Content uploaded by Joshua Foster
Author content
All content in this area was uploaded by Joshua Foster on Sep 27, 2018
Content may be subject to copyright.
Individual differences in narcissism: Inflated
self-views across the lifespan and around
the world
q
Joshua D. Foster,
a,*
W. Keith Campbell,
a
and Jean M. Twenge
b
a
Department of Psychology, University of Georgia, Athens, GA 30602, USA
b
Department of Psychology, San Diego State University, San Diego, CA 92182-4611, USA
Abstract
The present investigation examined associations among narcissism, age, ethnicity, world re-
gion, and gender, using a large (n¼3445) sample of participants representing several different
world regions and ethnicities. The results suggest that (1) reported narcissism declines in older
participants, (2) consistent with previous findings, males report being more narcissistic than
females, (3) that ethnic differences in reported narcissism are generally comparable to those
found in the self-esteem literature, and (4) that world region appears to exert influence on nar-
cissism, with participants from more individualistic societies reporting more narcissism. The
results are discussed in terms of how age and culture might impact narcissism and how future
research might address this topic.
Ó2003 Elsevier Science (USA). All rights reserved.
1. Introduction
Culture and development across the lifespan play crucial roles in shaping the self.
Personality and general character sometimes change as people age, especially as they
Journal of Research in Personality 37 (2003) 469–486
www.elsevier.com/locate/jrp
JOURNAL OF
RESEARCH IN
PERSONALITY
q
We thank Ilan Shrira and Hope Jackson for their insightful comments concerning a draft of this
article. We thank the Social Psychology Network (socialpsychology.org) for allowing us to advertise our
questionnaire on their website. We also thank the many individuals who helped us pilot test the website
and made suggestions about its content and style. Finally, we thank those who completed our
questionnaire and the many individuals who emailed us questions regarding it.
*
Corresponding author. Fax: 1-706-542-3275.
E-mail address: jfoster@uga.edu (J.D. Foster).
0092-6566/$ - see front matter Ó2003 Elsevier Science (USA). All rights reserved.
doi:10.1016/S0092-6566(03)00026-6
move through adolescence and young adulthood (e.g., Ozer & Gjerde, 1989). Culture
also exerts a great deal of pressure on the shaping of personality. For example, the
DSM-IV TR notes that people who have recently immigrated may appear to have
diagnosable personality disorders when, in fact, they are simply expressing personal-
ity traits common to their country of origin (American Psychiatric Association,
2000). Indeed, many authors have argued that culture strongly influences our per-
sonalities and views of self (e.g., Heine & Lehman, 1997; Markus & Kitayama,
1991). Even within a single country, culture may influence people to define them-
selves quite differently (Plaut, Markus, & Lachman, 2002).
In the present investigation, we examined how age and culture influence the per-
sonality construct of narcissism. In order to collect data from as diverse a sample as
possible, we used the Internet. The Internet is fast becoming recognized as a valid
and reliable tool for data collection and has been utilized in several large-scale pro-
jects that collected data from thousands of participants worldwide. For example,
Robins, Trzniewski, Tracy, Gosling, and Potter (2002) collected self-esteem reports
from a worldwide sample of participants. Another set of researchers used the Inter-
net to collect self-report personality questionnaires from a large set of respondents
representing different ages (Srivastava, John, Gosling, & Potter, in press). The results
of investigations such as these may yield valuable insight into the effects of age and
culture on personality.
The goal of the present investigation was twofold. First, we wanted to collect data
on narcissism from a larger and more inclusive sample of participants compared to
what one typically finds in the narcissism literature. Specifically, we wanted to gather
data on narcissism from people who represented various age and ethnic identity cat-
egories as well as different regions of the world. Second, we wanted to test several
specific hypotheses and conduct exploratory analyses with narcissism on this large
and inclusive sample. Specifically, we wanted to determine whether age, ethnic iden-
tity, and country of residence are related to narcissism. We also wanted to replicate
previous research showing that men usually report more narcissism than women
(e.g., Bushman & Baumeister, 1999; Farwell & Wohlwend-Lloyd, 1998; Joubert,
1998; Ladd, Welsh, Vitulli, Labbe, & Law, 1997). Before we state our specific predic-
tions we briefly address the general issue of narcissism and how it relates to other
psychological constructs.
Narcissism has a brief but rich history of psychological investigation. Early re-
search in this area centered on narcissism as a personality disorder. The DSM-IV
TR defines narcissistic personality disorder (NPD) as lack of empathy, need for ad-
miration, and a pattern of grandiosity (American Psychiatric Association, 2000).
Personality/social psychological researchers, however, focus on ‘‘sub-clinical’’ or
‘‘normal’’ narcissists—those who display some of the characteristics of NPD, but
not necessarily enough to be diagnosed with NPD. In the present paper, when
we refer to the term ‘‘narcissist,’’ we are using the personality/social psychological
definition.
Narcissism is correlated with several undesirable traits and behaviors. For exam-
ple, narcissists tend to be less agreeable (Bradlee & Emmons, 1992), tend to be mo-
tivated less by intrinsic and more by extrinsic desires (Kasser & Ryan, 1996), and
470 J.D. Foster et al. / Journal of Research in Personality 37 (2003) 469–486
tend to brag about their accomplishments and display a relatively arrogant attitude
(Paulhus, 1998). However, there are many positive aspects to narcissism as well. For
example, narcissists tend to be highly extraverted (Bradlee & Emmons, 1992),
socially bold (Emmons, 1984), less depressed, and less socially anxious (Watson &
Biderman, 1993).
In addition, narcissism is positively correlated with self-esteem (e.g., Emmons,
1984; Jackson, Ervin, & Hodge, 1992; Raskin, Novacek, & Hogan, 1991; Rhodewalt
& Morf, 1995; Watson & Biderman, 1993; Watson, Hickman, & Morris, 1996). Us-
ing meta-analytic techniques, Campbell (1999) found an average correlation of .29
between narcissism and self-esteem (as measured by the Rosenberg (1965) self-es-
teem scale), using the results of 11 studies with a total of 2963 participants. Many
researchers view high self-esteem as a positive trait (however, see Twenge & Camp-
bell, 2001; for a more tempered evaluation of the benefits of possessing high self-es-
teem). Or perhaps more accurately stated, possessing low self-esteem is frequently
viewed as a negative trait (e.g., Cutrona, 1982; Kanfer & Zeiss, 1983; Leary,
1983). Thus, despite their negative qualities, narcissists do appear to maintain several
positive characteristics, including high self-esteem.
