673192

research-article2017
PPSXXX10.1177/1745691616673192Kraus et al.Signs of Social Class

Perspectives on Psychological Science

Signs of Social Class: The Experience of 2017, Vol. 12(3) 422­–435

© The Author(s) 2017

Economic Inequality in Everyday Life https://doi.org/10.1177/1745691616673192

Reprints and permissions:
sagepub.com/journalsPermissions.nav
DOI: 10.1177/1745691616673192
www.psychologicalscience.org/PPS

Michael W. Kraus1, Jun Won Park1, and Jacinth J. X. Tan2

1
Yale University and 2University of California, San Francisco

Abstract
By some accounts, global economic inequality is at its highest point on record. The pernicious effects of this broad
societal trend are striking: Rising inequality is linked to poorer health and well-being across countries, continents,
and cultures. The economic and psychological forces that perpetuate inequality continue to be studied, and in this
theoretical review, we examine the role of daily experiences of economic inequality—the communication of social class
signals between interaction partners—in this process. We theorize that social class signals activate social comparison
processes that strengthen group boundaries between the haves and have nots in society. In particular, we argue
that class signals are a frequent, rapid, and accurate component of person perception, and we provide new data
and analyses demonstrating the accuracy of class signaling in 60-s interactions, Facebook photographs, and isolated
recordings of brief speech. We suggest that barriers to the reduction of economic inequality in society arise directly
from this class signaling process through the augmentation of class boundaries and the elicitation of beliefs and
behaviors that favor the economic status quo.

Keywords
social class, socioeconomic status, economic inequality, person perception, intergroup relations

Broad global trends suggest that economic inequality, at Integrating insights from research on social comparison
its historic peak according to some metrics (Organisation processes (Fiske, Cuddy, Glick, & Xu, 2002; Taylor &
for Economic Co-operation and Development, 2014; Lobel, 1989) and on status signaling (Ambady & Rosenthal,
Piketty & Saez, 2014), is a pressing societal problem 1993; Sapolsky, 2005), we suggest that economic inequal-
impacting the health and well-being of individuals: In ity is experienced daily as the communication of social
roughly 70% of studies examining the health impacts of class signals—behaviors that provide information about a
economic inequality, data indicate that societal health person’s income, educational attainment, or occupation
worsens as economic inequality intensifies (Wilkinson & status (Bourdieu, 1984; Kraus & Keltner, 2009; Veblen,
Pickett, 2006, 2009). When economic inequality deepens 1899/1973)—perceived and expressed in everyday inter-
on a societal scale, the evidence suggests that it is both actions. In this article, we consolidate findings from the
societies and individuals that suffer. psychological and economic sciences to derive three
In addition to its apparent relationship to well-being, basic predictions about social class signaling and the
the way in which economic inequality shapes the psy- experience of inequality. Specifically, we theorize that
chological experiences of individuals is a topic of grow- class signals (a) occur frequently, rapidly, and accurately
ing interest in the social and economic sciences, where in the social perception process; (b) augment group
research examines how people explain or justify inequal- boundaries between the haves and have nots in society;
ity (Kraus, Piff, & Keltner, 2009; Shariff, Wiwad, & Aknin, and (c) elicit psychological processes and behaviors that
2016), whether people are aware of its magnitude justify and maintain the current economic system (see
(Davidai & Gilovich, 2015; Norton & Ariely, 2011), and if
inequality influences behavior (DeCelles & Norton, 2016).
Corresponding Author:
In this article, we take a unique approach in that we Michael Kraus, Yale University - School of Management, 165 Whitney
examine economic inequality as it is experienced by Ave., New Haven, CT 06511
individuals in their everyday interactions with others. E-mail: michael.kraus@yale.edu
Signs of Social Class 423

ECONOMIC
SIGNS OF SOCIAL GROUP INEQUALITY
CLASS BOUNDARIES MAINTENANCE

Body
(Kinesic behavior, Sorting Dehumanization
physical appearance)

Voice
(Linguistic and
paralinguistic cues, Stereotyping Strategic sharing
word choice)

Culture
(Sartorial choices, leisure Class conflict Justifying ideology
activities, preferences)

Fig. 1.  A theoretical model detailing the modes of social class signaling and their influence on the augmentation of group boundaries
between the haves and have nots in society and, in turn, on social judgments and behaviors that perpetuate economic inequality on a
societal scale.

Fig. 1). Throughout, we detail existing evidence for each We believe our approach in this article represents a
of these predictions and chart the future directions they significant advance in how researchers examine eco-
suggest. nomic inequality. Prior studies have examined economic
For the purpose of this article, it is important to dwell inequality through the lens of a global economy (e.g.,
on the definition of social class, which we and others Piketty, 2015); as an economic trend perceived by indi-
have defined in the past as one’s position in the eco- viduals (e.g., Kluegel & Smith, 1986); or indirectly through
nomic hierarchy in society that arises from a combination an examination of the association between social class
of annual income, educational attainment, and occupa- position and emotion, cognition, or behavior (e.g., Kohn,
tion prestige (Adler et al., 1994; Oakes & Rossi, 2003). 1963; Kraus, Tan, & Tannenbaum, 2013). Moving forward
Though the experience of social class is shaped by this from this prior work, we examine economic inequality as
economic positioning, the actual impact of the construct the daily process of comparing one’s own socioeconomic
on social and psychological experience is wide-ranging standing to that of others based on the dynamic observa-
and multifaceted: Social class shapes behavior through tion of observable behaviors that signal social class. Such
cultural learning, such as socialization processes occur- an approach has several direct benefits: This approach
ring within a family whose members share a similar expands our understanding of the experience of eco-
socioeconomic background, and through social-cognitive nomic inequality beyond the economic realm to include
mechanisms, which include habitual response patterns to the dynamic social signaling processes it informs, it
an experience that is unique to a particular social class allows for a fuller understanding of the process by which
(e.g., Fiske & Markus, 2012). Social class can also deter- people come to accurately perceive their own social class
mine the groups an individual interacts with and belongs position in society relative to others (e.g., Goodman
to (Lareau & Conley, 2008). Because of its many facets, it et al., 2001; Hout, 2008), and it provides a framework for
is helpful to conceptualize social class (like race) as a understanding the processes by which economic inequal-
“bundle of sticks” that can be disaggregated and studied ity might perpetuate itself over time.
based on its specific elements (i.e., Sen & Wasow, 2016).
Signaling is the specific element of social class we exam-
ine in this article, with an eye toward better understand-
Signs of Social Class
ing how class signals might shape processes related to From primates to fiddler crabs, individuals across species
the creation of class boundaries, perceptions and experi- signal their broadly defined social status with nonverbal
ences of class mobility, and self-perceptions of an indi- behaviors (Hall, Coats, & LeBeau, 2005; Sapolsky, 2005;
vidual’s own socioeconomic position in society. Sergio et al., 2011), vocalizations (Dunbar & Burgoon,
424 Kraus et al.

Table 1.  Summary of Prior Studies Investigating Accuracy in Perceptions of Social Class From Observations
of Cultural Objects, Physical Appearance, Behavior, and Speech

Domains of signaling
Study reference Stimuli for observer judgments accuracy
Davis (1956) Living room photographs Cultural objects
Giles and Sassoon (1983) UK college students mimicking Speech
Cockney or standard English accents
Schmid Mast and Hall (2004) U.S. University employee Physical appearance
photographs Cultural objects
Kraus and Keltner (2009) 60 s interactions between USA Physical appearance
University students Cultural objects
Thin slices of behavior
Gillath, Bahns, Ge, and Crandall (2012) Shoe photographs Cultural objects
Becker, Kraus, and Rheinschmidt-Same U.S. Facebook profile photographs Physical appearance
(2017) Cultural objects
Bjornsdottir and Rule (2016) Standard facial photographs Physical appearance
Kraus, Park, and Tan (this manuscript) Seven spoken words Speech

