor will need to account
for any external circumstances that might affect one student
more than the other (e.g. Student Xsnancial situation or
extracurricular entanglements, Student Ys health issues).
To predict Student Xs performance, all of these factors must
be considered.
Now imagine instead that 30 students are enrolled in the
course. How would the professor go about predicting Student
*Correspondence to: Davidai, Shai, Cornell University, Department of
Psychology. Ithaca, NY, USA. E-mail: sd525@cornell.edu
Copyright © 2015 John Wiley & Sons, Ltd.
Journal of Behavioral Decision Making,J. Behav. Dec. Making,28: 491503 (2015)
Published online 6 March 2015 in Wiley Online Library (wileyonlinelibrary.com) DOI: 10.1002/bdm.1865
Xs performance in such a setting? In addition to assessing
Student Xs motivation and ability, the professor would
now need to consider each of the other 29 studentsattri-
butes. Furthermore, beyond considering the impact of exter-
nal circumstances shared by all students (e.g. time allotted
for the nal exam), the professor would need to consider
each students individual circumstances (e.g. Student Zs
family situation) as well as circumstances shared by subsets
of students (e.g. students with time extensions). Obviously,
computing the likely impact of factors that affect relative
performance in competitive settings is a challenge.
Given all the difculties involved, it is unlikely that peo-
ple predict performance in competitive settings the same
way they do in non-competitive settings, where performance
is measured relative to an objective benchmark. The sheer
number of factors that need to be taken into account compli-
cate the already challenging task of predicting uncertain
future events.
An upward mobility bias in predictions of relative
performance
It is not uncommon for people to substitute an easy assess-
ment that comes to mind automatically for the more chal-
lenging judgment thats actually required (Kahneman &
Frederick, 2002). When it comes to the computationally
complex task of predicting relative performance, a likely
substitution is the relatively simple assessment of likely ab-
solute performance. Focusing on a targets absolute attributes
rather than how those attributes compare to those of the tar-
gets competitors greatly simplies the prediction process.
Indeed, evidence from several lines of research supports the
idea that both personal (i.e. motivation and ability) and
impersonal (i.e. external circumstances) factors that might
affect the performance of the targets competitors are barely
taken into considerationif at all. That is, when making pre-
dictions of relative standing (Klar & Giladi, 1997; Kruger,
1999; Kruger & Burrus, 2004) or when predicting the out-
come of explicit competitions (Moore, 2005; Moore &
Kim, 2003; Radzevick & Moore, 2008; Windschitl et al.,
2003), people often focus exclusively on a targets personal
attributes and all but ignore its competitorscapabilities.
If, as past research suggests, people ignore the efforts,
abilities, and actions of the targets competitors, on what, ex-
actly, are they likely to focus? Following Heider (1958), we
predict that people tend to focus on a targets effort and abil-
ity. In many circumstances, furthermore, it is much easier to
assess whether the requisite effort is present than to assess
whether the target has sufcient ability. In this paper, we pro-
pose that the narrow focus on a targets motivation and effort
leads to an upward mobility bias in predictions of perfor-
mance: that is, people tend to predict that a rise in rankings
is more likely than a decline. Because people generally want
to improve their standing in a competition, a focus on the
motivation of a lowly ranked target will give credence to
the prospect of her moving up. At the same time, because
those who are highly ranked generally want to stay in the
higher echelon, focusing on a highly ranked targets motiva-
tion will make the prospects of a decline seem remote.
Declines, after all, are not intended, and happen despite a
persons efforts. But of course, rankings are by denition
zero-sum: one persons (or teams) rise in rankings is always
accompanied by anothers decline.
To illustrate, imagine once again that a student asks her
professor to predict her likely performance on the nal exam,
based on her midterm performance. If, as we argue, the pro-
fessor engages in a type of attribute substitution, she will fo-
cus her assessment on this particular students ability and
(especially) motivation, and neglect the attributes of the other
students in the class. As long as the professor believes that
her student is sufciently able and motivated, she is likely
to predict that Student X will rank higher (relative to the
other students) than she had ranked on the midterm. This
leads to a logical impossibility: students near the bottom of
the class will generally be seen as motivated to do better
and therefore be expected to rise in the rankings; at the same
time, students ranked near the top of the class will likely be
seen as motivated to remain at the top, and therefore ex-
pected to stand their ground. But because a rise of one stu-
dent necessitates the decline of another, these two
predictions are in conict. Although intuitively appealing
(studentsmotivation and ability should affect their perfor-
mance) predictions that focus solely on a given student and
disregard the motivations and talents of the rest of the class
are inherently awed.
To examine this account of the upward mobility bias, we
asked participants to predict the relative performance of a fo-
cal target in various settings, including athletics (Studies 1
and 3A), academia (Studies 2, 3B, 4, 5A, and 5B), and em-
ployment (Study 6). Participants considered an individual
(or team of individuals) previously ranked high or low in
the domain in question and were asked to predict the future
ranking of that same individual (or team). In line with the
upward mobility bias, we predicted that participants would
see a rise in rankings as more likely than a decline. We
predicted that this would be the case both when partici-
pants might be interested in seeing the targets succeed
(Study 1) and when they likely had no rooting interest
(Studies 26). Furthermore, we predicted that this effect
would result from participantsoverreliance on the attri-
butes of the target of prediction, especially the targets
perceived motivation. Among the consequences of this
focus on the targets motivation, we predicted that incre-
mental theorists,who believe that traits and abilities
are subject to change and who prize and attend to effort
(Dweck, 1999), would be especially prone to the upward
mobility bias.
STUDY 1: NATIONAL FOOTBALL LEAGUE (NFL)
RANKINGS
We rst asked participants to predict performance in a do-
main in which rankings and relative assessments are ubiqui-
tous: sports. We examined whether sports fans, who are
presumed to know a lot about different teams and are familiar
with the relativity of performance, would predict that a team
near the bottom of the standings is more likely to rise in
492 Journal of Behavioral Decision Making
Copyright © 2015 John Wiley & Sons, Ltd. J. Behav. Dec. Making, 28, 491503 (2015)
DOI: 10.1002/bdm
rankings than a team near the top of the standings is to drop.
We asked discussants in an on-line sports forum to assess the
future standing of NFL teams based on their past standing.
We predicted that participants would exhibit an upward mo-
bility bias and assign a higher likelihood of low-ranked
teams rising in rankings than high-ranked teams dropping
in them.
Method
Participants
Sixty-nine participants (1 female, M
age
= 24.26), recruited
from a popular NFL fan forum,
1
volunteered for this study.
Materials and procedure
Several times each year, the popular sports network ESPN
publishes a power rankingof the 32 National Football
League (NFL) teams based on the assessments of its sports
analysts. Ten days following the publication of ESPNs
post-NFL-draft power ranking (23 May 2014), participants
were presented with the rank assigned to eight teams and
asked to predict each teams ranking at the end of the follow-
ing season. To simplify their task, participants were asked to
assess the likelihood that next year a given team would nish
among the top or bottom 16 of the league. For example,
some participants estimated the likelihood that the Seattle
Seahawks (ranked #1) would either remain in the top 16 of
the league or drop to the bottom 16. The eight teams were
presented randomly, such that each participant predicted the
performance of one team from the top 4 teams, one team
from the second 4 (places 58), one team from the third 4
(places 912), and so forth.
2
For each participant, the teams
in the top and bottom halves of the rankings were matched
to be equally distant from the middle rank. For example, par-
ticipants who predicted the future standing of the Seattle
Seahawks (ranked #1), the New Orleans Saints (#5), the Phil-
adelphia Eagles (#9), and the Baltimore Ravens (#13) also
predicted the future standing of the Cleveland Browns
(#32), the Washington Redskins (#28), the Miami Dolphins
(#24), and the Detroit Lions (#20). Participants typed their
percentage estimates of the likelihood of a given team ending
up in the bottom half and top half of the rankings and were
unable to proceed if the sum did not equal 100%.
After eliciting their predictions, we assessed participants
interest in football with 3 questions: To what extent do you
consider yourself a football fan?(1I dont consider my-
self a fan at all, 5I am an avid fan), To what extent do
you follow the NFL?(1Not at all, 5Very much so),
and To what extent do you care about the outcomes of
games in the NFL?(1Not at all, 5Very much so). We
averaged these questions to create an index of participants
interest in football (Cronbachsα= 0.84). As expected given
the subject pool, all but 2 participants rated themselves as ex-
tremely avid fans (M = 4.72, SD = 0.47).
Results
We averaged participantslikelihood estimates for the four
low-ranking and the four high-ranking teams to create com-
posite measures of the anticipated likelihood of a rise in rank-
ings and a drop in rankings, respectively. For the rise in
rankings measure, we averaged participantsestimates that
each of the four teams from the bottom 16 of the current
power rankings (places 1732) would rise to the top 16 the
following year. For the drop in rankings measure, we aver-
aged participantsestimates that each of the four teams from
the top 16 of the current rankings (places 116) would drop
to the bottom 16 the following year.
Consistent with our hypothesis, participants exhibited an
upward mobility bias in their predictions of teamsrankings
from one year to the next. Participants estimated that a team
previously ranked in the bottom half of the league would be
signicantly more likely to rise to the top half of the rankings
(M = 33.93%, SD = 10.83) than a team ranked in the top half
would be to drop to the bottom half (M =28.91%,
SD = 10.80), t(68) = 2.99, p<.005. This was most pro-
nounced for teams farthest from the middle of the pack.