As stated previously, the present investigation sought to answer several questions
related to narcissism, age, and culture. The initial focal point of this investigation
centers upon narcissism and its possible relation to age. Specifically, we wanted to
answer the question: are younger people more narcissistic than older people? One
might predict that narcissism is unlikely to change with age, especially after early
adulthood. Although there is some change in personality during adolescence, a large
body of research finds that personality is very stable, based primarily on analyses of
correlational consistency, after the age of 30 (Costa & McCrae, 1988; Costa, McC-
rae, & Arenberg, 1980; Costa et al., 1986). Some cross-sectional research, on the
other hand, reveals significant mean-level personality change with age (Srivastava
et al., in press). In fact, a recent review of the research documented evidence that per-
sonality fluctuates across the lifespan in theoretically meaningful ways (Helson,
Kwan, John, & Jones, 2002).
Of course, it is possible that personality may demonstrate both correlational con-
sistency and mean-level fluctuations. For example, it is possible that throughout the
lives of two people, person A is always more agreeable than person B, but that they
both become more agreeable as they age. To examine correlational consistency, lon-
gitudinal designs must be employed. Mean-level changes, however, may be investi-
gated using cross-sectional paradigms. Because the primary measure of narcissistic
personality (the focus of the present examination), the NPI (Raskin & Terry,
1988), has only been in use in its current form for a relatively short period of time,
longitudinally based analyses of correlational consistency will have to wait. We can
and we do, however, in the present investigation, assess mean-level fluctuations in
narcissism across the lifespan.
Thus we return to the issue that has generated considerable debate in personality
research: is personality set in plaster after a certain age, or can it change? As for the
personality construct of narcissism, we believe that it does change across the lifespan.
There are at least three reasons to suggest why this assertion might be true: (1) what
J.D. Foster et al. / Journal of Research in Personality 37 (2003) 469–486 471
some clinicians call ‘‘disorder burnout,’’ (2) the incorporation of objective failure
into oneÕs self-concept (which we label the ‘‘reality principle model’’), and (3) cultural
changes that may produce birth cohort effects.
Our first reason to suspect age differences in narcissism is labeled ‘‘disorder burn-
out.’’ Clinicians have noted that certain personality disorders tend to ‘‘mellow’’ with
age. For example, the DSM-IV TR suggests that many of the characteristics associ-
ated with antisocial personality disorder may become less evident as one grows older
(American Psychiatric Association, 2000), and borderline personality disorder shows
a similar pattern. This pattern fits NPD as well; the DSM-IV TR suggests that some
common adolescent behaviors might even be mistaken for NPD. Thus, it seems
likely that narcissism will decrease with age because as people age, some of the char-
acteristics they possess that are associated with narcissism should assuage.
We also speculate that as one grows older one will experience more frequent op-
portunities for failure (and, in fact, fail more frequently), which we label the ‘‘reality
principle model.’’ For example, few children have experience with failing a test or
being objectively compared with others before they enter school. As one progresses
through life, however, the frequency of objective failure should increase simply be-
cause the number of instances when one might objectively succeed or fail increases.
The challenge presented by these encounters also increases. In many ways, adoles-
cence and young adulthood present evaluations with progressively higher stakes: col-
lege admissions, dating experiences, and job interviews may lead people to doubt
their self-esteem in a way they never did as children. Continuing through the life-
span, problems associated with family and general health provide further evidence
on oneÕs lessened ability. We believe that the more failure people experience, the less
narcissistic they are likely to be. Young people have simply not had the opportunity
to experience much failure. Older people have, and thus they should be less narcis-
sistic.
The disorder burnout and reality principle models suggest true age differences in
narcissism. However, age differences in a cross-sectional study might also be due to
birth cohort (Schaie, 1965; Twenge, 2002). Because people born at different times are
exposed to different cultures, birth cohort is a useful proxy for the overall socio-cul-
tural environment (Twenge, 2002). Several pieces of evidence suggest that recent co-
horts and environments might be more narcissistic. Christopher LaschÕs well-known
1979 book, The Culture of Narcissism, argued that the United States became progres-
sively more individualistic and self-focused during the 1970s. Other authors found
that this trend toward individualism was pervasive and long-standing (Baumeister,
1987; Gough, 1991). In addition, college students reported progressively higher
self-esteem between 1968 and 1994 (Twenge & Campbell, 2001), even though objec-
tive measures such as SAT scores and divorce rates would suggest a decrease in self-
esteem. The correlation between self-esteem and narcissism suggests that narcissism
scores might have risen as well (especially because the self-esteem rise was discon-
nected from true improvement or achievement). Thus, younger people (who were
born more recently) grew up in a culture focused more on self-esteem and individu-
alism. Older people, raised in more collectivistic eras (e.g., the 1950s), might be less
likely to harbor narcissistic traits. This again suggests that narcissism should
472 J.D. Foster et al. / Journal of Research in Personality 37 (2003) 469–486
decrease with age, though this model predicts that the difference would lie in birth
cohort rather than in development.
Thus far, we have discussed the reasoning behind our hypothesis that reported
narcissism will decline in older participants. We now continue and present evidence
for our second and third predictions, that narcissism will show world regional and
ethnic differences. A growing body of recent research suggests that self-concept vari-
ables differ among ethnic groups. Gray-Little and Hafdahl (2000) first identified a
self-esteem advantage for Black Americans versus White Americans. This result
was confirmed and extended by Twenge and Crocker (2002) in a meta-analysis of
self-esteem and ethnic identity. Specifically, they tested potential differences in self-
esteem among Whites, Blacks, Hispanics, Asians, and American Indians.
1
They
found what amounts to a self-esteem continuum, with Blacks reporting the highest
self-esteem, followed by Whites. Hispanics, Asians, and American Indians reported
the lowest self-esteem. These results corresponded exactly to the comparative levels
of individualism among these groups (Oyserman, Coon, & Kemmelmeier, 2002),
suggesting that narcissism might show a similar pattern. We therefore predict that
a similar continuum will be evident for narcissism. The basis for this prediction is
the reliable association between narcissism and self-esteem, which should translate
into a result for narcissism that is similar to the result for self-esteem found by
Twenge and Crocker (2002). Further, because individualistic thinking is probably
positively associated with increased narcissism, the fact that the ethnic differences
in self-esteem correspond directly to levels of individualism should mean that narcis-
sism will follow a similar pattern.