2005; Gregory & Webster, 1996; Ko, Sadler, & Galinsky, a high social comparison orientation—was positively
2015), and social symbols (Veblen, 1899/1973). These associated with reports of relative deprivation or feelings
status symbols assist individuals in avoiding costly aggres- of reduced resources relative to others (Buunk, Zurriaga,
sive encounters or in aligning with others who have the Gonzalez-Roma, & Subirats, 2003).
material resources necessary to facilitate group living Second, studies of social class signaling indicate that
(e.g., Ambady & Rosenthal, 1993; Zahavi & Zahavi, 1999). perceivers can rapidly and accurately judge the social
In humans, symbols of status are used in a variety of class of others based on only small amounts of informa-
contexts—for instance, to signal physical dominance tion. Theoretical accounts of social class suggest that it
(Kraus & Chen, 2013) or as cues of achieved success shapes people’s lives in persistent and enduring ways by
(Tracy & Matsumoto, 2008). Aligning directly with this constraining or expanding access via levels of economic
past research, we assert that social class is a form of soci- resources or by shaping behavior through cultural and
etal social status that is signaled and perceived accurately psychological means (Bourdieu, 1984; Fiske & Markus,
in interactions with others (e.g., Kraus & Keltner, 2009). 2012; Kraus, Piff, Mendoza-Denton, Rheinschmidt, &
Our first prediction that social class signaling occurs Keltner, 2012). Studies generally support this assertion:
during social perception is supported by two separate Class determines the foods people eat (Monisaivis &
lines of research. First, research on social comparison, or Drewnowski, 2009), the music and art they enjoy (DiMaggio
the process of comparing oneself to other people on & Useem, 1978; Peterson & Kern, 1996; Snibbe & Markus,
social characteristics and outcomes (Festinger, 1954; Taylor 2005; Van Eijck, 2001), the leisure activities people
& Lobel, 1989), indicates that people compare themselves engage in (Veblen, 1899/1973), the linguistic patterns
to others frequently on economic dimensions. Social they employ (Labov, 1964, 1990, 2006), and the clothing
comparison is often referred to as an unavoidable aspect they wear (Gillath, Bahns, Ge, & Crandall, 2012). This
of perception during social interactions (Brickman & Bul- analysis suggests that when people engage in social
man, 1977; Brown, Ferris, Heller, & Keeping, 2007), and interactions, some of their behaviors and cultural prac-
it functions as a means by which individuals gather infor- tices are infused with social class and, as a result, accu-
mation about the self, regulate their emotions and goals, rately communicate social class position to observers.
manage uncertainty, and judge the normalcy of personal Although several studies indicate class differences in
life events (e.g., Taylor & Lobel, 1989). Though compari- behavior (e.g., Bernstein, 1962; Lizardo & Skiles, 2015)
sons occur on many dimensions in social relationships, and language use (e.g., Labov, 2006) and inform this
evidence indicates that these comparisons are frequently work a great deal, relatively few examine the social class
tuned to economic outcomes: For instance, reward signaling accuracy of these behaviors (see Table 1). For
regions of the brain (e.g., the left ventral striatum) were instance, early work by Davis (1956) found that photo-
activated when engaging in comparison with an experi- graphs of living rooms allowed observers to accurately
ment partner who received a relatively lower monetary judge the occupation status and education of individuals.
reward rather than an equal reward (Fliessbach et al., In the realm of appearance, researchers found that social
2007). In another study, the tendency for nurses to have class can be accurately perceived by strangers based on
Signs of Social Class 425

such static appearance cues as shoes worn by partici- 70

pants (Gillath et  al., 2012), employee photographs
60
(Schmid Mast & Hall, 2004), and standard facial images
(Bjornsdottir & Rule, 2016). Two additional studies con- 50

Percent Correct
ducted by an author of this manuscript also bear out this 40
signaling prediction: In the first, 50 university student
dyads (n = 100) were recruited to a laboratory setting 30
where they had a 5-min “get acquainted” interaction with 20
a stranger (Kraus & Keltner, 2009). In the second, a study
10
of cultural practices online, the 20 most recent Facebook
photographs were collected from a sample of 113 univer- 0
sity students and adults from an online crowdsourcing Correct Quartile Correct Half
platform (i.e., MTurk; Becker, Kraus, & Rheinschmidt- Chance Behavior Facebook
Same, 2017). In both studies, participants provided infor- Speech Across Modalities
mation about their own family social class by reporting
Fig. 2. The tendency for observers to guess the correct social class
their annual income, parental educational attainment,
quartile (left) and half (right) of participants based on observing 60 s of
and self-perceived position on a 10-rung ladder repre- behavior, 20 Facebook photographs, listening to seven isolated spoken
senting ascending levels of income, education, and occu- words, or across all signaling modalities. The light grey bars indicate
pation status (Adler, Epel, Castellazzo, & Ickovics, 2000). percentages expected if observers are guessing at rates indistinguish-
able from chance. In all cases, observers showed above chance accu-
A separate sample of observers then watched the first 60 s racy in perceiving social class.
of the “get acquainted” interaction or viewed the Face-
book photographs and estimated participant social class
on the same 10-rung society ladder. Across both studies, practices shared through photographs on Facebook are
an overall composite metric of social class computed by sufficient to accurately signal social class.
averaging the z-scored index of each individual social One of the most pervasive domains of person percep-
class item was significantly positively correlated with per- tion is the voice, and much research supports the utility of
ceptions of social class made by observers based only on aspects of speech—including word choice and linguistic
60 s of recorded behavior, r(98) = .28, p = .005, or on and paralinguistic vocal patterns—in signaling various
viewing 20 Facebook photographs, r(111) = .27, p = .005. forms of social status. For instance, people modulate their
Moreover these associations held after accounting for the vocal pitch to make it more consistent with that of high
race and gender of participants. status individuals (Gregory & Webster, 1996), and power-
For the purpose of this article, we reanalyzed the data ful people use lowered pitch along with other vocal cues
from these prior studies to determine how well observers to suggest their physical dominance and authority (Hall
performed in estimating social class above chance guess- et al., 2005; Ko et al., 2015). We predict that speech style,
ing. For this analysis, observer estimates were divided even the linguistic and paralinguistic content of speech, is
into four quartiles and two halves based on prior research of considerable utility for perceiving the social class of
indicating that the United States is divided into four social others (Bugental, Henker, & Whalen, 1976; Ellis, 1967;
class categories (i.e., lower, working, middle, and upper Labov, 2006; Wolfram & Schilling-Estes, 2006). Though
class), but that about 90% of people identify with the two speech style varies by region and cultural background, it
middle categories (Hout, 2008). The overall metric of par- is also determined by social class: For instance, studies
ticipant social class was also divided into quartiles and find that students from lower social class backgrounds are
halves. We then compared observer estimates to partici- more likely to speak using nonstandard dialects than rela-
pant overall social class. The observer estimates were tively upper class individuals (Fogel & Ehri, 2000). As
considered to be correct if the quartile or half chosen by well, discrimination occurs more often against individuals
observers matched that of participants, whereas mis- in the United States with nonstandard dialects (Gluszek &
matches were incorrect (e.g., Hertenstein, Holmes, Dovidio, 2010), and people from traditionally low status
McCullough, & Keltner, 2009). The results of this analysis groups in society sometimes code switch away from non-
are shown in Figure 2: When compared to chance guess- standard speech to more standard forms in order to blend
ing (i.e., 25% for quartiles, and 50% for halves), observers in while in high status contexts (Brannon & Markus, 2013;
performed significantly better than chance when estimat- Labov, 1972, 2006). At the other end of the social class
ing the correct social class quartile, x2(1)Behavior = 5.33, spectrum, the material resources and opportunities
p = .02; x2(1)Facebook = 5.45, p = .02, and half, x2(1)Behavior = afforded by high social class (Kraus et al., 2012; Oishi &
4.00, p = .046; x2(1)Facebook = 4.68, p = .03, of participants Talhelm, 2012) allow individuals to travel and to interact
using 60s of recorded behavior or Facebook photo- with others who are well traveled, thereby contributing to
graphs. Short dynamic bursts of behavior and cultural less locally defined and more standard patterns of speech.
426 Kraus et al.