Despite being equally distant from the mid-rank, participants
thought that a team ranked in the bottom 4 of the league was
signicantly more likely to rise to the top 16 (M =25.41%,
SD = 23.49) than a team ranked in the top 4 was to drop to
the bottom 16 (M = 12.78%, SD = 18.05), t(68) = 5.07,
p<.0001. They also thought that a team ranked 25th28th
was more likely to rise to the top 16 (M = 27.68%,
SD = 16.75) than a team ranked 5th8th was to drop to the
bottom 16 (M = 21.88%, SD = 16.77), t(68) = 2.08, p<.05.
Participantsestimates for the teams ranked 2124 and
912 and the teams ranked 1720 and 1316 did not differ
signicantly from reach other.
3
STUDY 2: BUSINESS SCHOOL RANKINGS
Participants in Study 1 exhibited an upward mobility bias in
predictions of performance in the NFL: they estimated that
low-ranked teams are more likely to rise in the standings than
high-ranked teams are to drop. Of course, this is impossible:
given that NFL standings are a zero-sum game, the rise of one
team requires the drop of another. It is noteworthy that partic-
ipants exhibited this bias even though they predicted the per-
formance of specic teams (e.g. the Saints and Dolphins)
rather than abstract entities (the 5
th
rankedand 24
th
ranked
teams) and despite making several predictions (which should
highlight the relativity of each teamsperformance).
1
http://www.reddit.com/r/n/
2
Participants were randomly presented with eight teams drawn from one of
the following four sets: (#1, #5, #9, #13, #20, #24, #28, #32), (#2, #6,
#10, #14, #19, #23, #27, #31), (#3, #7, #11, #15, #18, #22, #26, #30), or
(#4, #8, #12, #16, #27, #21, #25, #29).
3
Not surprisingly given that there was so little variability in the measures of
participantsinterest in the NFL, interest in professional football did not
qualify any of our results.
The Upward Mobility Bias 493S. Davidai and T. Gilovich
Copyright © 2015 John Wiley & Sons, Ltd. J. Behav. Dec. Making, 28, 491503 (2015)
DOI: 10.1002/bdm
In Study 2, we examined the robustness of these ndings
and addressed a potential artifact in Study 1. It is likely that
participants (who were avid football fans) were reluctant to
indicate that their favorite team would drop in the rankings.
To the extent that participants were likely to invest psycho-
logically in winners and root for teams ranked at the top of
the league, they may have inated their estimates of highly
ranked teams staying at the top out of self-interest. Indeed,
when asked about their favorite NFL team, 69% of partici-
pants listed a team in the top half of the league, and 37%
named a team in the Top 4. Therefore, the upward mobility
bias found in Study 1 may have been the result of partici-
pantsunwillingness to imagine their team dropping in rank-
ings (rather than a general tendency to posit a great deal of
movement upwards on the part of lower-ranked teams). In
Study 2, we examined predictions in a domain in which par-
ticipants had little or no prior exposure to the entities in ques-
tion and were unlikely to have much of a rooting interest in
their rankings.
Method
Participants
One hundred six Mechanical Turk participants (62 females,
M
age
= 35.30) completed the study in exchange for modest
monetary compensation.
Materials and procedure
Every two years, Bloomberg Businessweek magazine pub-
lishes its ranking of the top full-time MBA programs in the
United States, based on past and present studentssatisfac-
tion with their education, the views of corporate recruiters,
and statistics such as a schools placement record in top
rms. Participants were presented with the rankings of the
Top 20 business schools as ranked in 2012. They were also
informed of each programs cost, enrollment, geographical
location, and current dean. They were reminded of the im-
portance of such rankings in inuencing applicantsenroll-
ment decisions and were asked to predict the schools2014
rankings. To engage participants in the prediction task, they
were asked to consider the factors that are likely to play a
role in determining the future ranking of the school. Partici-
pants were then given time to type their thoughts about what-
ever factors seemed important to them before continuing to
the prediction task.
After reviewing the rankings and the information about
each school, participants estimated the likelihood that a given
schools standing would change from one ranking to the
next. Specically, participants were asked to assess the like-
lihood that a given school would nish among the top 5, the
second 5, the third 5, or the bottom 5 of the rankings in 2014
based on its rank in 2012. For example, participants who
were presented with the University of Chicago (ranked #1)
were asked to estimate the likelihood that it would either re-
main in the top 5 or drop to each of the other quadrants in the
rankings. Participants were randomly presented with two
schools: one school from the top 10 and one from the bottom
10 of the rankings. They typed their percentage estimates of
the likelihood of a given school ending up in each of the
quadrants, and were unable to proceed if the sum did not
equal 100%.
Results
Each participant made two sets of predictionsone for a
high-ranking school and one for a low-ranking school. We
therefore computed for each participant two likelihood as-
sessmentsthat a school in the bottom 10 would rise to the
top 10 and that a school in the top 10 would drop to the bot-
tom 10. For the rise in rankings measure, we summed partic-
ipantsestimates that a school from the bottom 10 (i.e. ranks
1120) would rise to either the top 5 or the second 5 in the
rankings. For the drop in rankings measure, we summed par-
ticipantsestimates that a school from the top 10 would either
drop to the third 5 or the bottom 5.
Participants exhibited an upward mobility bias in their
predictions of business school rankings. They estimated that
a school previously ranked in the bottom 10 would be signi-
cantly more likely to rise to the top 10 (M = 27.21%,
SD =20.86) than a school ranked in the top 10 would be to drop
to the bottom 10 (M = 16.05 %, SD = 17.87), t(105) = 4.68,
p<.0001. Furthermore, despite being equally distant from
the middle rank, participants believed that a school ranked
in the bottom 5 (16
th
20
th
) would be signicantly more
likely to rise to the top 10 (M= 23.30 %, SD = 22.42) than
a school ranked in the top 5 would be to drop to the
bottom 10 (M = 9.21%, SD = 12.13), t(32) = 3.22, p<.005,
and that a school ranked in the third 5 (11
th
15
th
) would
be more likely to rise to the top 10 (M = 34.92, SD = 20.61)
than a school ranked in the second 5 (6
th
10
th
) would be
to drop to the bottom 10 (M = 20.92, SD = 17.10), t(24) = 2.93,
p<.01.
STUDIES 3A AND 3B
Studies 1 and 2 provided support for the upward mobility
bias in predictions of relative performance. When predicting
the future standing of NFL teams or MBA programs, partic-
ipants considered a rise in rankings to be more likely than a
decline. In both cases, participants had quite a bit of informa-
tion about the targets in question. Participants in Study 1
were avid NFL fans and likely knew a great deal about the
personnel and off-season personnel changes of the different
NFL teams. Participants in Study 2 were given detailed infor-
mation about the top 20 MBA programs in the country. This
information may have facilitated the upward mobility bias by
enabling peoples tendency to process information in ways
that conrm preconceived notions (Cohen, 1981; Gilovich,
1991; Lord et al.1979). That is, participants may have selec-
tively attended to the information about the team or school
whose ranking they were asked to predict and construed it
in ways that implied a rise in rankings. For example, when
asked to predict the performance of a school with a high en-
rollment rate, participants may have seen the enrollment as a
signal of the schools popularity and therefore as supportive
of its likely rise in rankings. However, when asked to predict
494 Journal of Behavioral Decision Making
Copyright © 2015 John Wiley & Sons, Ltd. J. Behav. Dec. Making, 28, 491503 (2015)
DOI: 10.1002/bdm
the performance of a school with a low enrollment rate, par-
ticipants may have treated that piece of information as indic-
ative of the schools selectivity and therefore as similarly
implying a rise in the rankings.
Although biased processing of specic items of informa-
tion may play some role in the upward mobility bias, we
dont believe it is the main determinant. Instead, we argue
that the bias is mainly due to peoples overreliance on the tar-
gets motivation. When predicting the performance of a low-
ranked team, school, or individual, people are likely to focus
on its drive to succeed and therefore believe that it is likely to
rise in the rankings. In contrast, when predicting the perfor-
mance of a high-ranked team, school, or individual, people
are likely to focus on its drive to remain successful and there-
fore believe that it is likely to remain near the top of the rank-
ings. Indeed, when an entity experiences a drop in rank, it is
generally despite its efforts, not because of them. If such a
narrow focus on motivation is the driving force behind the
upward mobility bias, then the bias should still be present
when predicting the performance of abstract entitiesthose
about whom one has no concrete, detailed knowledge. We
tested this idea in Studies 3A and 3B. Participants in Study
3A predicted the future rankings of sport teams based on
their past standings. Unlike Study 1, these predictions were
made about abstract entities (e.g. the 11
th
ranked team) and
hence in the absence of any detailed information about the
team in question. Study 3B was a conceptual replication in
the domain of class standing.
STUDY 3A: NATIONAL BASKETBALL LEAGUE
(NBA) RANKINGS
Method
Participants
One hundred eighty-eight Mechanical Turk participants (100
females, M
age
= 35.41) completed the study in exchange for
modest monetary compensation.