This prediction, however, may mask important cultural and regional differences in
narcissism. For example, Fukunishi et al. (1996) found that the Chinese are more
narcissistic than Americans, but the Japanese are less narcissistic than Americans.
Thus even within Asia there appears to be some variance as to how narcissistic peo-
ple are. Plaut et al. (2002) found regional differences in self-description within the
United States. For example, people from the West South Central region (e.g., Texas,
Oklahoma) report being more outspoken and self-confident, compared to other re-
gions of the country. People from the New England area report being particularly
concerned with being softhearted and caring. One might infer from such results, then
that if different regions of the United States produce self-concepts laden with either
self or other-focused ideologies one might expect a construct such as narcissism to
also vary from region to region.
1
Twenge and Crocker (2002) used the terms ‘‘Black’’ to denote Black non-Hispanic, ‘‘White’’ to
denote White non-Hispanic, ‘‘Hispanic’’ to denote people of Hispanic origin, ‘‘Asian’’ to denote people of
Asian or Pacific Island decent, and ‘‘American Indian’’ to denote people of American Indian, Eskimo, or
Aleut decent. We do so as well for simplicity and because the sample we rely upon for the present
investigation includes participants of various ethnic identities, who reside in various countries. Thus, for
example, to label all Black non-Hispanic participants as African-Americans would not accurately reflect
the actual makeup of our sample, which includes Black participants from parts of the world other than the
United States of America. We also use the terms ‘‘ethnic identity’’ and ‘‘ethnicity’’ in place of such
characterizations of racial makeup. We realize, however, that none of these characterizations fully
describes the social-identities of the groups under investigation.
J.D. Foster et al. / Journal of Research in Personality 37 (2003) 469–486 473
There is also a tremendous amount of literature suggesting that self-conceptualiza-
tion varies across international boundaries. Much of this research focuses on dif-
ferences stemming from collectivistic versus individualistic cultures. In a recent
meta-analytic review, Oyserman et al. (2002) reported consistent differences in collec-
tivistic and individualistic orientations when comparing Americans with Europeans,
Asians, Africans, and Middle-Easterners. Other researchers have reported the cross-
cultural effects of these social orientations on behaviors and perceptions. For example,
Kitayama, Markus, and Matsumoto (1997) examined how collectivistic and individu-
alistic culture shapes situational perceptions. American participants were more likely
to identify situations where self-esteem enhancement was likely whereas Japanese par-
ticipants were more likely to identify situations where self-criticism was the likely out-
come. Other research suggests that people from individualistic cultures, in comparison
to people from collectivistic cultures, agree more strongly with self-relevant positive
emotions (Lee, Jones, & Mineyama, 2002), are less modest (Kurman & Sriram,
2002), are more likely to project their own feelings onto others and recall personal
situations from their own perspective as opposed to the perspective of others (Cohen
& Gunz, 2002), are more likely to engage in agentic self-enhancement (Kurman, 2001),
and tend to report well-being as more closely associated with emotions that are
interpersonally distancing (e.g., pride) (Kitayama, Markus, & Kurokawa, 2000).
These findings all point to a clear delineation between collectivistic and individu-
alistic cultures in terms of self-concept and perception. As the research suggests, in-
dividualism encourages greater focus on the self whereas collectivism promotes
greater focus on the group. Thus, individualistic promotion of self-focus over
other-focus should be reflected in greater narcissism being expressed in people from
more individualistic cultures. This leads us to our third prediction: regional world
differences in narcissism will reflect differences in individualism across world regions.
Therefore, we expect to find that people from countries with increased individualism
will also report being more narcissistic.
Our remaining prediction is a replication of previously reported results, which
find that men are more narcissistic than women (e.g., Bushman & Baumeister,
1999; Farwell & Wohlwend-Lloyd, 1998; Joubert, 1998; Ladd et al., 1997). In fact,
Campbell (1999) found an average correlation of .18 (males higher than females) be-
tween gender and narcissism across 20 samples with 3668 participants. Thus, our fi-
nal hypothesis is that male participants in the present investigation will report more
narcissism compared to female participants. The data in the present article, obtained
from a large and diverse sample, should help increase the generalizability of this al-
ready reliable finding.
Previous research has neglected the potentially crucial roles that age and culture
play on narcissism. Thus, our knowledge of potential developmental and cultural in-
fluences is lacking. The present investigation will for the first time demonstrate the
efficacy of these predictors on the personality construct of narcissism, employing a
sample that is larger than any previously found in the narcissism literature. To sum-
marize, the present research sought to test four hypotheses concerning narcissism,
age, and culture. First, we predict that narcissism will be negatively related to age.
Second, we predict that ethnic differences in reported narcissism will fit the same
474 J.D. Foster et al. / Journal of Research in Personality 37 (2003) 469–486
pattern as ethnic differences in self-esteem (Blacks highest, followed by Whites, His-
panics, and then Asians; Twenge & Crocker, 2002). Third, we predict that people
from world regions that are more individualistic will also report being more narcis-
sistic. Finally, we predict that men will be more narcissistic than women. We tested
each prediction using a sample of participants who completed Internet versions of a
narcissism and demographic questionnaire.
2. Method
2.1. Participants
A total of 3445 participants [average age of 24.5 years (SD ¼9:1 years)] partici-
pated in this study. Participant ages ranged from 8 to 83 years, with at least 10 par-
ticipants representing each age from 13 to 50 years and at least 20 participants
representing each age from 14 to 42 years. Consistent with popularly quoted demo-
graphic characteristics of internet users, the majority of the participants ranged in
age from 16 to 26 years, with at least 100 participants representing each of these ages.
Finally, the most popular age range of participants in this study was 17–22 years, with
at least 175 participants representing each of these ages. An examination of age by
gender revealed that males (M¼25:4 years; SD ¼9:2 years) were slightly, but signif-
icantly older than females (M¼24:2 years; SD ¼9:1 years), tð3443Þ¼3:2;p<:01.