In one example of the social stratification of linguistic Moreover, this association held after accounting for
cues, Labov (1972) found that New York City department speaker characteristics—race, gender, and vocal pitch—
store clerks used different pronunciations for the “r” in as well as geographic region characteristics—zip code
words like “floor” and “fourth” based on social class mea- level population density, median income, and high school
sured in terms of the prestige of each of the stores. Upon graduation rates, β = .14, t(192) = 2.12, p = .036. Percep-
being asked to repeat their pronunciations, all clerks tions of social class were also significantly positively
tended to emphasize the “r” sound in the words more associated with actual speaker social class in three of the
than they did in spontaneous speech—a behavior indi- four Census defined regions of the United States (all but
cating conscious awareness of the class signaling impli- in the Midwest), region effect F(3, 1632) = 46.78, p <
cations of their speech (Labov, 1972). In other direct .001).
evidence for signaling of social class through speech, We also examined how observer estimates of social
Giles and Sassoon (1983) asked participants to mimic a class performed relative to chance using the prior quar-
Cockney or standard English accent, finding that Cock- tile and half analysis. Splitting the sample for this analysis
ney accents were judged lower in social class than stan- appeared to be valid: For instance, the bottom half of
dard English ones. speakers had an average educational attainment of 1.83,
The above analysis suggests that speech style is an roughly equivalent to an education of high school gradu-
accurate signal of social class even when the semantic ation and some college, and an occupation code of
content of speech is held constant. In a new study, we 4209.54, similar to the occupation status of sales (4000)
sought to test the very limits of this assertion by exposing and service workers (5000). The top half of the speaker
the brief speech of individuals (i.e., seven words spoken sample had an average educational attainment of 3.38,
out of context) to a separate panel of observers with the consistent with obtaining a four-year college degree, and
prediction that observers would accurately judge social an occupation score of 1477.33, similar to the occupation
class based solely on speech style. status of professional athletes and journalists (1500). As
To test this hypothesis we used a sample of individual Figure 2 illustrates, observers showed above chance
speakers from across the United States recorded reading accuracy in judging the correct quartile, x2(1) = 15.38,
one of two literary passages by the International Dialects p < .001, and half, x2(1) = 12.21, p < .001, of speaker
of English Archive (http://dialectsarchive.com). Of the social class based only on their speaking style—evident
246 speakers collected by the archive across 48 of the 50 upon listening to just seven words spoken out of context.
states in the United States, 213 reported their educational Moreover, the overall chance accuracy estimate across
attainment and occupation status, and these speakers the modalities of social class signaling (see Fig. 2, right-
comprised our sample for the signaling analysis. We then most bars) was significant, x2(1)Quartile = 25.92, p < .001;
created isolated speech recordings for seven spoken x2(1)Half = 20.74, p < .001.
words—(i.e., “and,” “from,” “thought,” “beautiful,” “imag- Overall, these studies across diverse domains of
ine,” “yellow,” and “the”) chosen because they occurred speech and behavior provide some preliminary yet prom-
in both read passages—for each of the speakers and pre- ising evidence suggesting that social class is rapidly and
sented these recordings to a separate U.S. panel sample of accurately perceived in the early stages of social percep-
observers recruited as part of Qualtrics panels (n = 568). tion. Several future directions are suggested by this
Just as in the prior studies, our observer sample estimated research: For instance, does accuracy of social class per-
speaker social class using a 10-rung ladder representing ception improve linearly with the addition of more
ascending levels of social class in society (M = 5.85, SD = behavioral information about social class? Based on prior
0.68). We then compared averaged observer estimates of research, it is possible that diminishing returns in social
social class to a standard composite (M = −0.02, SD = perception accuracy occur for social class due to the
0.87) of speaker educational attainment, coded based on onset of confirmation biases (Nickerson, 1998). Likewise,
four ascending categories (1 = did not finish high school, though we demonstrate that the information passed on in
2 = high school graduation or some college, 3 = four-year the initial stages of impression formation is sufficient for
college graduation, 4 = advanced degree), and occupa- class signaling, data using more sensitive time measure-
tion status, coded using an established four digit occupa- ments might be able to determine precisely when accu-
tion status coding scheme (Ganzeboom, De Graaf, & racy in class signaling emerges in the social perception
Treiman, 1992). See the online Supplemental Material for process (e.g., at the level of seconds or milliseconds). It
more detail on the samples, stimuli, and analyses. is also interesting to speculate about the other modalities
The results support our prediction that speech style is of social perception where class signals are communi-
a social class signal. Observers provided judgments of cated. Given health disparities between high and low
speaker social class that were significantly positively cor- social class individuals (Adler et al., 1994), perhaps phys-
related with speaker social class, r(211) = .22, p = .002. ical features that reveal health are also a reliable indicator
Signs of Social Class 427

of social class as preliminary work suggests (Bjornsdottir people from differing social class backgrounds develop
& Rule, 2016). unique norms and cultural models for behavior and
Social class signals are also the result of two processes— self-expression discussed earlier (Stephens, Markus, &
perceiver judgments and target behaviors—and thus far Fryberg, 2012; Weininger & Lareau, 2009).
studies have only started to examine the specific contribu- In addition to the historical structural barriers that con-
tion of each process to accurate class signaling. Experi- tribute to the sorting of social class in society (Desmond,
ments that systematically constrain target behavior, by for 2016), we contend that social class signals expressed and
instance, carefully controlling the context and conditions perceived in interactions with others facilitate this sorting
for speech (Labov, 1972), can better tease apart both the process. Specifically, we assert that social class signals are
specific behaviors that elicit class signals and the minimal causal in the maintenance of structural social class
conditions necessary for perceivers to judge social class. In boundaries because they reduce class mobility and limit
this vein, signal detection theory methods are needed to cross-class contact. Several lines of evidence are sugges-
determine how sensitive social class perceptions are to tive of this prediction: The notion that similarity enhances
variations in class signals (McNicol, 2005; Swets, 2014). liking and relationship formation is a core principle in
Lastly, one of the central claims of this analysis is that peo- relationship science (Bergeron & Zanna, 1973; Montoya,
ple experience economic inequality principally through Horton, & Kirchner, 2008). When individuals accurately
this process of sending and perceiving social class signals signal and perceive social class in interactions with oth-
in daily interactions and comparatively less in terms of ers, signals that communicate differences in social class
macroeconomic societal shifts in wealth and markets. One are likely to create barriers for relationship formation
implication of this logic is that in countries or cultures across class boundaries. In this fashion, class signals are
where societal economic inequality is lower, signals of particularly strong in determining the likelihood of rela-
social class might actually be harder to discern than in tionship formation and in constraining access to net-
countries like the United States where inequality is height- works of influence and opportunity and to individuals of
ened (e.g., Côté, House, & Willer, 2015). similar (higher) social class upbringing.
The first work directly assessing this prediction was
Social Class Signals Augment conducted by Bergeron and Zanna (1973). In their study,
Peruvian university students read opinion pieces that
Group Boundaries were manipulated to be written by someone who shared
Having detailed evidence suggesting that social class is their social class (i.e., attended their same university, an
perceived early on in the social perception process, we elite social club, or a neighboring university) or who did
now turn to the potential consequences of the accuracy not (i.e., a member of the Peruvian indigenous popula-
of social class perceptions for groups in society. We make tion not attending college). Participants reported desiring
three specific predictions for how class signals augment to form a relationship with opinion writers who shared
group boundaries—by facilitating the sorting of individu- their social class far more than they did the relatively
als into social class groups, activating social class stereo- lower class indigenous writers not attending college, and
types, and enhancing conflict between the haves and this effect was independent of participants’ level of agree-
have nots in society. ment with the arguments in the opinion piece (Bergeron
Signals of social class create group boundaries by sort- & Zanna, 1973). In more recent experimental evidence,
ing individuals into social class categories. The sorting of describing a potential interaction partner using upper
people in society based on social class is readily appar- class signals (e.g., plans to travel to Europe during the
ent from a cursory examination of social life. Schools and summer) led to greater affiliation motivation among
neighborhoods are often segregated in terms of social upper class participants, whereas describing a partner
class, with many explicit home lending policies favoring using lower class signals (e.g., plans to get a job over the
neighborhood separation based on social classes (Desmond, summer to help with family bills) led to greater affiliation
2016). People’s romantic preferences also seem to follow motivation among lower class participants (Côté, Kraus,
along social class lines, as the educational attainment, Piff, Beerman, & Keltner, 2014). Aligning with this
annual income, and occupational prestige of romantic research, several correlational studies find that first gen-
partners tends to be highly positively correlated (Schwartz, eration college students report feeling more alienated
2013). Social class sorting is also supported by low levels from university environments than do their continuing
of economic mobility in places like the United States, generation counterparts—a finding that suggests that sig-
where the chances of moving up or down in the social nals of social class on college campuses, which are pre-
class hierarchy are constrained, particularly relative to dominantly populated by people from highly educated
people’s lay estimates (Davidai & Gilovich, 2015; Kraus & families, reduce feelings of belonging among relatively
Tan, 2015). This sorting is one of the likely reasons why lower class individuals (e.g., Bufton, 2003; Ostrove &
428 Kraus et al.