Materials and procedure
The National Basketball Association (NBA) has 30 teams
that, at the end of each regular season, can be ranked accord-
ing to win-loss record (which is done, for example, to help
establish the order in which teams can draft collegiate and el-
igible foreign players). Participants were asked to estimate
the probability that a teams ranking would differ from one
season to the next. To control for the strength of the teams
participants were asked to consider, we specically drew
their attention to teams ranked in the top or bottom tertiles.
In the rise in rankings condition, participants were asked to
estimate the likelihood that a team ranked 30
th
,26
th
,or21
st
one year would either remain in the bottom 10 of the league
or rise to the middle or top 10 the following year. In the drop
in rankings condition, participants were asked to estimate the
likelihood that a team ranked 1
st
,5
th
,or10
th
in one year
would either remain in the top 10 or drop to the middle or
bottom 10 the following year. Note that teams ranked 1
st
and 30
th
are equidistant from the middle 10 of the league,
as are teams ranked 5
th
and 26
th
, and teams ranked 10
th
and
21
st
. Participants typed their percentage estimates, and were
unable to proceed if their estimates about each team did not
sum to 100%.
After eliciting their predictions, we assessed participants
interest in basketball with 3 questions: To what extent do
you consider yourself a basketball fan?(1I dont consider
myself a fan at all, 5I am an avid fan), To what extent do
you follow the NBA?(1Not at all, 5Very much so),
and To what extent do you care about the outcomes of
games in the NBA?(1Not at all, 5Very much so).
We averaged these questions to create an index of interest
in the NBA (Cronbachsα= 0.95).
Results
We summed participantsestimates of the likelihood that a
teams standing would differ from one year to the next. In
the rise in rankings conditions, we summed participantses-
timates that a team previously ranked in the bottom 10
(ranked 30
th
,26
th
,or21
st
) would be ranked in the middle
or top 10 in the following year. In the drop in rankings con-
ditions, we summed participantsestimates that a team previ-
ously ranked in the top 10 (ranked 1
st
,5
th
,or10
th
) would be
ranked in the middle or bottom 10.
Replicating the ndings from Studies 1 and 2, participants
exhibited an upward mobility bias in their predictions of
teamsranking from one year to the next. As shown in
Figure 1, participants estimated that a team previously
ranked 30
th
(M = 53.62%, SD = 21.07), 26
th
(M = 50.96%,
SD = 20.35) or 21
st
(M = 59.19 %, SD = 19.02) would be sig-
nicantly more likely to rise in the rankings than a team
ranked 1
st
(M = 43.71%, SD = 25.84), 5
th
(M = 41.78%,
SD = 16.89) or 10
th
(M = 48.30%, SD = 16.98) would be to
drop in the rankings. A planned contrast between the three
rise in rankings estimates and the three drop in rankings es-
timates was signicant, F(1,182) = 11.23, p= .001. Further-
more, despite being equally distant from the middle 10,
participants believed that a team ranked 21
st
(M = 41.15%,
Figure 1. Perceived likelihood of a rise/drop in NBA rankings by
previous season ranking (Study 3A)
The Upward Mobility Bias 495S. Davidai and T. Gilovich
Copyright © 2015 John Wiley & Sons, Ltd. J. Behav. Dec. Making, 28, 491503 (2015)
DOI: 10.1002/bdm
SD = 17.13) was signicantly more likely to move to the
middle 10 than a team ranked 10
th
(M = 31.45%,
SD = 10.79), F(1, 182) = 7.93, p= .005, that a team ranked
26
th
(M = 32.08%, SD = 11.36) was marginally more likely
to do so than a team ranked 5
th
(M = 25.94%, SD = 10.43),
F(1, 182) = 3.03, p= .08, and that a team ranked 30
th
(M = 37.50%, SD = 17.36) was signicantly more likely to
do so than a team ranked 1
st
(M = 26.22%, SD = 15.19),
F(1, 182) = 10.02, p= .002. These effects remained un-
changed after controlling for participantsself-reported in-
terest in the NBA.
STUDY 3B: CLASS STANDING
Method
Participants
One hundred twenty-two Cornell University undergraduates
(96 females, M
age
= 19.75) participated for extra credit in
their psychology and human development courses.
Materials and procedure
Participants were asked to estimate the probability that a stu-
dents grade would differ from one exam to the next. Specif-
ically, participants were asked to imagine a course in which
30 students were graded on a curve, such that the top 10 stu-
dents on each exam receive an A, the middle 10 receive a B,
and the bottom 10 receive a C. They then estimated the like-
lihood of a student with a particular grade on the rst exam
receiving each of the three grades on the second exam. In
the rise in rankings condition, participants estimated the like-
lihood that a randomly selected student who received a C on
the rst exam would receive an A, B, or C on the second. In
the drop in rankings condition, participants estimated the
likelihood that a randomly selected student who received
an A on the rst exam would receive each of the three grades
on the second. Participants wrote their estimates in percent-
ages, with a written reminder that they should total 100%.
4
Results
In the rise in rankings condition, we summed participants
estimates that a student who received a C on the midterm
would do better (receive an A or B) on the nal; in the drop
in rankings condition, we summed participantsestimates
that a student who received an A on the midterm would do
worse (receive a B or C) on the nal.
As expected, participants exhibited an upward mobility
bias, estimating that a rise in rankings is signicantly more
likely than a decline. Whereas participants in the rise in rank-
ings condition estimated that a student who was previously at
the bottom 10 of the class had a 56% (SD = 12.35) chance of
rising to the middle 10 or higher, participants in the drop in
rankings condition estimated that a student who was previ-
ously in the top 10 had only a 43% (SD = 15.69) chance of
dropping to the middle 10 or lower, t(117) = 4.91,
p<.0001. Furthermore, despite being equally distant from
the middle 10, participants believed that a randomly selected
student from the bottom 10 had a 35% (SD = 8.29) chance of
rising to the middle 10 (i.e. receive a B following a C), but
that a student from the top 10 had only a 28% (SD = 7.98)
chance of dropping to the same middle third, t(117) = 4.75,
p<.0001.
5
STUDY 4: AN ALTERNATIVE ACCOUNT TO THE UP-
WARD MOBILITY BIAS
Studies 3A and 3B testify to the robustness of the upward
mobility bias by showing that it occurs even when partici-
pants do not have any detailed information about the target
entities in question and are estimating the performance of ab-
stract targets (e.g. the 12
th
ranked team). Participants esti-
mated that low-ranked teams (Study 3A) and students (3B)
were signicantly more likely to rise in the rankings than
high-ranked targets were to drop.
We contend that the upward mobility bias is due to a kind
of attribute substitution: that peoples lay theory of motiva-
tionthat most people strive to improve a weak performance
but dont strive to get worseis highly accessible and is all-
but substituted for the difcult assessment of how the person or
team in question will stack up to the competition. An alterna-
tive account of this bias is that people simply anchor on the
most salient position in any rankingsthe top positionand
insufciently adjust from it (Epley & Gilovich, 2001, 2006;
Tversky & Kahneman, 1974). That is, the salience of 1or
topexerts a pull on peoples estimates ofrising and declining
in rankings, leading the former to be higher than the latter.
We tested this alternative account in Study 4 by manipu-
lating the desirability of rising in ranking. Being highly
ranked is often desirable. It is desirable for a country to be
ranked high in per capita income, for a university to be
ranked high in terms of Nobel Prize winners, or for a basket-
ball team to be ranked high in assists. In those circumstances,
the upward mobility bias can be explained both by an an-
choring account as well as an attribute substitution account.
But sometimes a drop in ranking is more desirable. It is not
desirable for a country to be ranked high in corruption, for
4
Three participantslikelihood assessments exceeded 100%. Since these par-
ticipants either did not understand or did not pay attention to the directions,
we excluded their data from the analyses.
5
Because participants were asked to estimate the likelihood that a student
would receive each of the three grades (A, B or C), it could be that we inad-
vertently cued them to think about performance in the classroom in absolute
rather than relative terms. We therefore replicated this study with minor
changes. We presented 148 Mechanical Turk participants (81 females,
M
age
= 33.92) with a similar scenario, but rather than having them predict a
random students grade, they were asked to predict the students ranking
on the second exam based on her ranking in the rst exam (e.g. given that
a student was ranked in the bottom 10, what is the likelihood that she will
rise to the middle 10). The instructions made no reference to specic grades.
Participants still exhibited an upward mobility bias, estimating that a student
from the bottom 10 had a 51% (SD = 19.88) chance of rising to the middle
10 or higher, but that a student from the top 10 had only a 35%
(SD = 21.44) chance of dropping to the middle 10 or lower t(146) = 4.56,
p<.0001.
496 Journal of Behavioral Decision Making
Copyright © 2015 John Wiley & Sons, Ltd. J. Behav. Dec. Making, 28, 491503 (2015)
DOI: 10.1002/bdm
a university to be ranked high in sexual assault charges, or
for a basketball team to be ranked high in turnovers. In these
circumstances, the two accounts make different predictions.
According to the anchoring account, people anchor on the
top rank and insufciently adjust from it, leading to estimates
of a rise that are higher than estimates of a decline both when
a rise in ranking is desirable and when it is not. In contrast,
the desirability of a rise in ranking is critical to the attribute
substitution account. When being highly ranked is undesir-
able, the relevant targets are motivated to drop in rankings.