Of these participants, approximately 75% were female. Approximately 74% of the
participants were White, though all other sampled ethnicities were represented (7%
Hispanic, 7% Asian, 6% Black, 1% Native American, 1% Middle Eastern, and 4%
‘‘Other’’). Though a majority of the participants reported that they resided in the
United States (74%), participants also reported living in Europe (9%), Canada
(6%), Asia (5%), the Middle East (3%), Africa (2%), South America (1%), and Aus-
tralia (0.2%). More than half of the participants (61%) reported that they earned less
than $20,000 (US$) per year, with 20% reporting that they earned between $20,000
and $40,000 annually, 9% reporting yearly incomes between $40,000 and $60,000,
and 11% reporting that they earned more than $60,000 per year.
2.2. Materials and procedure
An Internet website was constructed for this study (www.psycdawgs.com) that
was advertised at various Internet search engines and a social psychology themed
website. The advertisement solicited potential participants to complete a question-
naire study on personality across the lifespan. All participant data were collected be-
tween July 2001 and June 2002.
When participants visited the website, they were first presented with a consent
form that gave the option to either proceed to the questionnaire or to leave the web-
site. All participants were assured that no attempt would be made to associate any
identifying information (e.g., IP address) with their responses, and no identifying
information was requested in the questionnaire. Following the completion of the
J.D. Foster et al. / Journal of Research in Personality 37 (2003) 469–486 475
questionnaire, participants were directed to a debriefing page that explained the pur-
pose of the study and directed them to one of several appropriate Internet websites
for further information regarding the topic of the investigation. Participants were
also given the opportunity to email the investigators with any questions or concerns.
Narcissism was assessed with the Narcissistic Personality Inventory (NPI; Raskin
& Terry, 1988), which consists of 40 forced-choice items, each containing two alter-
native statements concerning the participant. Examples of the statements include ‘‘I
will be a success’’ (narcissistic) and ‘‘I am not too concerned about success’’ (non-
narcissistic) or ‘‘I find it easy to manipulate people’’ (narcissistic) and ‘‘I donÕt like
it when I find myself manipulating people’’ (non-narcissistic). Total scores on the
NPI can range from 0 to 40 with higher scores indicating increased narcissism. Par-
ticipants in the present study reported a mean NPI score of 15.2 (SD ¼6:7). Also as-
sessed in the questionnaire were age, gender, ethnicity, income level (in $US), and the
world region where the participant resided. Income level, specifically, was assessed in
increments of US$20,000 increments with options ranging from under $20,000 to
more than $200,000 per annum.
3. Results
Is age related to narcissism? Recall that we anticipated a negative correlation be-
tween the two, hypothesizing that younger people would be more narcissistic than
older people. We initially examined age differences by performing a simple regression
analysis with age predicting NPI scores. The result revealed a significant negative
correlation, r¼:17;p<:001 (d¼:35), revealing that older participants reported
being less narcissistic than did younger participants. Because both age
(r¼:30;p<:001) and NPI (r¼:08;p<:001) were significantly correlated with re-
ported annual income, we added income to the equation to assess the unique contri-
bution of age. This increased the correlation between narcissism and age,
b¼:22;p<:001 (d¼:42). We then added gender to the equation, which resulted
in a nearly identical correlation between narcissism and age, b¼:22;p<:001
(d¼:43). Thus, our primary hypothesis, that age would be negatively related to nar-
cissism was solidly confirmed by our data, particularly after we controlled for con-
founding variables.
2
As can be seen in Fig. 1, narcissism appears to be a personality
trait that fluctuates a great deal depending on the age of the participant. In fact, the
difference in reported NPI scores from our youngest participants to our oldest par-
ticipants is nearly a full standard deviation.
Next, we compared NPI scores among the four ethnic identity groups with at least
100 participants: Whites (n¼2564), Blacks (n¼222), Asians (n¼237), and Hispan-
ics (n¼230).
3
There were significant differences among the four ethnic identity
2
We also tested for potential curvilinear relations between age and narcissism. However, the results
suggested that the pattern of association was linear in nature.
3
Twenge and Crocker (2002) also compared a sample of Native Americans. In our investigation,
however, the Native American sample was too small to make any comparison feasible.
476 J.D. Foster et al. / Journal of Research in Personality 37 (2003) 469–486
groups, Fð3;3249Þ¼9:4;p<:001 (g2
p¼:01). A set of comparisons revealed that
both White participants (M¼14:8;SD ¼6:8) and Asian participants
(M¼14:9;SD ¼6:5) reported being slightly, but significantly (p’s<:05) less narcis-
sistic than both Black participants (M¼16:7;SD ¼6:5) and Hispanic participants
(M¼16:5;SD ¼6:4). No other ethnicity differences were revealed. These results re-
flect to some degree the pattern for ethnicity differences found in self-esteem re-
search: in the self-esteem literature, Asians generally report the lowest self-esteem
and Hispanics report lower self-esteem than do Whites, with Blacks reporting the
highest self-esteem. Consistent with this, Blacks reported the highest levels of narcis-
sism, though not significantly more than Hispanics. Whites, however, reported less
narcissism than did Hispanics and equivalent narcissism with Asians.
Black and Hispanic participants reported slightly lower ages and income levels
compared to White and Asian participants, r’s¼:05 and ).04, p’s<:01 and .05,
respectively.
4
We controlled for these differences in a regression analysis that
Fig. 1. The negative association between narcissism and age is shown. Ages are grouped into five year
intervals beginning at less than 15 years of age and ending with more than 59 years of age. At each age
interval, the mean NPI score is calculated and plotted. Error bars represent +/)1 standard error of the
mean. Also reported is the number of total participants in each age grouping.
4
For this set of zero-order correlation and regression analyses, we coded all participants who reported
to be either White or Asian as ‘‘)1’’ and all participants who reported to be either Black or Hispanic as
‘‘1.’’ Thus, a positive association between this ethnic variable and narcissism would indicate greater
reported narcissism for Black/Hispanic participants compared to White/Asian participants.
J.D. Foster et al. / Journal of Research in Personality 37 (2003) 469–486 477
compared NPI scores for White/Asian participants versus Black/Hispanic partici-
pants. The result revealed a small, yet significant difference between the two ethnicity
groupings even after reported age and income level were controlled, b¼:09;p<:001
(d¼:19). Finally, we examined any potential interaction between ethnic identity
group and gender, but found none. In summary, the results for narcissism were gen-
erally comparable to those found by self-esteem researchers (e.g., Gray-Little & Haf-
dahl, 2000; Twenge & Crocker, 2002) with a few notable exceptions.