Long, 2007; Reay, Crozier, & Clayton, 2009). Other work adjacent to poor neighborhoods (Pellowski, Kalichman,
on job hiring finds a similar pattern: Interviews reveal Matthews, & Adler, 2013). To the extent health is related
that top-tier business firms rely on social class informa- to group conflict (e.g., Gallo & Matthews, 2003), this
tion in extracurricular activities and club memberships to research is suggestive of the possibility that social class
seek out higher social class applicants (Rivera, 2016), and signals have the capacity to increase conflict between the
experiments find that men described as working class classes.
were less likely to receive a hypothetical job offer for a Several studies support the notion that the presence of
higher class managerial position relative to middle-class class signals contribute to rises in class conflict. On air-
men (Rivera & Tilcsik, 2016). planes, the odds of air rage, defined as extreme antiso-
Class signals also create group boundaries through cial behavior reported during commercial airline flights,
stereotyping—that is, social class signals activate unique increase when class signals are visible—through the
stereotypes about people from distinct ends of the eco- presence versus absence of a first class cabin, or through
nomic spectrum. Though the accuracy of stereotypes is a boarding the plane at the front of the aircraft where the
matter of some controversy (McCauley, Jussim, & Lee, first class cabin is located versus boarding at the rear
1995), several converging lines of evidence suggest that where it is not (DeCelles & Norton, 2016). In laboratory
the use of stereotypes in person perception aid in the research, class signals were directly manipulated during a
formation of group boundaries and the expression of competitive real estate negotiation, where two strangers
prejudice and discrimination (Fiske, 2005; Fiske et al., role-played as a buyer or seller of a biotechnology plant
2002). Studies also indicate that signals of social class are (Kraus & Mendes, 2014). Prior to the negotiation, one of
used as a basis for activating stereotypes. In the original the two participants was assigned to upper class clothing
studies of the stereotype content model by Fiske and col- (i.e., a business suit), lower class clothing (i.e., sweats
leagues (2002), participants rated 23 groups in society and t-shirt), or their own neutral clothing before engag-
based on how others in society view the groups on ing in the negotiation with a neutrally clothed participant
warmth and competence. The studies found that social unaware of the clothing change (Kraus & Mendes, 2014).
class was a reliable predictor of stereotype content across In the study, negotiation pairs where signals of social
studies and samples—the rich were viewed typically as class were manipulated experienced significant differ-
low in warmth but high in competence, whereas the ences in profits, concessions, and physiological states
poor were viewed as both low in warmth and compe- associated with threat vigilance between players whereas
tence (Fiske et al., 2002). It is important to note that the no differences emerged in pairs where class signals were
stereotype content association of high competence to not manipulated (Kraus & Mendes, 2014). In a more
high social class has been replicated across 37 cultures direct experimental test of this conflict hypothesis, par-
(Durante et al., 2013) and, in particular, when assessing ticipants were exposed to pictures of high or low wealth
the stereotypes using social class signals: In one exam- before answering questions about their aggressive ten-
ple, signals of wealth in Italian print media were associ- dencies. Participants exposed to high wealth signals in
ated with ratings of individuals as higher in competence pictures felt temporarily lower in social class and were
(Durante, Volpato, & Fiske, 2010). In other research, more likely to feel aggressive in general—particularly
more or less expensive houses were used to judge the toward a relatively upper-class PhD student who ostensi-
competence of occupants (Oldmeadow & Fiske, 2012). bly designed the wealth stimuli (Greitemeyer & Sagioglou,
Just as class signals separate individuals from different 2016). Though much of this research does not involve
class backgrounds and activate stereotypes, they are also specific interclass conflicts, the findings themselves are
the tinder for class conflict. Specifically, signals of social suggestive of the role of signals of social class in provok-
class increase conflict between individuals of different ing aggressive states for individuals and potentially for
social class groups because they make salient that social groups around class lines.
resources are unequally shared between individuals in Taken together, we have amassed some initial evi-
society. Though people across countries and cultures dence supporting our second prediction that class signals
have different expectations for the ways in which create group boundaries between the haves and have
resources are shared, visible inequalities such as those nots in society by sorting people into social classes, acti-
that are made apparent by social class signals violate vating stereotype content, and enhancing class conflicts.
assumptions about the generosity of neighbors and the These early findings are supportive of our theoretical
fairness of current economic systems. For instance, prior predictions and suggest several avenues of future
research indicates that residents of poor neighborhoods research. For instance, in many of the reviewed experi-
are impacted psychologically by visible class signals— ments, participants are exposed to class signals in con-
experiencing worse health outcomes when they live trolled laboratory settings (e.g., reading about a
adjacent to rich neighborhoods than they do living hypothetical job candidate)—it will be instructive to
Signs of Social Class 429

determine whether perception of class signals during suggest that conscious modification of class signals will
more naturalistic social perception elicits the same group disrupt individual cognitive performance at least until the
processes. As well, some of the prior research indicates behaviors become habitual.
that signals of higher social class are particularly strong
in eliciting conflict between social classes (DeCelles & Class Signals Perpetuate Economic
Norton, 2016; Kraus & Mendes, 2014). It will be interest-
ing in future research to examine if symbols of wealth and
Inequality
poverty exert asymmetric influences on class boundaries. The evidence we have reviewed thus far suggests that
People perceive others on a variety of group dimen- class signals are accurately perceived in brief social behav-
sions, including social categories like race and gender ior and facilitate the creation of class boundaries. In this
that have clear physical characteristics (e.g., skin color, final section, we detail the ways in which class signaling
secondary sexual characteristics) that individuals use processes have the potential to actually perpetuate the
during social perception. How social class group bound- economic conditions they arise from. We predict that class
aries and stereotypes inform, and are informed by, per- signals increase economic inequality maintenance by elic-
ceptions of other categories is an important area of future iting dehumanizing judgments of those at the bottom of
research (e.g., Fiske & Dupree, 2014; Gilens, 1999). In the class hierarchy and strategic sharing patterns that
this regard, the potential permeability of social class favor sharing resources with those who already have
groups is a domain where social class may differ from resources over those without, as well as justifying ideolo-
other categories: Aspirational motivations of relatively gies that elicit perceptions of economic structures as fair,
lower class individuals might lead these individuals to legitimate, mobile, and merit-based.
identify less with their social class group or even to stra- Beliefs about people who occupy the lowest positions
tegically adjust their behavior to signal the social class to on the social class hierarchy include that these individu-
which they aspire. More broadly, thinking of group als are both untrustworthy in social relationships and
boundaries in terms of social class has the potential to incapable of changing their own states (Fiske et al.,
inform theories of intergroup relations by both adding a 2002). Thus, class signals that elicit judgments that one is
social category upon which individuals define their social at the bottom of the economic hierarchy are likely to
groups and by highlighting the intersections of these cat- elicit patterns of social perception suggesting that indi-
egories. For instance, research on stereotype content sug- viduals are simultaneously incapable of achieving desired
gests that the valence of racial stereotypes depends on life outcomes and more likely to harm others. This mix of
social class (e.g., Fiske & Dupree, 2014). social judgments leads perceivers to scorn extremely low
Would people desiring upward mobility benefit from status groups in society (Fiske, 2011) and to discount
modifying their own signals to appear as if they belong their mental states and human characteristics (Cikara,
to the class they aspire to—as is attempted by the char- Bruneau, & Saxe, 2011)—a pattern we and others refer to
acter Eliza Doolittle in the play Pygmalion by George as dehumanization (Haslam, 2006).
Bernard Shaw (1913/2012)? Several studies are suggestive In one illustrative study of this dehumanizing process
of this possibility. For instance, the aforementioned directed at the poor, Harris and Fiske (2006) showed
research on stereotyping suggests that shifting one’s photographs of people from various groups in society.
social class signals to be more in line with higher social Pictures of homeless and drug-addicted individuals, those
class individuals will elicit others’ judgments of height- at the very bottom of the economic spectrum elicited
ened competence (e.g., Fiske et al., 2002). In the afore- self-reports of disgust—an emotion associated with
mentioned sartorial symbols manipulation study, people intense dislike and avoidance of outgroups (Harris &
who were assigned a business suit earned higher profits, Fiske, 2006). In a study with similar methods, photo-
made fewer concessions, and were perceived as more graphs of the poor were less likely to activate the medial
powerful in a dyadic negotiation relative to their interac- prefrontal cortex of perceivers, a brain region associated
tion partner, who was unaware of the clothing assign- with considering the mental states of others (Harris,
ment (Kraus & Mendes, 2014). These data suggest that Cikara, & Fiske, 2008). Signals of lower social class appear
modifying one’s class signals in aspirational ways has to reduce the extent that we consider the minds of the
immediate benefits; however, the sustainability of these poor.
modifications is a matter of future empirical inquiry Aligning with this dehumanization perspective, social
(Fussell, 1992; Labov, 2006). For instance, several studies class signals are likely to elicit strategic patterns of
indicate that awareness of one’s lower status relative to resource sharing. Part of the utility of signaling status in
others is cognitively demanding ( Johnson, Richeson, & interpersonal life is the capacity to weigh the costs and
Finkel, 2011) or likely to elicit stereotype threat (Croizet benefits of cooperation (Ambady & Rosenthal, 1993;
& Claire, 1998; Steele & Aronson, 1995), and these studies Kraus & Keltner, 2009). In this vein, sharing resources
430 Kraus et al.