Therefore, when it comes to undesirablerankings, people
should see a drop in ranking as more likely than a rise. In
Study 4, we tested this prediction by examining whether
participants would see a rise in rankings as more likely than
a decline only when a rise is desirable.
Method
Participants
One hundred twenty six Mechanical Turk participants (59 fe-
males, 61 males, 6 unspecied, M
age
= 33.89) completed the
study in exchange for modest monetary compensation.
Materials and procedure
Participants were asked to estimate the probability that a stu-
dents absentee rate would differ from one half of the semes-
ter to the next. Specically, participants were asked to
imagine a class of 30 students in which the teacher kept track
of the studentsnumber of absences. To examine the two
accounts of the upward mobility bias, we manipulated the
desirability of being highly ranked. In the desirable-ranking
condition, participants were told that the rankings ranged
from the students with the least absences (top 10) to those
with the most (bottom 10). Therefore, to get good grades,
the students try to rise in the rankings by minimizing their
absences.In the undesirable-ranking condition, participants
were told that the rankings ranged from students with the
most absences (top 10) to those with the least (bottom 10).
In this condition, participants read that to get good grades
the students try to drop in the rankings by minimizing their
absences.We crossed ranking desirability with direction of
prediction in a 2 × 2 between-subjects design. In the rise in
rankings condition, participants were asked to assess the
likelihood that a randomly selected student who was in the
bottom 10 of the class in absences in the rst half of the
semester would rank in each third of the distribution on the
second half. In the drop in rankings condition, participants
assessed the likelihood that a randomly selected student in
the top 10 of the class would rank in each third of the distri-
bution on the second half.
If the upward mobility bias is due to participants anchor-
ing on the top rank, then the desirability of being highly
ranked should be irrelevant to predictions of future absences.
However, if the bias is due to overweighting the impact of
the targets motivation to better her standing, the desirability
of being highly ranked should signicantly inuence partici-
pantsestimates.
Results
In the rise in rankings condition, we computed a measure of
estimated likelihood of change in the students ranking by
summing participantsestimates that a student who previ-
ously ranked at the bottom 10 would rank in the middle 10
or higher. In the drop in rankings condition, we summed
their estimates that a student who had previously ranked in
the top 10 would rank in the middle 10 or lower.
As expected, the interaction between condition (desirable-
ranking vs. undesirable-ranking) and direction of change
(rise or decline) was signicant, F(1, 125) = 10.82, p= .001
(see Figure 2). When a rise in ranking was desirable, partic-
ipants estimated that a rise in ranking (M = 53.61%,
SD = 19.85) is more likely than a decline (M = 37.33%,
SD = 24.22), t(125) = 2.83, p= .005. In contrast, when a rise
in ranking was undesirable, participants exhibited the oppo-
site pattern, estimating that a drop in ranking (M = 45.42%,
SD = 26.44) is marginally more likely than a rise
(M = 34.79%, SD = 20.01), t(125) = 1.82, p= .07. Although
these data indicate that anchoring may indeed play a role in
predictions of this sortstrengthening the upward mobility
bias in the desirable-ranking condition and weakening it in
the undesirable-ranking conditionthey cast doubt on a sim-
ple anchoring account of the bias documented in the studies
we report here. The signicant interaction provides clear sup-
port for the attribution substitution account.
STUDIES 5A AND 5B
Another alternative account of the upward mobility bias is
that people are simply motivated to think that positive events
are more likely than negative events. Because a rise in rank-
ings is a more positive event than a decline, participants may
be eager to see the former as more likely than the latter, even
when they have no attachment to the school (Study 2), team
(Study 3A), or student (Studies 3B and 4) in question.
We test this alternative account in the following studies by
manipulating the amount of control the target has over the
Figure 2. Perceived likelihood of a rise/drop in ranking when a rise
in ranking is desirable (desirable-ranking condition) and undesirable
(undesirable-ranking condition) (Study 4)
The Upward Mobility Bias 497S. Davidai and T. Gilovich
Copyright © 2015 John Wiley & Sons, Ltd. J. Behav. Dec. Making, 28, 491503 (2015)
DOI: 10.1002/bdm
outcome in question (Study 5A) and the motivation to exer-
cise that control (Study 5B). If participants are merely being
optimistic in thinking that a rise in rankings is more likely
than a decline, they should be equally likely to exhibit this
bias both when a person can exert considerable control over
her performance and when control is limited. A rise in rank-
ings is a positive outcome in either context. However, if the
upward mobility bias is due, as we claim, to overweighting
the targets motivation, participants should only exhibit the
bias when motivation is seen as both high and relevant to
performance. In Studies 5A and 5B, we predicted that partic-
ipants would see a rise in rankings as more likely than a
decline only when the target of prediction is willing and able
to inuence the outcome.
STUDY 5A: INFLUENCE OF PERCEIVED CONTROL
ON PREDICTIONS OF STUDENT PERFORMANCE
Method
Participants
One hundred three Mechanical Turk participants (47 females,
M
age
= 32.21) completed the study in exchange for modest
monetary compensation.
Materials and procedure
As in Study 3B, participants estimated the probability that a
students grade would differ from one exam to another. How-
ever, we manipulated the students perceived control over
her performance. In the high-control condition, participants
were told that the second exam was administered a month
following the rst one, that the students have a lot of extra
time to study between the rst and the second exam,and
that the available time is more than enough for the students
to rehearse the material again and make sure they know ev-
erything.In the low-control condition, participants were
told that the second exam was administered the day follow-
ing the rst exam, that the students do not have extra time
to study between the rst and the second exam,and that
the available time is not enough for the students to rehearse
the study material again and make sure that they know every-
thing. All the students can do is hope that the exam on the
second day will not include things they didnt know on the
rst day.In the rise in rankings condition, participants were
then asked to assess the likelihood that a randomly selected
student who was in the bottom 10 of the class on the rst
exam would rank in each third of the distribution on the sec-
ond exam. In the drop in rankings condition, participants
assessed the likelihood that a randomly selected student
who was in the top 10 on the rst exam would rank in each
third of the distribution on the second exam.
If the upward mobility bias is due to participants wanting
to see success as more likely than failure, the timing of the
two exams should be irrelevant to estimates of relative per-
formance. Participants should want a student who did poorly
on the rst exam to do better on the second one (rise in rank-
ings condition) and a student who already did well on the
rst exam to keep doing well (drop in rankings condition).
However, if the bias is due to overweighting the impact of
the targets motivation, the timing of the two exams is highly
relevant. Participants should see a rise in rankings as more
likely than a decline only when outcomes are (at least partly)
within the targets control. When the targets ability to affect
her performance is limited (e.g. when she doesnt have
enough time to learn from her mistakes on the rst exam),
participants should think of a rise and decline in rankings
as equally likely.
Results
As before, we computed a measure of change in ranking by
averaging participantsestimates of the likelihood that a stu-
dent would receive a grade that ranked her in a different third
of the class than where she ranked on the previous exam. In
the rise in rankings condition, we summed participantsesti-
mates that a student who previously ranked at the bottom 10
would rank in the middle 10 or higher. In the drop in
rankings condition, we summed their estimates that a student
who had previously ranked in the top 10 would rank in the
middle 10 or lower.
As expected, participants exhibited the upward mobility
bias only when the students were assumed to have consider-
able control over their performance on the second exam.
When the students could exert control over their performance
(the two exams were separated by a month, allowing moti-
vated students to practice the material), participants thought
that rising from the bottom 10 of the class to the middle 10
or higher was signicantly more likely (M = 60.53%,
SD = 19.67) than dropping from the top 10 to the middle
10 or lower (M = 34.89%, SD = 19.68), t(102) = 4.99,
p<.0001. However, when the studentsability to determine
their performance was limited (the two exams were on con-
secutive days, thus constraining the studentsability to study
the material), participants did not think that rising from the
bottom 10 was signicantly more likely (M = 41.65%,
SD = 19.99) than dropping from the top 10 (M = 35.17%,
SD = 21.15), t(102) = 1.05, ns. The interaction between con-
dition (high vs. low control) and direction of change (rise
or decline) was signicant, F(1,99) = 5.66, p= .019.
STUDY 5B: INFLUENCE OF PERCEIVED MOTIVA-
TION ON PREDICTIONS OF STUDENT
PERFORMANCE
In Study 5A we manipulated the amount of control targets
had over their performance and found that the upward mobil-
ity bias was evident only when participants thought the
targets could inuence their standing. In Study 5B, we
manipulated the targets motivation to control her standing.
We maintain that the upward mobility bias results from
overweighting the impact of the targets motivation. There-
fore, we predicted that participants would only exhibit the
bias when they thought the targets motivation to inuence
her standing was high.
498 Journal of Behavioral Decision Making
Copyright © 2015 John Wiley & Sons, Ltd. J. Behav. Dec. Making, 28, 491503 (2015)
DOI: 10.1002/bdm
Method
Participants
One hundred twenty-two Mechanical Turk participants (75
females, M
age
=33.48) completed the study in exchange for
modest monetary compensation.