We then examined whether NPI scores were associated with world region. Recall
that we predicted that people from world regions that are more individualistic would
report more narcissism compared to people from world regions that are less individ-
ualistic. We examined NPI scores for participants from each of five world regions
with at least 100 participants: the United States (n¼2546), Canada (n¼190), Eu-
rope (n¼304), Asia (n¼162), and the Middle East (n¼104). The result revealed
a significant difference among the five regions, Fð4;3301Þ¼2:7;p<:05 (g2
p¼
:003). An examination of the mean NPI scores revealed that participants from the
United States produced the highest levels of reported narcissism with an average
NPI score of 15.3 (SD ¼6:8). This was followed by Europe [M¼15:0(SD ¼6:3)],
Canada [M¼14:8(SD ¼6:9)], Asia [M¼14:3(SD ¼6:7)], and the Middle East
[M¼13:9(SD ¼6:9)]. Statistical comparisons revealed that participants from the
United States reported significantly (i.e., p’s<:05) greater narcissism compared to
participants from either Asia or the Middle East. We also tested for any potential
interaction between world region and gender, but found none (i.e., F<1).
To more directly test the influence of individualism on narcissism, we grouped
together participants from the United States, Canada, and Europe to create a
group of participants that we refer to as high individualists (n¼2898). We also cre-
ated a group, which we refer to as low individualists, consisting of participants
from Asia and the Middle East (n¼408). The decision of whether to include each
region in the ‘‘high individualist’’ or ‘‘low individualist’’ group was based on inter-
national individualism score data reported by Diener and Diener (1995). We com-
pared these two groups in terms of NPI scores and found that they differed
significantly from one another, r¼:05;p<:01 (d¼:11). The high individualist
group reported an average NPI score of 15.3 (SD ¼6:7) compared to 14.2
(SD ¼6:7) for the low individualistic group. We then assessed whether gender inter-
acted with our individualism variable and found that it did not (i.e., F¼1:2).
Therefore, the result for world region falls in line with our general prediction that
narcissism would be associated positively with individualism. World regions that
generally display higher individualism also reported somewhat greater narcissism.
Are males more narcissistic than females? We compared NPI scores of male and
female participants. Males reported a significantly higher average NPI score [16.5
(SD ¼6:9)] compared to the female average of 14.7 (SD ¼6:6), r¼:12;p<:001
(d¼:24). Because gender and income level were significantly related, r¼
:05;p<:01 (males reported earning higher incomes), reported income was con-
trolled in a second regression model. The result revealed that males still reported
greater narcissism compared to females, b¼:12;p<:001 (d¼:23). Finally,
gender was also significantly related to age, r¼:06;p<:01 (males reported being
478 J.D. Foster et al. / Journal of Research in Personality 37 (2003) 469–486
older than females). Thus, we also controlled for age (in addition to income) in a
third regression model. The result, once again, revealed that males reported greater
narcissism than did females, b¼:12;p<:001 (d¼:26). This result is consistent
with our prediction and served to replicate the general finding that males report be-
ing somewhat more narcissistic than females. Further, our result suggests that even
when age and reported income levels are simultaneously controlled, the gender
difference remains.
An intriguing possibility is that some of the NPI subscales might have influenced
the reported associations more than others. For example, vanity scores may be much
lower among older individuals, and that may lead to the lower overall NPI compos-
ite scores. To examine this possibility, we regressed each of the seven NPI subscales
onto the predictor variables of interest (along with any variables that we controlled
for in the previous analyses). Table 1 shows a summary of these analyses. When we
assessed age (controlling for gender and income) in relation to the seven NPI sub-
scales, the results were all consistent with those that used the composite measure.
That is, all subscales of the NPI decrease with age. Some of the subscales, however,
were more highly correlated with age than others. Specifically, exhibitionism, explo-
itativeness, vanity, and entitlement were most highly related to age. Self-sufficiency,
in particular, demonstrated a relatively weak relation to age. Authority and superi-
ority were more moderately related to age.
We next assessed how ethnic identity related to the NPI subscales. Specifically, we
compared Black and Hispanic participants to White and Asian participants. The re-
sults were again mostly consistent with what we reported when using the composite
measure. The ethnic identity groups differed on five of the seven subscales (i.e., Black
and Hispanic participants reported more authority, superiority, entitlement, self-suf-
ficiency, and vanity than did White and Asian participants). They did not differ in
terms of exhibitionism or exploitativeness. It is also important to note that these final
Table 1
Age Ethnic identity (1) World region (2) Gender (3)
NPI Composite ).22.09.05).12
NPI Subscales
Authority ).10.09.07).09
Exhibitionism ).21).02 ).02 ).02
Superiority ).11.07.07).06
Entitlement ).15.08.01 ).16
Exploitativeness ).18.03 .02 ).11
Self-sufficiency ).07.07.08).08
Vanity ).16.09.03 .01
Notes. Numbers represent standardized regression coefficients.
(1) Ethnic identity coded so that positive association indicates that Black/Hispanic participants higher
than White/Asian participants.
(2) World region coded so that positive association indicates that individualistic country higher than
collectivistic country.
(3) Gender coded so that negative association indicates that males higher than females.
*
Significantly (i.e., p<:05) different from zero.
J.D. Foster et al. / Journal of Research in Personality 37 (2003) 469–486 479
two subscales of the NPI are often considered to represent the elements most detri-
mental to interpersonal functioning.
The results for world region (specifically, individualistic compared to collectivistic
cultures) were somewhat less consistent. Significant differences were found on three
of the seven NPI subscales. Each of these three differences were in the hypothesized
direction. Specifically, participants from more individualistic countries reported
greater authority, superiority, and self-sufficiency. The two cultures did not differ
in terms of exhibitionism, entitlement, exploitativeness, or vanity.
Finally, the results for gender were fairly consistent, regardless of whether we as-
sessed the NPI composite or subscales. Males reported greater authority, superiority,
entitlement, exploitativeness, and self-sufficiency. The two genders did not differ sig-
nificantly in terms of exhibitionism or vanity.
In summary, whether we correlated our relevant predictors with the NPI compos-
ite or its subscales, the results were generally in agreement. With the exceptions
noted above, both overall narcissism and most of its components were correlated
with age, ethnic identity, world region, and gender.