with a person with low economic standing may be per- lead to the justification of economic inequality—particu-
ceived as risky behavior because such a sharing relation- larly for those at the top of the economic hierarchy who
ship has few direct economic benefits for the sharer. stand to gain the most from the current economic system.
Based on this analysis, one possible consequence is that Though no study to date has examined the role of class
visible class signals perpetuate inequality in society by signals in particular in eliciting specific justifying ideolo-
discouraging resource sharing with those that have less gies, several studies indirectly illustrate this in laboratory
in favor of sharing with others of similar class standing. settings. In one set of studies, temporarily manipulating
Several lines of evidence support this assertion. In an an individual’s relative standing to be higher than other
experiment directly testing whether visible class signals people in an economic game led to seeing the economic
influence strategic sharing, researchers created miniature game itself as fairer (Brown-Iannuzi, Lundberg, Kay, &
societies in an experimental economic game where par- Payne, 2015). In another study, asking participants to
ticipants shared resources with each other across time think of someone way below them on a social class lad-
(Nishi, Shirado, Rand, & Christakis, 2015). In the study, der in society elicited beliefs that social class mobility was
some societies were manipulated to have individuals higher in society than it actually was (Kraus & Tan, 2015).
with visible resources whereas in other societies resources To the extent that class signals activate these same com-
were invisible. When resources were visible, participants parison processes, we expect these signals to elicit similar
shared their resources with other resource-rich individu- judgments of unequal economic systems as more fair,
als, thereby perpetuating economic inequality. In con- mobile, merit-based, and justified.
trast, when resources were invisible, participants shared The above evidence suggests that class signals activate
regardless of the resources of their partners, thereby individual and group processes that perpetuate economic
reducing economic inequality (Nishi et al., 2015). These inequality. Several lines of future research are necessary
findings suggest that visible class signals perpetuate eco- to more rigorously test this theoretical prediction. For
nomic inequality by eliciting strategic resource sharing instance, much of this work relies on laboratory studies
patterns. In other research, higher income individuals in which class signals are artificially made visible (Nishi
gave less in a single trial dictator game to an anonymous et al., 2015) or brought to mind (Brown-Iannuzi et al.,
other when living in states that were more unequal or 2015). How actual class signals activate these same pro-
when exposed to high levels of economic inequality cesses as they are expressed in everyday social interac-
(Côté, House, & Willer, 2015). These results suggest that tions remains a promising area of future research. For
awareness of social class differences elicits lower levels instance, the concealment of visible social class is a rea-
of resource sharing among the wealthy. son used to justify school uniform policies in more than
Finally, we assert that signals of social class can also 20% of schools in the United States (Ball, Bowe, &
activate basic dissonance processes that perpetuate eco- Gewirtz, 1996). Do uniforms reduce social class signaling
nomic inequality. Cognitive dissonance occurs when and class conflict, or alternatively, do the myriad and
people are forced to reconcile conflicting attitudes or subtle ways in which behaviors signal social class render
behaviors (Festinger & Carlsmith, 1959; Kunda, 1990), such uniform policies ineffective in concealing social
and economic inequality is an economic state that fre- class?
quently elicits dissonance processes (Kraus & Tan, 2015; Also, little is known about the mechanisms that might
Shariff et al., 2016; c.f., Jost, Banaji, & Nosek, 2004; Jost, counteract the ways that class signaling perpetuates eco-
Pelham, & Carvallo, 2002). In essence, when levels of nomic inequality. It is possible, for instance, that elevat-
economic inequality widen, beliefs that society is fair and ing compassionate responses to the needy, such as by
just are directly challenged. These dissonance processes eliciting judgments of high self–other similarity with
associated with economic inequality, in part, explain why lower class others (Oveis, Horberg, & Keltner, 2009), will
people underestimate levels of economic inequality countervail the psychological processes related to eco-
(Norton & Ariely, 2011), class mobility (Davidai & Gilov- nomic inequality maintenance. However, the evidence
ich, 2015), and CEO pay relative to the average worker we have reviewed thus far suggests that barriers to the
(Kiatpongsan & Norton, 2014). experience of compassion across social classes are sig-
Evidence we have reviewed thus far indicates that class nificant: People of higher social class standing view lower
signals are perceived rapidly and accurately, activate class social class others as outgroup members who are less
stereotypes, and augment class boundaries. Many of these human, less warm, and less competent than their same
psychological consequences of class signaling are likely class counterparts and each of these perceptions is likely
to elicit uncomfortable social comparisons about one’s to create a compassion barrier. Given these patterns, pol-
standing in society relative to others (Buunk et al., 2003). icies that aim to improve the lives of the poor by relying
Through this process, we suggest that class signals are on wealthy individuals and corporations to act compas-
also likely to elicit economic dissonance processes that sionately, with the best interests of lower class individuals
Signs of Social Class 431

in mind, do not rest on the available empirical evidence white supremacy (Kharakh & Primack, 2016) are also
(e.g., Bresnahan & Rogers, 2016). both potential examples of realistic conflict playing out in
current events. To the extent that these broad societal
trends are influenced by everyday signals of social class
Class Signals as Everyday Inequality
is an empirical question worth scrutiny.
In this review, we argue that the global rising tide of eco- The reviewed evidence suggests a few additional
nomic inequality is experienced at the interpersonal future empirical directions of note. In particular, there are
level. We contend that subtle differences in the ways in several implications for the process of class mobility
which people behave toward and even speak about the highlighted by our theory: The evidence for class signal-
haves and have nots of society are perceived rapidly and ing suggests that one possible barrier to upward social
accurately and that, as a consequence of these percep- class mobility is that social class is perceived rapidly and
tions, social class divides sectors of society and economic accurately in basic person perception and that these
inequality is perpetuated. We support this theoretical judgments may elicit assessments of one’s lower compe-
position with basic research studies from the psychologi- tence that could impact promotion and hiring decisions
cal and economic sciences that examine class signaling in (Rivera & Tilcsik, 2016) or feelings of belonging at a uni-
the context of observational studies of visible economic versity that are so crucial for academic achievement and
inequality (DeCelles & Norton, 2016), as well as con- striving (Stephens, Hamedani, & Destin, 2014). As such,
trolled laboratory experiments (Nishi et al., 2015). perhaps individual differences in code switching apti-
If our theoretical predictions are accurate, how might tude, the ability to switch between group contexts and to
social class signals shape society? One prediction derives blend in with others (Labov, 2006; Steele, 2011), predict
from the daily experience of relatively lower class indi- an individual’s capacity to move up the class hierarchy.
viduals who, according to our theory, will see themselves How societal institutions—like schools and employers—
as chronically lower in societal status relative to others can reduce the impact of class signaling, or alternatively,
across the domains of their lives. A consequence of these weigh those signals in admission and hiring decisions is
chronic perceptions might be widespread disengagement also a crucial area of future inquiry.
from politics as lower class individuals come to distrust a Related to this point, the theoretical model we advo-
political system that ignores their own daily economic cate for here relates to past research examining the psy-
struggles. We see some evidence for these political trends chological experience of individuals at the bottom of the
currently in the United States where lower class individu- class hierarchy. This research indicates that lower social
als participate less in general elections than their rela- class individuals are more vigilant of threats in their envi-
tively upper class counterparts (Census.gov; Laurison, ronments (Gallo & Matthews, 2003), exhibit reduced
2016; McElwee, 2016) and where both members of Con- executive functioning (Mullainathan & Shafir, 2013), and
gress and current Presidential candidates have the high- experience poorer health and well-being outcomes rela-
est disapproval ratings on record (Enten, 2016). To the tive to their more advantaged counterparts (Adler et al.,
extent that social class signals contribute to these trends 1994; for a review, see Kraus et al., 2012). Future research
in politics is an empirical question worthy of further could uncover the extent that class signals are causal in
research. eliciting these patterns, and some research is suggestive
For relatively lower class Whites, daily perceptions of of this possibility: As we mentioned previously, environ-
subordinate societal status are also likely to sow inter- ments where inequality is more visible elicit greater
group conflict, particularly with respect to other tradi- threat-related aggressive tendencies (DeCelles & Norton,
tionally disadvantaged groups (cf., Gilens, 1999): When 2016; Greitemeyer & Sagioglou, 2016) and more negative
new immigrants, refugees, women, or ethnic minorities health outcomes for the poor (Pellowski et al., 2013).
receive (or are simply perceived to receive) benefits from Overall then, the impact of economic inequality is both
the government that are not explicitly shared by lower significant and widespread. Significant because it predicts
class Whites, these benefits have the potential to create the health and well-being of members of society, and
realistic group conflict (e.g., Bobo, 1983) because gov- widespread because of the ways in which people can
ernment policies seem to respond to one kind of injus- experience inequality in the briefest of social interactions
tice (e.g., racial discrimination) and not another (e.g., with strangers, while boarding airplanes, and in travel
conditions of poverty). One example of this dynamic is between schools and neighborhoods. Although societal
the belief among Whites that Blacks are the sole and ille- economic inequality has received growing empirical and
gitimate beneficiaries of welfare (Gilens, 1999). Wide- public attention, the everyday experience of that inequal-
spread concern about taking on Syrian refugees across ity is a largely unexamined phenomenon in need of
the United States and Europe and the existence of a cur- greater scrutiny: Macro-economic policy changes may not
rent U.S. President who has openly expressed racism and receive widespread public support, regardless of their
432 Kraus et al.