Materials and procedure
As before, participants estimated the probability that a stu-
dents grade would differ from one exam to the next. How-
ever, we manipulated participantssense of the students
motivation by altering the stakes involved in the second
exam. In the high-stakes condition, participants were told
that the students were about to graduate and look for a
joband that they were convinced that their nal grades play
a large role in their employment prospects. In contrast, in the
low-stakes condition, participants were told that the students
had already sent out all their job applications and therefore
were convinced that their performance on the second exam
wouldntinuence their prospects. In the rise in rankings
condition, participants were then asked to assess the likeli-
hood that a randomly selected student who was in the bottom
10 of the class on the rst exam would rank in each third of
the distribution on the second exam. In the drop in rankings
condition, participants assessed the likelihood that a ran-
domly selected student who was in the top 10 on the rst
exam would rank in each third of the distribution on the sec-
ond exam.
Results
We summed participantsestimates of the likelihood that a
student would receive a grade that ranked her in a different
third of the class than where she ranked on the previous
exam. As expected, participants exhibited the upward mobil-
ity bias only when the students were assumed to be highly
motivated to succeed (Figure 3). When the stakes for the
second exam were high (i.e. the students believed their job
prospects depended on their nal grade), participants thought
that rising from the bottom 10 of the class to the middle 10 or
higher was signicantly more likely (M = 60.81%,
SD = 18.19) than dropping from the top 10 to the middle
10 or lower (M = 34.48%, SD = 21.24), t(121) = 4.85,
p<.0001. However, when the stakes were low (i.e. the stu-
dents believed their nal grade would not inuence their
job prospects), participants did not think that rising from
the bottom 10 was signicantly more likely (M = 49.80%,
SD = 25.82) than dropping from the top 10 (M = 41.36%,
SD = 23.43), t(121) = 1.42, ns. The interaction between con-
dition (high vs. low stakes) and direction of change (rise or
decline) was signicant, F(1,118) = 4.95, p= .028.
STUDY 6: IMPLICIT PERSONALITY THEORY AND
PREDICTIONS OF EMPLOYEE PERFORMANCE
Studies 5A and 5B provided support for an attribute substitu-
tion account of the upward mobility bias. Only when perfor-
mance was seen as within the focal individuals control, and
only when the individual was seen as highly motivated to ex-
ercise that control, did participants estimate that a rise in
rankings was more likely than a decline. Of course, people
vary in the degree to which they focus on (and attribute be-
havior to) motivation, variability that is likely to have impli-
cations for the strength of the upward mobility bias.
Researchers have documented considerable variability in
the extent to which people believe abilities are stable, and
consequently differ in the role they ascribe to effort in
inuencing performance (Dweck, 1999; Hong et al., 1999).
Whereas entity theorists tend to believe that personal attri-
butes are relatively xed (agreeing with such statements as
People can do things differently, but the important parts of
who they are cant really be changedand As much as I hate
to admit it, you cant teach an old dog new tricks), incre-
mental theorists are more likely to believe that ability is mal-
leable and subject to change due to effort (Dweck, 1999). We
therefore designed Study 5 to examine whether a persons
implicit personality theory or mindsetinuences the likeli-
hood that they would exhibit an upward mobility bias. More
specically, we examined whether the bias is more pro-
nounced for incremental theorists than for entity theorists.
We also wanted to further test the pervasiveness of the bias
by seeing whether it exists in yet another domainsales
performance.
Method
Participants
One hundred six Mechanical Turk participants (71 females, 1
unspecied, M
age
= 33.18) completed the study in exchange
for modest monetary compensation.
Materials and procedure
Participants read a scenario describing a retail store in which
30 salespeople were ranked at the end of each month accord-
ing to their sales, with the top 10 receiving a bonus. They
then estimated the likelihood that an employees ranking
would differ from one month to the next. In the rise in
rankings condition, participants estimated the likelihood that
an employee who was in the bottom 10 the previous month
would rank in each third of the distribution the next month.
Figure 3. Perceived likelihood of a rise/drop in class rankings when
the motivation is low (low-stakes condition) and high (high-stakes
condition) (Study 5B)
The Upward Mobility Bias 499S. Davidai and T. Gilovich
Copyright © 2015 John Wiley & Sons, Ltd. J. Behav. Dec. Making, 28, 491503 (2015)
DOI: 10.1002/bdm
In the drop in rankings condition, participants estimated the
likelihood that an employee who was previously ranked in
the top 10 would rank in each third of the distribution the
next month.
Following their predictions, participants completed a
measure of implicit personality theory (Dweck, 1999). They
indicated their agreement with eight statements (e.g. the
kind of person someone is, is something very basic about
them that cant be changed very much,everyone is a cer-
tain type of person, and there is not much that can be done
to really change that), presented in a different random order
for each participant, on a 6-point scale (1-strongly dis-
agree, 6-strongly agree). We reverse-scored the four entity
items and averaged them with the incremental items to
create an incrementalist composite (Cronbachsα= 0.88;
Levy et al., 1998).
Results
We summed participantsestimates of the likelihood that
an employees ranking would differ from one month
to the next. In the rise in rankings condition, we summed
the estimates that an employee who previously ranked in
the bottom 10 would rank in the middle 10 or higher.
In the drop in rankings condition, we summed partici-
pantsestimates that an employee who previously ranked
in the top 10 would subsequently rank in the middle 10 or
lower.
Participants exhibited an upward mobility bias, giving
higher estimates of the likelihood of a rise in rankings
than a decline. Whereas participants in the rise in
rankings condition thought that an employee who was
previously in the bottom 10 in sales had a 56%
(SD = 18.88) chance of rising to the middle 10 or higher,
participants in the drop in rankings condition thought that
an employee who was previously in the top 10 had only a
40% (SD = 15.84) chance of dropping to the middle 10 or
lower, t(104) = 4.74, p<.0001. Furthermore, whereas par-
ticipants believed that a randomly selected employee from
the bottom 10 had a 34% (SD = 11.37) chance of rising to
the middle 10, they thought that an employee from the
top 10 had only a 26% (SD = 8.25) chance of dropping
to the same third, t(104) = 4.15, p<.0001.
As predicted, the interaction between condition (rise
vs. drop in rankings) and participantspersonality mindset
was signicant, F(3,102) = 5.13, p= .006. Figure 4 pre-
sents the estimated predictions for any participant scoring
1 standard deviation above and below the mean on the
mindset measure (Aiken & West, 1991). As can be seen,
high incremental theorists were signicantly more likely
to exhibit the upward mobility bias than low incremental
theorists. High incremental theorists thought that a rise
in rankings is signicantly more likely than a drop in
rankings, t(102) = 5.39, p<.001, but no such tendency
was observed for low incremental theorists, t(102) = 1.39,
p= .17. It thus appears that the belief that performance
is malleable and subject to ones effort plays a role in
the upward mobility bias.
GENERAL DISCUSSION
Eight studies provided consistent evidence of an upward mo-
bility bias in predictions of performance in competitive set-
tings. Regardless of whether they were predicting athletic
(Studies 1 and 3A), academic (Studies 2, 3B, 4, 5A, and
5B), or vocational (Study 6) performance, participants
thought that a rise in rankings was signicantly more likely
than a decline. This bias is not the product of simple opti-
mism or wishful thinking. Rather, we obtained evidence that
the upward mobility bias is due to the emphasis given to a
targets motivation, and a relative disregard of contextual
factors inuencing the targets standing, such as the likely
performance of competitors. Only when an individual was
seen as willing and able to inuence her performance was a
rise in rankings judged to be more likely than a decline.
When an individuals ability to control her own perfor-
mance was limited or when she was seen as not suf-
ciently motivated to do so, a rise and fall in rankings
were judged equally likely. Study 6 provided evidence
that some people are predictably more prone to this bias
than others. Participants who consider ability to be mal-
leable rather than stable (Dweck, 1999), and who are
therefore more likely to emphasize motivation and effort
in their predictions, indicated that a rise in rankings is
more likely than a decline.
Although the upward mobility bias results from entirely
different psychological mechanisms, it bears some resem-
blance to the tendency for people to give more weight to
facilitory than inhibitory factors in judgment. Hansen and
Hall (1985), for example, found that people are more likely
to attribute the outcome of a competition to the strength of
the winner than the weakness of the loser. Similarly, peoples
causal attributions tend to be more inuenced by factors that
promote a given action (e.g. taking a class because the pro-
fessor is popular) than factors that inhibit it (e.g. it meets
early in the morning) (Newtson, 1974). With the respect to
the upward mobility bias, people give great weight to a par-
ticular facilitory cause on the part of low ranked targets (their
motivation to rise in the ranking) while giving little or no
weight to inhibitory causes acting on those who are already
at the top (regression, the decline in motivation that often ac-
companies success).