4. Discussion
This research examined associations between narcissism and age, gender, ethnic
identity, and world region using a large sample of participants of many ages and
from around the world. As predicted, narcissism was negatively related to age, espe-
cially after income level and gender were controlled. Further, we found ethnic differ-
ences in narcissism somewhat similar to those in self-esteem. These differences
remained even after reported income level and age were controlled. We also found
evidence that narcissism varies across world region, with Americans reporting the
highest levels of narcissism, followed by Europeans, Canadians, Asians, and Middle
Easterners. By grouping the world regions into those who display greater individu-
alism compared to those who display less individualism we found that higher indi-
vidualism was associated with more reported narcissism. Finally, males reported
being more narcissistic than females even after age and income level were controlled.
Below, we discuss potential explanations for the above results and how they relate to
previous research and theory.
4.1. Narcissism and age
We found that younger people report more narcissism than older people. This
finding is consistent with several of the theories we discussed in the introduction.
Both the disorder burnout model and the reality principle model suggest that age
should be negatively related to narcissism. Generational differences also predicted
more narcissism in younger people, who grew up in eras characterized by high indi-
vidualism and emergent narcissism (Lasch, 1979; Twenge & Campbell, 2001). As a
cross-sectional study, however, our data cannot pinpoint the exact cause of the cor-
relation between narcissism and age: it could be due to either lifespan development
480 J.D. Foster et al. / Journal of Research in Personality 37 (2003) 469–486
or to birth cohort/culture. Future studies should explore this issue, possibly with the
use of longitudinal designs (which would pinpoint developmental changes) or cross-
temporal meta-analysis, (which would pinpoint birth cohort differences; e.g., Twenge
& Campbell, 2001). Unfortunately, the (40-item) NPI was first published in 1988,
making it difficult to address changes across many birth cohorts.
Our cross-sectional results are consistent with the findings of some researchers
that many personality constructs appear to operate in a rather fluid manner across
the lifespan, even beyond the age of 30 (see Helson et al., 2002; for review). As seen
in Fig. 1, the drop in mean level narcissism did not cease at the age of 30. In fact, it
appears to drop well into later adulthood. Of course, cross-sectional investigations,
such as the present, cannot untangle the relationship between narcissism and age. Is
the relationship due to culture or development? Though we know that narcissism ap-
pears to diminish with age, is the maintenance of individual differences preserved?
We simply do not know at the present time. However, the results of this, as well
as several other investigations, might challenge the notion that personality is essen-
tially set in plaster beyond a certain age. At the very least, they provide the impetus
for further examination of the role that age plays in regard to personality constructs,
such as narcissism.
It should be noted that the correlation between age and narcissism, though reli-
able, is small. Age explains approximately 4% of the variance in reported narcissism,
not a large amount by any measure (although it is about the same size as the age ef-
fect for self-esteem, which is between 2% and 6%; Twenge & Campbell, 2001). How-
ever, as Helson et al. (2002) alluded to, the ability of age to predict personality
fluctuation at all is important given the enormous number other contributing factors.
We therefore suggest that age is an important, yet one of many potential factors that
reliably predict reported variations in narcissism.
4.2. Narcissism and culture
The differences in reported narcissism levels among the contrasted ethnic identi-
ties are similar to those in self-esteem (i.e., Gray-Little & Hafdahl, 2000; Twenge
& Crocker, 2002). Though our results were not exactly the same, we did find
evidence to support a similar association between narcissism and ethnic identity. Re-
sults for both narcissism and self-esteem show that Blacks score the highest and
Asians either the lowest or near the lowest. The largest discrepancy between our
results and those of self-esteem researchers is for Hispanic participants, who reported
greater narcissism than Whites; the self-esteem analyses found that Hispanics scored
slightly lower than Whites on self-esteem measures. Overall, however, the similarity
of these results to those for self-esteem suggests that the two concepts share similar
causes and roots. Twenge and Crocker (2002) concluded, that race differences in
self-esteem stem primarily from cultural conceptions of individualism. Although
the precise antecedents of self-esteem and narcissism are not known, the race and
birth cohort differences in each imply a strong link with individualism. Cultures that
emphasize individualism, independence, and ‘‘standing out from the crowd’’ are
likely to produce more narcissism among their members. Our results showed the
J.D. Foster et al. / Journal of Research in Personality 37 (2003) 469–486 481
highest reported narcissism among Blacks and more recent birth cohorts, both
groups that display higher levels of individualism (Gough, 1991; Oyserman et al.,
2002; Twenge & Campbell, 2001).
Keeping in line with the general conclusion that individualism is associated with
narcissism, we found that world regions that generally display greater individualism
also tend to display more narcissism. Indeed, when we contrasted the five world re-
gions with one another, the only significant differences that emerged suggest that the
United States displays significantly more narcissism than does either Asia or the
Middle East. This result is consistent with the level of individualism displayed by
these world regions (see Diener & Diener, 1995). The United States is generally con-
sidered to be one the most individualistic nations while countries within Asia and the
Middle East are relatively low in individualism. Therefore, it is not surprising that
participants from the United States also reported more narcissism than participants
from either Asia or the Middle East. Further, after grouping the world regions into
either high or low individualism categories, we found that regional individualism dif-
ferences appear related to regional differences in reported narcissism, with high indi-
vidualism associated with greater reported narcissism. This result provides further
evidence that higher levels of individualism appear to be associated with higher levels
of narcissism.
Finally, we were able to report further evidence that men report more narcissism
than women. This difference has been documented in prior research. However, this
sample is larger and more inclusive than any previously used and thus can increase
confidence in the direction and magnitude of the gender difference. Because of the
large and inclusive sample utilized in the present investigation, the result should per-
haps be given more credence. Previous research (see Cross & Madson, 1997; for re-
view) suggests that males conceptualize the self more in terms of independence
whereas females tend to be more interdependent. Therefore, it is possible that the dif-
ference in narcissism between the genders reflect another difference in self-concept
with males being more individualistic and thus also more narcissistic.
4.3. General limitations and conclusions
It should first be reiterated that the associations reported (especially those related
to ethnicity and world region) were quite small. Therefore, although theoretically
meaningful, these results should not be interpreted as diagnostic in nature or as pos-
sessing significant predictive utility. In other words, the within-group variance is
vastly larger than any between-group differences in narcissism. Just as the normal
curves of most social psychological variables greatly overlap when one compares
groups of people, those associated with narcissism are no different. One would be
na€
ııve to assume anything more than this on the basis of the present investigation.