benefits, if those policies do not also articulate the ways Becker, J., Kraus, M. W., & Rheinschmidt-Same, M. L. (2017).
in which they contend with social class dynamics in daily Cultural expressions of social class and their implications
social interactions. To wit, passage of the UK referendum for beliefs and behavior. Journal of Social Issues.
to leave the European Union occurred with the support of Bergeron, A. P., & Zanna, M. P. (1973). Group membership and
belief similarity as determinants of interpersonal attraction
relatively lower class individuals whose concerns cen-
in Peru. Journal of Cross-Cultural Psychology, 4, 397–411.
tered more on the importance of allocating resources to
Bernstein, B. (1962). Social class, linguistic codes and gram-
national programs, and less on the impact that leaving matical elements. Language and Speech, 5, 221–240.
might have on the economic fortunes of banks and multi- Bjornsdottir, R. T., & Rule, N. O. (2016, January). Conspicuous
national corporations (Barr, 2016). How nations, cultures, class: Wealth is accurately perceived from the face.
and societies respond to economic inequality is a conten- Presentation at the Emerging Psychology of Social Class
tious and complex issue with significant social and eco- Pre-Conference of the Society for Personality and Social
nomic implications for societies and, as we have argued, Psychology, San Diego, CA.
the daily social and psychological experiences of indi- Bobo, L. (1983). Whites’ opposition to busing: Symbolic racism
viduals who populate them. or realistic group conflict? Journal of Personality and Social
Psychology, 45, 1196–1210.
Bourdieu, P. (1984). Distinction: A social critique of the judge-
Acknowledgments
ment of taste (R. Nice, Trans.). Cambridge, MA: Harvard
Data for this manuscript is publically available at the Open Sci- University Press.
ence Framework: https://osf.io/wq9gy/. For sample audio stimuli, Brannon, T. N., & Markus, H. R. (2013). Social class and race:
please visit https://soundcloud.com/michael-kraus-678964220/ Burdens but also some benefits of chronic low rank.
sets/oregon-2. The authors wish to thank Paul Meier of the Psychological Inquiry, 24, 97–101.
International Dialects of English Archive for providing access to Bresnahan, J., & Rogers, D. (2016). New House Republican
the speaker audio files, Jennifer MacCormack for early assis- “anti-poverty plan” repackages GOP proposals. Politico.
tance with this project, Stephen Dionne for help with article Retrieved from http://www.politico.com/story/2016/06/
formatting, and Serena Chen, Dacher Keltner, Wendy Mendes, house-gop-poverty-223969
and Bill Kraus for helpful comments that shaped our thinking Brickman, P., & Bulman, R. J. (1977). Pleasure and pain in
about this work. social comparison. In J. M. Suls & R. L. Miller (Eds.), Social
comparison processes: Theoretical and empirical perspec-
Declaration of Conflicting Interests tives (pp. 149–186). Washington, DC: Hemisphere.
The authors declared that they had no conflicts of interest with Brown, D. J., Ferris, D. L., Heller, D., & Keeping, L. M. (2007).
respect to their authorship or the publication of this article. Antecedents and consequences of the frequency of upward
and downward social comparisons at work. Organizational
Behavior and Human Decision Processes, 102, 59–75.
Supplemental Material Brown-Iannuzzi, J. L., Lundberg, K. B., Kay, A. C., & Payne,
Additional supporting information may be found online. B. K. (2015). Subjective status shapes political preferences.
Psychological Science, 26, 15–26.
Bufton, S. (2003). The lifeworld of the university student:
References
Habitus and social class. Journal of Phenomenological
Adler, N. E., Boyce, T., Chesney, M. A., Cohen, S., Folkman, Psychology, 34, 207–234.
S., Kahn, R. L., & Syme, S. L. (1994). Socioeconomic sta- Bugental, D. B., Henker, B., & Whalen, C. K. (1976). Attributional
tus and health: The challenge of the gradient. American antecedents of verbal and vocal assertiveness. Journal of
Psychologist, 49, 15–24. Personality and Social Psychology, 34, 405–411.
Adler, N. E., Epel, E. S., Castellazzo, G., & Ickovics, J. R. (2000). Buunk, B. P., Zurriaga, R., Gonzalez-Roma, V., & Subirats, M.
Relationship of subjective and objective social status with psy- (2003). Engaging in upward and downward comparisons
chological and physiological functioning: Preliminary data as a determinant of relative deprivation at work: A longitu-
in healthy white women. Health Psychology, 19, 586–592. dinal study. Journal of Vocational Behavior, 62, 370–388.
Ambady, N., & Rosenthal, R. (1993). Half a minute: Predicting Cikara, M., Bruneau, E. G., & Saxe, R. R. (2011). Us and them
teacher evaluations from thin slices of nonverbal behav- intergroup failures of empathy. Current Directions in
ior and physical attractiveness. Journal of Personality and Psychological Science, 20, 149–153.
Social Psychology, 64, 431–460. Côté, S., House, J., & Willer, R. (2015). High economic inequal-
Ball, S. J., Bowe, R., & Gewirtz, S. (1996). School choice, social ity leads higher-income individuals to be less generous.
class and distinction: The realization of social advantage in Proceedings of the National Academy of Sciences, USA, 112,
education. Journal of Education Policy, 11, 89–112. 15838–15843.
Barr, C. (2016, June 24). The areas and demographics where Côté, S., Kraus, M., Piff, P. K., Beermann, U., & Keltner, D. J.
the Brexit vote was won. The Guardian. Retrieved from (2014). Social class clash: A dyadic model of social affili-
https://www.theguardian.com/news/datablog/2016/ ation in cross-class and same-class interactions (Rotman
jun/24/the-areas-and-demographics-where-the-brexit-vote- School of Management Working Paper No. 2503031).
was-won Toronto, Canada: University of Toronto.
Signs of Social Class 433