Figure 4. Perceived likelihood of a rise/drop in employee rankings
by participantspersonality mindset (entity/incremental) at +/
1SD (Study 6)
500 Journal of Behavioral Decision Making
Copyright © 2015 John Wiley & Sons, Ltd. J. Behav. Dec. Making, 28, 491503 (2015)
DOI: 10.1002/bdm
The upward mobility bias may contribute to the pervasive
tendency to root for the underdog. In competitive sports
(Frazier & Snyder, 1991), horse racing (Grifth, 1949), busi-
ness endeavors (Michniewicz & Vandello, 2013), political
campaigns (Goldschmied & Vandello, 2009), international
relations (Vandello et al. 2007) and artistic competitions
(Kim et al., 2008) people are often drawn to those who face
long odds. The underdog bias has been attributed to the ten-
dency to sympathize and identify with the struggle of com-
petitors with low a-priori chances of winning (Kim et al.,
2008) as well as to spectatorsmotivation to remain engaged
in the competition (Frazier & Snyder, 1991). But the upward
mobility bias may play a role as well. As we have shown,
people are relatively sanguine about the likelihood of success
for low-ranked individuals or teams. As long as the underdog
is seen as sufciently motivated to succeed, people believe
that a rise in their rank is quite possible. Indeed, people gen-
erally expect low-performing targets to take actions that
better their performance, and hence improve their future
standings (Lawrence & Makridakis, 1989; see also
Lawrence et al., 2006). Thus, an inated sense of the like-
lihood of future success may contribute to the appeal of
rooting for the underdog. Because peoples beliefs about
a targets motivation to improve is an important compo-
nent of the upward mobility bias, it should be the case
that the more salient an underdogsmotivationtosucceed,
the more likely it is that observers will overestimate its
success and therefore be drawn to it.
The psychological processes underlying the upward mo-
bility bias no doubt contribute to peoples asymmetrical be-
liefs about upward and downward economic mobility. In
previous work (Davidai & Gilovich, 2015), we found that
people believe that a person is more likely to move up the
economic ladder than down. Whereas people overestimated
the likelihood that an individual born to a family in the bot-
tom income quintile would rise to the middle quintile or
higher as an adult, they underestimated the likelihood of an
individual born to a family in the top income quintile
dropping to the middle quintile or lower. Consistent with
the results of the present studies, people appear to overweight
the impact of othersmotivation to better their life circum-
stances and underweight the strong situational factors that
work against their efforts to do so. The upward mobility bias
therefore may make it easier for people to accept consider-
able economic inequality because it leads to an inated sense
of how likely it is to better ones economic position.
CONCLUSION AND FUTURE RESEARCH
We have found that people believe that a rise in ranking is
more likely than a decline, a belief that appears to be robust
across a variety of domains. It remains to be seen whether it
generalizes across cultures as well. Because participants from
interdependent cultures pay more attention to contextual fac-
tors than participants from independent cultures (Masuda &
Nisbett, 2001; Miyamoto et al., 2006), the former may give
more weight to various impersonal causal inuences, such
as competitorsmotivation and ability, and therefore be less
likely to exhibit the upward mobility bias. At the same time,
we have shown that people who hold an incremental theory
of personality are more likely to fall prey to the bias. Given
that people from more interdependent cultures are more
likely to hold incremental beliefs regarding the malleability
of personal attributes (Norenzayan et al., 2002), they may
be more likely to exhibit the upward mobility bias. We there-
fore remain agnostic about the generalizability of this bias
across cultures and urge further research to explore this
question.
To probe the generalizability of this bias, we had partici-
pants assess the likelihood that individuals or teams would
move up or down in rankings when they had rich, detailed in-
formation about the targets (Studies 1 and 2) and when they
did not (Studies 36). However, our results do not imply that
the upward mobility bias will trump all types of detailed in-
formation when predicting performance in competitive set-
tings. Knowing that a student was often absent from class
would probably affect predictions of her academic success
in both absolute and relative terms, leading her to be seen
as less likely to rise in rankings and more likely to drop.
Note, however, that a students frequent absence from lec-
tures should not in and of itself affect predictions of the stu-
dents relative performance and is diagnostic only to the
extent that it relays comparative information (e.g. the amount
the student was absent relative to her peers). However, as we
noted above, people often substitute absolute assessments for
relative ones (Klar & Giladi, 1997, Klar & Giladi, 1999;
Kruger & Burrus, 2004; Kruger et al., 2008; Moore, 2005;
Windschitl et al., 2003). Therefore, when given specic in-
formation about a students lack of motivation (or ability),
people may jump to an overly pessimistic conclusion that a
drop in ranking is more likely than a rise. Devoid of this in-
formation, participants in our studies seemed to believe that
competitors (e.g. students, NBA teams, employees, etc.) are
motivated to succeed and proceeded to overweight this
assumption in their predictions, resulting in an upward
mobility bias.
Motivation is likely to be seen as especially relevant when
there appears to be considerable room for improvement. Peo-
ple may therefore assign considerable weight to a lower-
ranked targets motivation because those ranked near the
bottom have plenty of room for improvement. People may
assign less weight to a higher-ranked targets motivation
because, as hard as such a target might try, there is relatively
little room to improve. Future research might therefore examine
whether the tendency to overweight a targets motivation that
we have documented in these studies is inversely related to a
targets ranking: the lower a target is ranked, the more weight
people assign to the motivation to improve.
It is also worth considering how the upward mobility bias
may be inuenced by the number of contestants in a compe-
tition. The difculty of predicting a targets relative perfor-
mance is likely to increase with the number of competitors.
As a result, the more competitors there are, the more likely
it is that the targets desire to move up or down in the rank-
ings will be substituted for the likelihood that the target will
do so. We would expect, then, that as the number of compet-
itors increases, so should the belief that a rise in ranking is
more likely than a decline.
The Upward Mobility Bias 501S. Davidai and T. Gilovich
Copyright © 2015 John Wiley & Sons, Ltd. J. Behav. Dec. Making, 28, 491503 (2015)
DOI: 10.1002/bdm
Might it be benecial to believe that a rise in rankings is
more likely than a decline? To the extent that the upward mo-
bility bias leads to inated estimates of future success, it may
increase engagement in competitions both as spectators and
competitors. However, to the extent that people overestimate
the probability of success of lower-ranking competitors, the
bias can lead to post-competition disappointment. It is possi-
ble, therefore, that people are especially likely to exhibit the
bias prior to the beginning of a competition, and gradually
correct their predictions as more information about the com-
petitorsrelative advantages and disadvantages is revealed.
Given that information about a favorite NBA team is highly
accessible and salient during pre-season, a fan may be more
likely to overestimate the teams success before the season
commences, when information about the strength of the
teams rivals is not as salient. However, as league play gets
underway, information about other teams and their relative
abilities and motivations may then command more attention,
thereby reducing the bias. As fans of many struggling NBA
franchises have been painfully shown, the prospect of rising
in the leagues rankings can seem very likely right up until
the tip-off on opening night.
REFERENCES
Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing
and interpreting interactions. Newbury Park, CA: Sage.
Cohen, C.E. (1981). Person categories and social perception: Test-
ing some boundaries of the processing effects of prior knowl-
edge. Journal of Personality and Social Psychology,40,
441452. DOI: 10.1037/0022-3514.40.3.441
Davidai, S., & Gilovich, T. (2015). Building a more mobile Amer-
icaOne income quintile at a time. Perspectives on Psycholog-
ical Science,10(1), 6071.
Dweck, C. S. (1999). Self-theories: Their role in motivation,
personality and development. Philadelphia: Taylor and
Francis/Psychology Press.
Epley, N., & Gilovich, T. (2001). Putting adjustment back in
the anchoring and adjustment heuristic: Differential proces-
sing of self-generated and experimenter-provided anchors.
Psychological Science,12(5), 391396. DOI: 10.1111/1467-
9280.00372
Epley, N., & Gilovich, T. (2006). The anchoring-and-adjustment
heuristic: Why the adjustments are insufcient. Psychological
Science,17(4), 311318. DOI: 10.1111/j.1467-9280.2006.
01704.x
Frazier, J. A., & Snyder, E. E. (1991). The underdog concept in
sport. Sociology of Sport Journal,8(4), 380388.
Gilovich, T. (1991). How we know what isnt so: The fallibility of
human reason in everyday life. New York, NY: Free Press.
Goldschmied, N., & Vandello, J. A. (2009). The advantage of dis-
advantage: Underdogs in the political arena. Basic and Applied
Social Psychology,31(1), 2431. DOI: 10.1080/0197353080
2659810
Grifth, R. M. (1949). Odds adjustments by American horse-race
bettors. The American Journal of Psychology,62 (2), 290294.
DOI: 10.2307/1418469
Hansen, R. D., & Hall, C. A. (1985). Discounting and augmenting
facilitative and inhibitory forces: The winner takes almost all.
Journal of Personality and Social Psychology,49(6),
14821493. DOI: 10.1037/0022-3514.49.6.1482
Heider, F. (1958). The psychology of interpersonal relations. New
York: Wiley.
Hong, Y. Y., Chiu, C. Y., Dweck, C. S., Lin, D. M. S., & Wan, W.
(1999). Implicit theories, attributions, and coping: A meaning
system approach. Journal of Personality and Social Psychology,
77(3), 588599. DOI: 10.1037/0022-3514.77.3.588
Kahneman, D., & Frederick, S. (2002). Representativeness
revisited: Attribute substitution in intuitive judgment. In T.
Gilovich, D. Grifn & D. Kahneman (Eds.), Heuristics and
biases: The psychology of intuitive judgment (pp. 4981). Cam-
bridge: Cambridge University Press.
Kim, J., Allison, S. T., Eylon, D., Goethals, G. R., Markus, M. J.,
Hindle, S. M., & McGuire, H. A. (2008). Rooting for (and then
abandoning) the underdog. Journal of Applied Social Psycho-
logy,38(10), 25502573. DOI: 10.1111/j.1559-1816.2008.