As with all correlational research, no causal inferences can be made about the re-
sults reported in this article. However, we did not intend to provide causal evidence
of the links among narcissism, age, and culture. Instead, we sought to report rela-
tions among these variables that stemmed from a sample that is larger and more in-
clusive than those used in prior investigations. Future research should examine these
482 J.D. Foster et al. / Journal of Research in Personality 37 (2003) 469–486
relations to determine specific factors associated with age and culture that lead to
variations in reported narcissism.
In general, one might question whether Internet respondents answer question-
naires honestly. However, this concern is present in the traditional survey method
as well, and there is no reason to believe that people are more dishonest when com-
pleting an Internet questionnaire.
Another concern is the representativeness of a sample collected on the Internet.
Recent studies have reviewed the literature and concluded that these samples are just
as representative as college undergraduate samples, and possibly more representative
(Srivastava et al., in press; Robins et al., 2002; Williams, Cheung, & Choi, 2000). We
would be remiss, however, if we did not acknowledge the possibility that self-selec-
tion might have biased the results of some cultural groups more than others (this is
especially true with regard to our cross-nationality comparisons). Our sample,
though larger and more inclusive than most previously reported, still largely consists
of white females from the United States who are college aged and middle-class. It is
possible that those participants who do not fit into this demographic are less repre-
sentative of their population. It would be advisable for future studies to investigate
the issues outlined in this paper by assessing larger, potentially more representative
samples of participants from multiple cultures. Studies that address cross-cultural
differences in narcissism by administering the NPI to randomly selected samples will
likely shed valuable insight into the nature of these potential differences.
With regard to the present investigation, it might speak to the generalizability of
our results to compare the results we have attained with an Internet sample to those
attained when using more conventional sampling methods. We are not aware of any
collection of studies that specifically addressed the relatedness of narcissism, age, eth-
nicity, or world region. The vast majority of research on narcissism to date has been
conducted on college student populations in the US. The data in the current research
literature simply do not offer age or cultural comparisons of narcissism. There have
been, however, several studies that have reported the link between narcissism and
gender (at least within this limited demographic). As stated in the introduction of
this paper, Campbell (1999), using meta-analytic techniques (20 independent sam-
ples, over 3500 participants), found an average correlation between narcissism and
gender of .18 (males reported being more narcissistic). Our current study yielded a
correlation of .12 (again, males reporting higher narcissism). Though not precisely
identical, the present correlation is close and falls well within the range of those pre-
viously reported (rÕs ranged from .00 to .32). This perhaps lends more confidence to
the generalizability of the results reported herein, reducing the concerns associated
with self-selection biases.
In conclusion, the present investigation demonstrates how age and culture are
linked to narcissism. Although we cannot pinpoint the exact causes of these rela-
tions, differences in individualism across continent, ethnicity, birth cohort, and gen-
der may provide the most parsimonious answer. Focusing on the self may lead to
high self-esteem and high psychological well-being, but it can also lead to narcissism
under certain conditions. Understanding this potential pitfall to unconditional posi-
tive regard for the self may shed light on sociological trends such as increasing
J.D. Foster et al. / Journal of Research in Personality 37 (2003) 469–486 483
divorce rates and the recent influx of high profile violent acts. Perhaps future re-
search can demonstrate how people can develop positive self-concepts without also
displaying the negative behaviors associated with narcissism.
References
American Psychiatric Association. (2000). Diagnostic and statistical manual (4th ed.). Washington, DC:
Text Revision.
Baumeister, R. F. (1987). How the self became a problem: A psychological review of historical research.
Journal of Personality and Social Psychology, 52, 163–176.
Bradlee, P. M., & Emmons, R. A. (1992). Locating narcissism within the interpersonal circumplex and the
five-factor model. Personality and Individual Differences, 13, 821–830.
Bushman, B. J., & Baumeister, R. F. (1999). Threatened egotism, narcissism, self-esteem, and direct and
displaced aggression: Does self-love or self-hate lead to violence? Journal of Personality and Social
Psychology, 76, 367–376.
Campbell, W. K. (1999). Narcissism in everyday life. Unpublished manuscript, Case Western Reserve
University.
Cohen, D., & Gunz, A. (2002). As seen by the other... Perspectives on the self in the memories and
emotional perceptions of Easterners and Westerners. Psychological Science, 13, 55–59.
Costa, P. T., & McCrae, R. R. (1988). Personality in adulthood: A six-year longitudinal study of self-
reports and spouse ratings on the NEO Personality Inventory. Journal of Personality and Social
Psychology, 54, 853–863.
Costa, P. T., McCrae, R. R., & Arenberg, P. (1980). Enduring dispositions in adult males. Journal of
Personality and Social Psychology, 38, 793–800.
Costa, P. T., McCrae, R. R., Zonderman, A. B., Barbano, H. E., Lebowitz, B., & Larson, D. M. (1986).
Cross-sectional studies of personality in a national sample: 2. Stability in neuroticism, extraversion,
and openness. Psychology and Aging, 1, 144–149.
Cross, S. E., & Madson, L. (1997). Models of the self: Self-construals and gender. Psychological Bulletin,
122, 5–37.
Cutrona, C. E. (1982). Transition to college: Loneliness and the process of social adjustment. In L.
A. Peplau, & D. Perlman (Eds.), Loneliness: A sourcebook of current theory, research, and therapy
(pp. 291–309). New York: Wiley.
Diener, E., & Diener, M. (1995). Cross-cultural correlates of life satisfaction and self-esteem. Journal of
Personality and Social Psychology, 68, 653–663.
Emmons, R. A. (1984). Factor analysis and construct validity of the narcissistic personality inventory.
Journal of Personality Assessment, 48, 291–300.
Farwell, L., & Wohlwend-Lloyd, R. (1998). Narcissistic processes: Optimistic expectations, favorable self-
evaluations, and self-enhancing attribution. Journal of Personality, 66, 65–83.
Fukunishi, I., Nakagawa, T., Nakamura, H., Li, K., Hua, Z. Q., & Kratz, T. S. (1996). Relationships
between Type A behavior, narcissism, and maternal closeness for college students in Japan, the United
States of America, and the PeopleÕs Republic of China. Psychological Reports, 78, 939–944.