Croizet, J. C., & Claire, T. (1998). Extending the concept of ste- Fogel, H., & Ehri, L. C. (2000). Teaching elementary stu-
reotype threat to social class: The intellectual underperfor- dents who speak Black English Vernacular to write in
mance of students from low socioeconomic backgrounds. Standard English: Effects of dialect transformation practice.
Personality and Social Psychology Bulletin, 24, 588–594. Contemporary Educational Psychology, 25, 212–235.
Davidai, S., & Gilovich, T. (2015). Building a more mobile Fussell, P. (1992). Class: A guide through the American status
America—One income quintile at a time. Perspectives on system. New York, NY: Simon & Schuster.
Psychological Science, 10, 60–71. Gallo, L. C., & Matthews, K. A. (2003). Understanding the asso-
Davis, J. A. (1956). Status symbols and the measurement of ciation between socioeconomic status and physical health:
status perception. Sociometry, 19, 154–165. Do negative emotions play a role? Psychological Bulletin,
DeCelles, K. A., & Norton, M. I. (2016). Physical and situational 129, 10–51.
inequality on airplanes predicts air rage. Proceedings of the Ganzeboom, H., De Graaf, P. M., & Treiman, D. J. (1992). A
National Academy of Sciences, USA, 113, 5588–5591. standard international socio-economic index of occupa-
Desmond, M. (2016). Evicted: Poverty and profit in the American tional status. Social Science Research, 21, 1–56.
city. New York, NY: Broadway Books. Gilens, M. (1999). Why Americans hate welfare. Chicago, IL:
DiMaggio, P., & Useem, M. (1978). Social class and arts con- University of Chicago Press.
sumption. Theory and Society, 5, 141–161. Giles, H., & Sassoon, C. (1983). The effect of speaker’s accent,
Dunbar, N. E., & Burgoon, J. K. (2005). Perceptions of power social class background and message style on British lis-
and interactional dominance in interpersonal relationships. teners’ social judgements. Language & Communication, 3,
Journal of Social and Personal Relationships, 22, 207–233. 305–313.
Durante, F., Fiske, S. T., Kervyn, N., Cuddy, A. J., Akande, A. Gillath, O., Bahns, A. J., Ge, F., & Crandall, C. S. (2012). Shoes
D., Adetoun, B. E., & Barlow, F. K. (2013). Nations’ income as a source of first impressions. Journal of Research in
inequality predicts ambivalence in stereotype content: How Personality, 46, 423–430.
societies mind the gap. British Journal of Social Psychology, Gluszek, A., & Dovidio, J. F. (2010). The way they speak: A
52, 726–746. social psychological perspective on the stigma of non-
Durante, F., Volpato, C., & Fiske, S. T. (2010). Using the native accents in communication. Personality and Social
Stereotype Content Model to examine group depictions in Psychology Review, 14, 214–237.
fascism: An archival approach. European Journal of Social Goodman, E., Adler, N. E., Kawachi, I., Frazier, A. L., Huang, B.,
Psychology, 40, 465–483. & Colditz, G. A. (2001). Adolescents’ perceptions of social
Ellis, D. S. (1967). Speech and social status in America. Social status: Development and evaluation of a new indicator.
Forces, 45, 431–437. Pediatrics, 108(2), e31.
Enten, H. (2016, May 5). Americans’ distaste for both Trump Gregory, S. W., Jr., & Webster, S. (1996). A nonverbal signal in
and Clinton is record-breaking. Five Thirty Eight. Retrieved voices of interview partners effectively predicts communica-
from http://fivethirtyeight.com/features/americans-distaste- tion accommodation and social status perceptions. Journal
for-both-trump-and-clinton-is-record-breaking/ of Personality and Social Psychology, 70, 1231–1240.
Festinger, L. (1954). A theory of social comparison processes. Greitemeyer, T., & Sagioglou, C. (2016). Subjective socioeco-
Human Relations, 7, 117–140. nomic status causes aggression: A test of the theory of social
Festinger, L., & Carlsmith, J. M. (1959). Cognitive consequences deprivation. Journal of Personality and Social Psychology,
of forced compliance. The Journal of Abnormal and Social 111, 178–194.
Psychology, 58, 203–210. Hall, J. A., Coats, E. J., & LeBeau, L. S. (2005). Nonverbal behav-
Fiske, S. T. (2005). Social cognition and the normality of pre- ior and the vertical dimension of social relations: A meta-
judgment. In J. F. Dovidio, P. Glick, & L. A. Budman (Eds.), analysis. Psychological Bulletin, 131, 898–924.
On the nature of prejudice: Fifty years after Allport (pp. Harris, L. T., Cikara, M., & Fiske, S. T. (2008). Envy as predicted
36–53). Malden, MA: Blackwell. by the stereotype content model: Volatile ambivalence. In
Fiske, S. T. (2011). Envy up, scorn down: How status divides us. R. Smith (Ed.), Envy: Theory and research (pp. 133–147).
New York, NY: Russell Sage Foundation. New York, NY: Oxford University Press.
Fiske, S. T., Cuddy, A. J., Glick, P., & Xu, J. (2002). A model of Harris, L. T., & Fiske, S. T. (2006). Dehumanizing the lowest of
(often mixed) stereotype content: Competence and warmth the low neuroimaging responses to extreme out-groups.
respectively follow from perceived status and competition. Psychological Science, 17, 847–853.
Journal of Personality and Social Psychology, 82, 878–902. Haslam, N. (2006). Dehumanization: An integrative review.
Fiske, S. T., & Dupree, C. (2014). Gaining trust as well as respect Personality and Social Psychology Review, 10, 252–264.
in communicating to motivated audiences about science Hertenstein, M. J., Holmes, R., McCullough, M., & Keltner, D.
topics. Proceedings of the National Academy of Sciences, (2009). The communication of emotion via touch. Emotion,
USA, 111, 13593–13597. 9, 566–573.
Fiske, S. T., & Markus, H. R. (2012). Facing social class. New Hout, M. (2008). How class works: Objective and subjective
York, NY: Russell Sage Foundation. aspects of class since the 1970s. In A. Lareau & D. Conley
Fliessbach, K., Weber, B., Trautner, P., Dohmen, T., Sunde, U., (Eds.), Social class: How does it work? (pp. 25–64). New
Elger, C. E., & Falk, A. (2007). Social comparison affects York, NY: Russell Sage Foundation.
reward-related brain activity in the human ventral striatum. Johnson, S. E., Richeson, J. A., & Finkel, E. J. (2011). Middle
Science, 318, 1305–1308. class and marginal? Socioeconomic status, stigma, and
434 Kraus et al.