00403.x
Klar, Y., & Giladi, E. E. (1997). No one in my group can be below
the groups average: A robust positivity bias in favor of
anonymous peers. Journal of Personality and Social Psycho-
logy,73(5), 885901. DOI: 10.1037/0022-3514.73.5.885
Klar, Y., & Giladi, E. E. (1999). Are most people happier than their
peers, or are they just happy?. Personality and Social Psychology
Bulleting,25(5), 586595.
Kruger, J. (1999). Lake Wobegon be gone! The below-average ef-
fectand the egocentric nature of comparative ability judgments.
Journal of Personality and Social Psychology,77, 221232.
DOI: 10.1037/0022-3514.77.2.221
Kruger, J. & Burrus, J. (2004). Egocentrism and focalism in unreal-
istic optimism (and pessimism). Journal of Experimental Social
Psychology,40, 332340. DOI: 10.1016/j.jesp.2003.06.002
Kruger, J., Windschitl, P. D., Burrus, J., Fessel, F., & Chambers, J.
R. (2008). The rational side of egocentrism in social compari-
sons. Journal of Experimental Social Psychology,44(2),
220232. DOI: 10.1016/j.jesp.2007.04.001
Lawrence, M., Goodwin, P., OConnor, M., & Önkal, D. (2006).
Judgmental forecasting: A review of progress over the last
25 years. International Journal of Forecasting,22(3),
493518. DOI: 10.1016/j.ijforecast.2006.03.007
Lawrence, M., & Makridakis, S. (1989). Factors affecting judgmen-
tal forecasts and condence intervals. Organizational Behavior
and Human Decision Processes,43(2), 172187. DOI:
10.1016/0749-5978(89)90049-6
Levy, S. R., Stroessner, S. J., & Dweck, C. S. (1998). Stereotype
formation and endorsement: The role of implicit theories. Jour-
nal of Personality and Social Psychology,74(6), 14211436.
DOI: 10.1037/0022-3514.74.6.1421
Lord, C., Ross, L., & Lepper, M. (1979). Biased assimilation and at-
titude polarization: The effects of prior theories on subsequently
considered evidence. Journal of Personality and Social Psycho-
logy,37(11), 20982109. DOI: 10.1037/0022-3514.37.11.2098
Masuda, T., & Nisbett, R. E. (2001). Attending holistically versus
analytically: Comparing the context sensitivity of Japanese
and Americans. Journal of Personality and Social Psychology,
81(5), 922934. DOI: 10.1037/0022-3514.81.5.922
Michniewicz, K. S., & Vandello, J. A. (2013). The attractive under-
dog: When disadvantage bolsters attractiveness. Journal of So-
cial and Personal Relationships,30(7), 942952. DOI:
10.1177/0265407513477629
Miyamoto, Y., Nisbett, R. E., & Masuda, T. (2006). Culture and the
physical environment holistic versus analytic perceptual
affordances. Psychological Science,17(2), 113119. DOI:
10.1111/j.1467-9280.2006.01673.x
Moore, D. A. (2005). Myopic biases in strategic social prediction:
Why deadlines put everyone under more pressure than everyone
else. Personality and Social Psychology Bulletin,31(5),
668679. DOI: 10.1177/0146167204271569
Moore, D. A., & Kim, T. G. (2003). Myopic social prediction and the
solo comparison effect. Journal of Personality and Social Psychol-
ogy,85(6), 11211135. DOI: 10.1037/0022-3514.85.6.1121
Newtson, D. (1974). Dispositional inference from effect of actions:
Effects chosen and effects foregone. Journal of Experimental
Social Psychology,10, 489496.
502 Journal of Behavioral Decision Making
Copyright © 2015 John Wiley & Sons, Ltd. J. Behav. Dec. Making, 28, 491503 (2015)
DOI: 10.1002/bdm
Norenzayan, A., Choi, I., & Nisbett, R. E. (2002). Cultural simila-
rities and differences in social inference: Evidence from behav-
ioral predictions and lay theories of behavior. Personality and
Social Psychology Bulletin,28(1), 109120. DOI: 10.1177/
0146167202281010
Radzevick, J. R., & Moore, D. A. (2008). Myopic biases in
competitions. Organizational Behavior and Human Decision
Processes,107(2), 206218. DOI: 10.1016/j.obhdp.2008.02.010
Tversky, A., & Kahneman, D. (1974). Judgment under uncertainty:
Heuristics and biases. Science,185, 11241131. DOI: 10.1126/
science.185.4157.1124
Vandello, J. A., Goldschmied, N. P., & Richards, D. A. (2007). The
appeal of the underdog. Personality and Social Psychology Bul-
letin,33(12), 16031616. DOI: 10.1177/0146167207307488
Windschitl, P. D., Kruger, J., & Simms, E. N. (2003). The inuence
of egocentrism and focalism on peoples optimism in competi-
tions: When what affects us equally affects me more. Journal
of Personality and Social Psychology,85(3), 389408. DOI:
10.1037/0022-3514.85.3.389
Authorsbiographies:
Shai Davidai is a PhD candidate in the Department of Psychology
at Cornell University. His research examines judgment and decision
making in everyday behavior.
Thomas Gilovich is the Irene Blecker Rosenfeld Professor of Psy-
chology at Cornell University. He received his PhD in Psychology
from Stanford University in 1981. His research has examined every-
day judgment and decision making, behavioral economics, and
egocentrism.
Authorsaddresses:
Shai Davidai, Department of Psychology, Cornell University,
Ithaca, NY, USA.
Shai Davidai, Department of Psychology, Cornell University,
Ithaca, NY, USA.
The Upward Mobility Bias 503S. Davidai and T. Gilovich
Copyright © 2015 John Wiley & Sons, Ltd. J. Behav. Dec. Making, 28, 491503 (2015)
DOI: 10.1002/bdm
... The salience of shared disadvantages is further enhanced by people's sensitivity to cues about rank and hierarchy (Davidai & Gilovich, 2015;Gurin & Townsend, 1986). Thus, when leaders use underdog narratives, they leverage people's natural sensitivities to information about disadvantages and create a shared awareness of an organization's unfavorable situation. ...
Article
Full-text available
Underdog stories are ubiquitous––the disadvantaged and outmatched protagonist overcoming the odds. Leaders across industries, from telecom to sports, employ these narratives to inspire members of their organizations. However, little is understood about how underdog narratives influence the actions and attitudes of members of organizations. To address this gap, we explain how a leader’s communication of an underdog narrative may instill confidence in members of the organization that together they can overcome their shared disadvantage and achieve a clear set of organizational aspirations. In doing so, we introduce a conceptual model that draws on the sensemaking and social identity literatures to explain how a leader’s underdog narrative fosters the adoption of a collective underdog identity by members of an organization. Further, we explain that the specific attributes of the leader’s underdog narrative influence how this unique type of collective identity leads to varied outcomes for members of an organization. As such, this paper aims to contribute to the understanding and utilization of a prevalent but under-examined organizational phenomenon.
... For instance, whether people perform well or poorly in a competition is determined by their inherent ability or skill, by the amount of effort they devote to the competition, by their opponents' abilities and skills, by the amount of effort devoted by each of their opponent, by external factors that advance or hinder their and their opponents' performance, and so forth. People typically focus on only a subset of such factors when thinking about their and others' performance ( Davidai and Gilovich, 2015 ). Yet, the myriad of elements that influence performance provide people with sufficient flexibility to feel as if their relative inferiority is due to factors outside their control rather than personal inadequacy. ...
Article
Full-text available
Inaccurate beliefs about procedural fairness often motivate people to act in self-serving and selfish manners. We investigate whether information about a level playing field might mitigate such behaviors. In a pre-registered behavioral experiment (n = 444), using a competitive and real-effort task, we manipulate whether participants are informed about the fairness of a competition or not. Following the competition, participants (who either won or lost the competition) decided how to distribute earnings between themselves and their opponent. We show that informing participants about the fairness of the competition reduces selfish behavior among losers, while behavior among winners remains unaffected. Moreover, we show that losers who were not informed about the fairness of the competition incorrectly viewed it as having been unfairly stacked against them (i.e., believing that they encountered significantly more difficult tasks than their opponents). Our findings suggest that information about a level playing field reduces selfish behavior and is important for understanding when and why motivated reasoning about procedural fairness helps people uphold a positive self-image.
... This upward-trajectory theme is common in U.S. culture, as evident in presidential state-of-the-union speeches, media focused on economic success (e.g., magazines, blogs, podcasts, etc), and "rags to riches" literature. This upward focus also aligns with the fact that people tend to have positive beliefs about their future selves (Markus & Nurius, 1986), make positive relative judgments in other performance domains (Davidai & Gilovich, 2015a), and tend to plan and work toward their generally upward future goals (Snyder, 2002). ...
Chapter
Full-text available
People commonly hold beliefs about social mobility, that is, perceptions of the likelihood to move up or down in society. What impact, if any, do these beliefs have on inequality, the societal status quo, and people’s own lives? How accurate are these views of where people end up in life and why? How can we characterize these beliefs? Through a review of existing research, we clarify the nature of social mobility beliefs and conceptualize how these beliefs affect society and individuals. For society, a consistent pattern emerges. Belief in social mobility leads to defense of the status quo and tolerance for inequality, but as we conceptualize, this occurs for abstract rather than concrete attitudes. On the personal level, social mobility beliefs relate to well-being and can affect the attainment of status-related goals, but this depends on individual difference factors, as well as the outcome being considered. This chapter also outlines opportunities for further research.