Gough, H. (1991). Scales and combinations of scales: What do they tell us, what do they mean? Paper
presented at the 99th Annual Convention of the American Psychological Association, San Francisco,
August 1991.
Gray-Little, B., & Hafdahl, A. R. (2000). Factors influencing racial comparisons of self-esteem:
A quantitative review. Psychological Bulletin, 126, 26–54.
Heine, S. J., & Lehman, D. R. (1997). The cultural construction of self-enhancement: An examination of
group-serving biases. Journal of Personality and Social Psychology, 72, 1268–1283.
Helson, R. H., Kwan, V. S. Y., John, O. P., & Jones, C. (2002). The growing evidence of personality
change in adulthood: Findings from research with personality inventories. Journal of Research in
Personality, 36, 287–306.
484 J.D. Foster et al. / Journal of Research in Personality 37 (2003) 469–486
Jackson, L. A., Ervin, K. S., & Hodge, C. N. (1992). Narcissism and body image. Journal of Research in
Personality, 26, 357–370.
Joubert, C. E. (1998). Narcissism, need for power and social interest. Psychological Reports, 82, 701–702.
Kanfer, R., & Zeiss, A. M. (1983). Depression, interpersonal standard setting, and judgments of self-
efficacy. Journal of Abnormal Psychology, 92, 319–329.
Kasser, T., & Ryan, R. M. (1996). Further examining the American dream: Differential correlates of
intrinsic and extrinsic goals. Personality and Social Psychology Bulletin, 22, 280–297.
Kitayama, S., Markus, H. R., & Kurokawa, M. (2000). Culture, emotion, and well-being: Good feelings in
Japan and the United States. Cognition and Emotion, 14, 93–124.
Kitayama, S., Markus, H. R., & Matsumoto, H. (1997). Individual and collective processes in the
construction of the self: Self-enhancement in the United States and self-criticism in Japan. Journal of
Personality and Social Psychology, 72, 1245–1267.
Kurman, J. (2001). Self-enhancement: Is it restricted to individualistic cultures? Personality and Social
Psychology Bulletin, 27, 1705–1716.
Kurman, J., & Sriram, N. (2002). Interrelationships among vertical and horizontal collectivism, modesty,
and self-enhancement. Journal of Cross-Cultural Psychology, 33, 71–86.
Ladd, E. R., Welsh, M. C., Vitulli, W. F., Labbe, E. E., & Law, J. L. (1997). Narcissism and causal
attribution. Psychological Reports, 80, 721–722.
Lasch, C. (1979). The culture of narcissism: American life in an age of diminishing expectations. NY: Norton.
Lee, J. W., Jones, P. S., & Mineyama, Y. (2002). Cultural differences in response to a Likert scale.
Research in Nursing and Health, 25, 295–306.
Leary, M. R. (1983). Understanding social anxiety: Social, personality, and clinical perspectives. Berverly
Hills: Sage.
Markus, H. R., & Kitayama, S. (1991). Culture and the self: Implications for cognition, emotion, and
motivation. Psychological Review, 98, 224–253.
Oyserman, D., Coon, H. M., & Kemmelmeier, M. (2002). Rethinking individualism and collectivism:
Evaluation of theoretical assumptions and meta-analyses. Psychological Bulletin, 128, 3–72.
Ozer, D. J., & Gjerde, P. F. (1989). Patterns of personality consistency and change from childhood
through adolescence. Journal of Personality, 57, 483–507.
Paulhus, D. L. (1998). Interpersonal and intrapsychic adaptiveness of trait self-enhancement: A mixed
blessing? Journal of Personality and Social Psychology, 74, 1197–1208.
Plaut, V. C., Markus, H. R., & Lachman, M. E. (2002). Place matters: Consensual features and regional
variations in America well-being and self. Journal of Personality and Social Psychology, 83, 160–184.
Raskin, R. N., Novacek, J., & Hogan, R. (1991). Narcissism, self-esteem, and defensive self-enhancement.
Journal of Personality, 59, 19–38.
Raskin, R. N., & Terry, H. (1988). A principle components analysis of the Narcissistic Personality
Inventory and further evidence of its construct validity. Journal of Personality and Social Psychology,
54, 890–902.
Rhodewalt, F., & Morf, C. C. (1995). Self and interpersonal correlates of the narcissistic personality
inventory. Journal of Research in Personality, 29, 1–23.
Robins, R. W., Trzniewski, K. H., Tracy, J. L., Gosling, P. D., & Potter, J. (2002). Global self-esteem
across the lifespan. Psychology and Aging, 17, 423–434.
Rosenberg, M. (1965). Society and the adolescent self-image. Princeton, NJ: Princeton University Press.
Schaie, K. W. (1965). A general model for the study of developmental problems. Psychological Bulletin, 64,
92–107.
Srivastava, S., John, O. P., Gosling, S. D., & Potter, J. (in press). Development of personality in
adulthood: Set like plaster or persistent change? Journal of Personality and Social Psychology.
Twenge, J. M. (2002). Birth cohort, social change, and personality: The interplay of dysphoria and
individualism in the 20th century. In D. Cervone, & W. Mischel (Eds.), Advances in Personality Science
(pp. 196–218). New York: Guilford.
Twenge, J. M., & Crocker, J. (2002). Race and self-esteem: Meta-analyses comparing Whites, Blacks,
Hispanics, Asians, and American Indians and Comment on Gray-Little and Hafdahl (2000).
Psychological Bulletin, 128, 371–408.
J.D. Foster et al. / Journal of Research in Personality 37 (2003) 469–486 485
Twenge, J. M., & Campbell, W. K. (2001). Age and birth cohort differences in self-esteem: A cross-
temporal meta-analysis. Personality and Social Psychology Review, 5, 321–344.
Watson, P. J., & Biderman, M. D. (1993). Narcissistic Personality Inventory factors, splitting, and self-
consciousness. Journal of Personality Assessment, 61, 41–57.
Watson, P. J., Hickman, S. E., & Morris, R. J. (1996). Self-reported narcissism and shame: Testing the
defensive self-esteem and continuum hypotheses. Personality and Individual Differences, 21, 253–259.
Williams, K. D., Cheung, C. K. T., & Choi, W. (2000). CyberOstracism: Effects of being ignored over the
Internet. Journal of Personality and Social Psychology, 79, 748–762.
486 J.D. Foster et al. / Journal of Research in Personality 37 (2003) 469–486