self-regulation at an elite university. Journal of Personality Labov, W. (2006). The social stratification of English in New
and Social Psychology, 100, 838–852. York City. London, United Kingdom: Cambridge University
Jost, J. T., Banaji, M. R., & Nosek, B. A. (2004). A decade of sys- Press.
tem justification theory: Accumulated evidence of conscious Lareau, A., & Conley, D. (Eds.). (2008). Social class: How does it
and unconscious bolstering of the status quo. Political Psyc­ work? New York, NY: Russell Sage Foundation.
hology, 25, 881–919. Laurison, D. (2016). Social class and political engagement in the
Jost, J. T., Pelham, B. W., & Carvallo, M. R. (2002). Non-conscious United States. Sociology Compass, 10, 684–697.
forms of system justification: Implicit and behavioral pref- Lizardo, O., & Skiles, S. (2015). Musical taste and patterns
erences for higher status groups. Journal of Experimental of symbolic exclusion in the United States 1993–2012:
Social Psychology, 38, 586–602. Generational dynamics of differentiation and continuity.
Kharakh, B., & Primack, D. (2016, March 22). Donald Trump’s Poetics, 53, 9–21.
social media ties to white supremacists. Fortune. Retrieved McCauley, C. R., Jussim, L. J., & Lee, Y. T. (1995). Stereotype accu-
from http://fortune.com/donald-trump-white-supremacist- racy: Toward appreciating group differences. Washington,
genocide/ DC: American Psychological Association.
Kiatpongsan, S., & Norton, M. I. (2014). How much (more) McElwee, S. (2016). The income gap at the polls. Politico.
should CEOs make? A universal desire for more equal pay. Retrieved from http://www.politico.com/magazine/story/
Perspectives on Psychological Science, 9, 587–593. 2015/01/income-gap-at-the-polls-113997
Kluegel, J. R., & Smith, E. R. (1986). Beliefs about inequality: McNicol, D. (2005). A primer of signal detection theory. Mahwah,
Americans’ views of what is and what ought to be. New NJ: Erlbaum.
Brunswick, NJ: Transaction Publishers. Monsivais, P., & Drewnowski, A. (2009). Lower-energy-density
Ko, S. J., Sadler, M. S., & Galinsky, A. D. (2015). The sound of diets are associated with higher monetary costs per kilo-
power conveying and detecting hierarchical rank through calorie and are consumed by women of higher socioeco-
voice. Psychological Science, 26, 3–14. nomic status. Journal of the American Dietetic Association,
Kohn, M. L. (1963). Social class and parent-child relationships: An 109, 814–822.
interpretation. American Journal of Sociology, 68, 471–480. Montoya, R. M., Horton, R. S., & Kirchner, J. (2008). Is actual
Kraus, M. W., & Chen, T. W. D. (2013). A winning smile? Smile similarity necessary for attraction? A meta-analysis of actual
intensity, physical dominance, and fighter performance. and perceived similarity. Journal of Social and Personal
Emotion, 13, 270–279. Relationships, 25, 889–922.
Kraus, M. W., & Keltner, D. (2009). Signs of socioeconomic status: Mullainathan, S., & Shafir, E. (2013). Scarcity: Why having too
A thin-slicing approach. Psychological Science, 20, 99–106. little means so much. Macmillan.
Kraus, M. W., & Mendes, W. B. (2014). Sartorial symbols of Nickerson, R. S. (1998). Confirmation bias: A ubiquitous phe-
social class elicit class-consistent behavioral and physiolog- nomenon in many guises. Review of General Psychology,
ical responses: A dyadic approach. Journal of Experimental 2, 175–220.
Psychology: General, 143, 2330–2340. Nishi, A., Shirado, H., Rand, D. G., & Christakis, N. A. (2015).
Kraus, M. W., Piff, P. K., & Keltner, D. (2009). Social class, sense Inequality and visibility of wealth in experimental social
of control, and social explanation. Journal of Personality networks. Nature, 526, 426–429.
and Social Psychology, 97, 992–1004. Norton, M. I., & Ariely, D. (2011). Building a better America—
Kraus, M. W., Piff, P. K., Mendoza-Denton, R., Rheinschmidt, One wealth quintile at a time. Perspectives on Psychological
M. L., & Keltner, D. (2012). Social class, solipsism, and Science, 6, 9–12.
contextualism: How the rich are different from the poor. Oakes, J. M., & Rossi, P. H. (2003). The measurement of SES in
Psychological Review, 119, 546–572. health research: Current practice and steps toward a new
Kraus, M. W., & Tan, J. J. (2015). Americans overestimate social approach. Social Science & Medicine, 56, 769–784.
class mobility. Journal of Experimental Social Psychology, Organisation for Economic Co-operation and Development.
58, 101–111. (2014). OECD factbook 2014: Economic, environmental
Kraus, M. W., Tan, J. J., & Tannenbaum, M. B. (2013). The and social statistics. Paris: Author. Retrieved from http://
social ladder: A rank-based perspective on social class. dx.doi.org/10.1787/factbook-2014-en
Psychological Inquiry, 24, 81–96. Oishi, S., & Talhelm, T. (2012). Residential mobility: What
Kunda, Z. (1990). The case for motivated reasoning. Psychological psychological research reveals. Current Directions in
Bulletin, 108, 480–498. Psychological Science, 21, 425–430.
Labov, W. (1964). Phonological correlates of social stratifica- Oldmeadow, J. A., & Fiske, S. T. (2012). Contentment to resent-
tion. American Anthropologist, 66, 164–176. ment: Variation in stereotype content across status systems.
Labov, W. (1972). The social stratification of (r) in New York Analyses of Social Issues and Public Policy, 12, 324–339.
City department stores. In W. Labov (Ed.), Sociolinguistic Ostrove, J. M., & Long, S. M. (2007). Social class and belonging:
patterns (Vol. 4, pp. 43–70). Philadelphia, PA: University of Implications for college adjustment. The Review of Higher
Pennsylvania Press. Education, 30, 363–389.
Labov, W. (1990). The intersection of sex and social class in Oveis, C., Horberg, E. J., & Keltner, D. (2009). Compassion,
the course of linguistic change. Language Variation and pride, and social intuitions of self-other similarity. Journal
Change, 2, 205–254. of Personality and Social Psychology, 98, 618–630.
Signs of Social Class 435

Pellowski, J. A., Kalichman, S. C., Matthews, K. A., & Adler, N. Steele, C. M. (2011). Whistling Vivaldi: And other clues to how
(2013). A pandemic of the poor: Social disadvantage and stereotypes affect us (issues of our time). New York, NY: W.W.
the US HIV epidemic. American Psychologist, 68, 197–209. Norton.
Peterson, R. A., & Kern, R. M. (1996). Changing highbrow taste: Steele, C. M., & Aronson, J. (1995). Stereotype threat and the
From snob to omnivore. American Sociological Review, 61, intellectual test performance of African Americans. Journal
900–907. of Personality and Social Psychology, 69, 797–811.
Piketty, T. (2015). The economics of inequality. Cambridge, MA: Stephens, N. M., Hamedani, M. G., & Destin, M. (2014).
Harvard University Press. Closing the social-class achievement gap a difference-
Piketty, T., & Saez, E. (2014). Inequality in the long run. Science, education intervention improves first-generation students’
344, 838–843. academic performance and all students’ college transition.
Reay, D., Crozier, G., & Clayton, J. (2009). ‘Strangers in para- Psychological Science, 25, 943–953.
dise’? Working-class students in elite universities. Sociology, Stephens, N. M., Markus, H. R., & Fryberg, S. A. (2012). Social
43, 1103–1121. class disparities in health and education: Reducing inequal-
Rivera, L. A. (2016). Pedigree: How elite students get elite jobs. ity by applying a sociocultural self model of behavior.
Princeton, NJ: Princeton University Press. Psychological review, 119, 723–744.
Rivera, L. A., & Tilcsik, A. (2016). Class advantage, commitment Swets, J. A. (2014). Signal detection theory and ROC analysis
penalty: The gendered effect of social class signals in an in psychology and diagnostics: Collected papers. New York,
elite labor market. American Sociological Review. Advance NY: Psychology Press.
online publication. Taylor, S. E., & Lobel, M. (1989). Social comparison activity
Sapolsky, R. M. (2005). The influence of social hierarchy on under threat: Downward evaluation and upward contacts.
primate health. Science, 308, 648–652. Psychological Review, 96, 569–575.
Schmid Mast, M., & Hall, J. A. (2004). Who is the boss and who Tracy, J. L., & Matsumoto, D. (2008). The spontaneous expres-
is not? Accuracy of judging status. Journal of Nonverbal sion of pride and shame: Evidence for biologically innate
Behavior, 28, 145–165. nonverbal displays. Proceedings of the National Academy of
Schwartz, C. R. (2013). Trends and variation in assortative mat- Sciences, USA, 105, 11655–11660.
ing: Causes and consequences. Annual Review of Sociology, Van Eijck, K. (2001). Social differentiation in musical taste pat-
39, 451–470. terns. Social forces, 79, 1163–1185.
Sen, M., & Wasow, O. (2016). Race as a bundle of sticks: Designs Veblen, T. (1973). The theory of the leisure class. New York, NY:
that estimate effects of seemingly immutable characteristics. Houghton Mifflin Harcourt. (Original work published 1899)
Annual Review of Political Science, 19, 499–522. Weininger, E. B., & Lareau, A. (2009). Paradoxical pathways:
Sergio, F., Blas, J., Blanco, G., Tanferna, A., López, L., Lemus, An ethnographic extension of Kohn’s findings on class and
J. A., & Hiraldo, F. (2011). Raptor nest decorations are a childrearing. Journal of Marriage and Family, 71, 680–695.
reliable threat against conspecifics. Science, 331, 327–330. Wilkinson, R. G., & Pickett, K. E. (2006). Income inequality and
Shariff, A. F., Wiwad, D., & Aknin, L. B. (2016). Income mobil- population health: A review and explanation of the evi-
ity breeds tolerance for income inequality cross-national dence. Social Science & Medicine, 62, 1768–1784.
and experimental evidence. Perspectives on Psychological Wilkinson, R. G., & Pickett, K. E. (2009). Income inequality and
Science, 11, 373–380. social dysfunction. Annual Review of Sociology, 493–511.
Shaw, G. B. (2012). Pygmalion. New York, NY: Simon & Wolfram, W., & Schilling-Estes, N. (2006). American English:
Schuster. (Original work published 1913) Dialects and variation (2nd ed.). Malden, MA: Black­well.
Snibbe, A. C., & Markus, H. R. (2005). You can’t always get Zahavi, A., & Zahavi, A. (1999). The handicap principle: A
what you want: Educational attainment, agency, and choice. missing piece of Darwin’s puzzle. New York, NY: Oxford
Journal of Personality and Social Psychology, 88, 703–720. University Press.