... Once underdog support in the United States was demonstrated consistently, further research attempted to delineate the motivations behind the effect. Ideas such as the upward mobility bias (Davidai and Gilovich, 2015) demonstrate the considerable focus placed on the underdog's motivation to rise up while the intentions of the favorite to suppress this effort are often overlooked. Alternatively, Goldschmied and Vandello (2012) argued for a memory bias rooted in the availability heuristic, such that when having to make a decision whether to support an underdog entity or not, past instances of underdog struggles are recalled. ...
Article
Full-text available
Jesus told his disciples, “Truly I tell you, it is hard for someone who is rich to enter the kingdom of heaven. Again I tell you, it is easier for a camel to go through the eye of a needle than for someone who is rich to enter the kingdom of God.” (Matthew 19:23-24). Ditto for heroes. The current study suggests that ‘humble beginnings’ is also a prerequisite for one to become an adulated entity. Participants from China, Israel, and Japan read of two sports teams with disparate expectations and/or financial resources about to face each other. Support was extended to the lesser one. When the two domains of comparison were contrasted, participants wished the lower resources/ high expectations team to win the game. This finding was interpreted as an impetus to maintain basic fairness based on competency assessments, both fundamental and universal psychological needs, at the root of the choice to support underdogs. In conclusion, we explore how support underdog relates generally to the concept of heroism.
Article
How do people forecast an actor’s future rank after observing a rank change and what are the factors that shape these forecasts? In this research, we shed new light on the attributions that people make when they observe an actor change rank and on how these attributions explain where people expect the actor to rank in the future. Specifically, in Studies 1a and 1b we document an asymmetric extrapolation bias, whereby people extrapolate upward rank trajectories more steeply into the future than downward trajectories – a pattern of results that differ in both magnitude and direction from actual rank change patterns over time. In Studies 2 and 3 we provide evidence of the different attributions that explain people’s asymmetric extrapolation through measurement and manipulation. Finally, in Study 4 we demonstrate a practical downstream consequence of this asymmetric extrapolation bias (i.e., promotion recommendation) (Study 4). Theoretical and practical implications are discussed.
Article
Objective. To analyze and summarize the results obtained in various social sciences regarding the emergence and spread of prejudice. Background. Prejudice is a complex issue, which is represented in different spheres of social science. In psychology, researchers analyze individual or group-based factors, while in economics and sociology they investigate objective indicators of the socioeconomic development at the country-level. Today the data in the mentioned fields are not integrated. Conclusions. Previous studies showed that GDP might indicate a significant factor in reducing bias, but only in countries with low social inequality or stable economic growth. Moreover, people with lower socioeconomic status are more vulnerable to the consequences of economic and socio-political changes. Psychological factors, in particularly, the perceived inequality or intergroup threat defined additional perspective in the explanation of the relationship between objective socio-economic indicators and the level of prejudice. In addition, the ideological attitudes (e.g., system justification) also influence the perceived inequality or threat, and bias. These findings are used to define future directions for research related to integration of objective socio-economic indicators and psychological variables to analyze the nature of prejudices.
Article
The prospect of socioeconomic mobility has long been a prominent concept in American life; however, research has shown that economic inequality can weaken Americans’ beliefs in the prospect of upward mobility for poorer individuals in their country. In one correlational study, two experimental studies, and an internal meta‐analysis, we extend this work by demonstrating that Americans’ perceptions of inequality can also influence their beliefs about another form of mobility that is necessary in a truly mobile society: downward mobility for richer individuals. Specifically, we found that the more that Americans’ perceptions of inequality in their country shifted toward extreme levels—toward the view that a small minority of the population holds a much greater proportion of the state’s wealth than all other groups combined—the more likely they were to believe that both upward and downward mobility were unlikely. That is, they believed that social class groups in their country were largely ossified and impermeable, and thus that Americans were unlikely to move out of the groups they were born into. We discuss the potential implications of these findings for important motivational and behavioural outcomes.
Article
Although economic mobility is an objectively defined term, lay beliefs about mobility—the configuration of ideas and attitudes about economic mobility that is not necessarily grounded in economic data—are often formed in a subjective manner. Drawing on research from the United States and beyond, we propose a novel framework for understanding how people construe, think about, and understand economic mobility. We highlight the importance of systematically examining the type, time frame, trajectory, and target of mobility that people have in mind for understanding when they most and least strongly believe in it. In addition, our framework offers a conceptual roadmap for examining the factors that influence lay beliefs about mobility, including individual differences in these lay beliefs and their important downstream consequences. Finally, we outline several important open questions that are highlighted by our framework as a guide for future research on lay beliefs about economic mobility.
Article
I examine how and why current-period performance shapes investors' evaluations of future-oriented disclosures. Three experiments provide evidence that a firm's current-period performance shapes investors' beliefs about the appropriateness of managerial optimism, which, in turn, affects investors' evaluation of firms that focus on either challenges or opportunities in future-oriented disclosures. When a firm is performing poorly, investors believe that managers can best achieve success by being optimistic about the future and, therefore, invest more when the firm focuses on opportunities rather than challenges in future-oriented disclosures. When a firm is performing well, on the other hand, investors believe that managers can best achieve success by being realistic about the future and, therefore, invest more when the firm focuses on challenges rather than opportunities.
Article
Full-text available
Most people judge themselves to be content with their lives. However, they also judge themselves to be more content than the others in their group, which is a logical impossibility. In line with previous speculations, the authors found in two studies that comparative contentment judgments were highly related to judgments of one’s own contentment but entirely unrelated to judgments of comparison of others’ contentment. That is, comparative contentment judgments are predominantly self-focused. Researchers asking the question, “How content are you relative to your peers?” should be aware that the response might well be to the question “How content are you?”
Article
Full-text available
Political candidates are often labeled as underdogs, either by the press or themselves. This paper explores connotations associated with the underdog label in the political arena. We argue that being labeled an underdog has a strategic advantage because it is associated with positive qualities, particularly likeability. The current studies demonstrate that partisan supporters prefer to see their favored candidate as an underdog compared to people not aligned with the candidate, and underdog-labeled entities are perceived to be warmer, and no less competent, than frontrunners. Discussion focuses on the advantages and risks of carrying the underdog label in the political arena.
Article
The tension and excitement of competitive sport is created by the indeterminacy of the contest that is based on an approximate equity between the contestants. Yet players and teams vary in competence and prestige, and those with less competence are frequently labeled as the underdog. While winning is valued, cross-cutting values often create sentiments for the underdog, that is, the desire for the underdog to overcome the inferior status and upset the favored opponent. Social support for the underdog reflects a utilitarian perspective that helps maintain an emotional interest in a contest; additionally, underdogs receive support from the social value of equity. At a microlevel, the underdog status is often used to increase the level of motivation and performance. Data gathered from university students are used to support the positions taken in the paper.
Article
While intuition suggests and much research has shown that people are attracted to advantaged individuals, the present study explored the conditions under which people might be attracted to disadvantaged individuals. We hypothesized that perceiving someone as unfairly disadvantaged can motivate attributions of positive personal characteristics and, consequently, judgments of heightened attractiveness. Seventy-eight participants were randomly assigned to read about a job applicant facing a fair or an unfair application process resulting in a competitive advantage or disadvantage. In support of our hypothesis, participants judged unfairly disadvantaged applicants (i.e., underdogs) and fairly advantaged applicants as more physically attractive and suitable as dates compared with fairly disadvantaged and unfairly advantaged applicants. These results highlight the role of situational factors in judgments of one's attractiveness.
Article
Many decisions are based on beliefs concerning the likelihood of uncertain events such as the outcome of an election, the guilt of a defendant, or the future value of the dollar. Occasionally, beliefs concerning uncertain events are expressed in numerical form as odds or subjective probabilities. In general, the heuristics are quite useful, but sometimes they lead to severe and systematic errors. The subjective assessment of probability resembles the subjective assessment of physical quantities such as distance or size. These judgments are all based on data of limited validity, which are processed according to heuristic rules. However, the reliance on this rule leads to systematic errors in the estimation of distance. This chapter describes three heuristics that are employed in making judgments under uncertainty. The first is representativeness, which is usually employed when people are asked to judge the probability that an object or event belongs to a class or event. The second is the availability of instances or scenarios, which is often employed when people are asked to assess the frequency of a class or the plausibility of a particular development, and the third is adjustment from an anchor, which is usually employed in numerical prediction when a relevant value is available.
Article
A core tenet of the American ethos is that there is considerable economic mobility. Americans seem willing to accept vast financial inequalities as long as they believe that everyone has the opportunity to succeed. We examined whether people's beliefs about the amount of economic mobility in the contemporary United States conform to reality. We found that (a) people believe there is more upward mobility than downward mobility; (b) people overestimate the amount of upward mobility and underestimate the amount of downward mobility; (c) poorer individuals believe there is more mobility than richer individuals; and (d) political affiliation influences perceptions of economic mobility, with conservatives believing that the economic system is more dynamic-with more people moving both up and down the income distribution-than liberals do. We discuss how these findings can shed light on the intensity and nature of political debate in the United States on economic inequality and opportunity. © The Author(s) 2014.