Compassionate Liberals and Polite Conservatives: Associations of Agreeableness With Political Ideology and Moral Values

Abstract

Political conservatism has been characterized by resistance to change and acceptance of inequality, with liberalism characterized by the polar opposite of these values. Political attitudes are heritable and may be influenced by basic personality traits. In previous research, conservatism (vs. liberalism) has been associated positively with Conscientiousness and negatively with Openness-Intellect, consistent with the association of conservatism with resistance to change. Less clear, however, are the personality traits relating to egalitarianism. In two studies, using a personality model that divides each of the Big Five into two aspects, the present research found that one aspect of Agreeableness (Compassion) was associated with liberalism and egalitarianism, whereas the other (Politeness) was associated with conservatism and traditionalism. In addition, conservatism and moral traditionalism were positively associated with the Orderliness aspect of Conscientiousness and negatively with Openness-Intellect. These findings contribute to a more nuanced understanding of personality's relation to political attitudes and values.

Full text

Article
Personality and Social
Psychology Bulletin
36(5) 655 –664
© 2010 by the Society for Personality
and Social Psychology, Inc
Reprints and permission:
sagepub.com/journalsPermissions.nav
DOI: 10.1177/0146167210366854
http://pspb.sagepub.com
Compassionate Liberals and Polite
Conservatives: Associations of
Agreeableness With Political
Ideology and Moral Values
Jacob B. Hirsh1, Colin G. DeYoung2,
Xiaowen Xu1, and Jordan B. Peterson1
Abstract
Political conservatism has been characterized by resistance to change and acceptance of inequality, with liberalism characterized
by the polar opposite of these values. Political attitudes are heritable and may be influenced by basic personality traits. In
previous research, conservatism (vs. liberalism) has been associated positively with Conscientiousness and negatively with
Openness-Intellect, consistent with the association of conservatism with resistance to change. Less clear, however, are the
personality traits relating to egalitarianism. In two studies, using a personality model that divides each of the Big Five into
two aspects, the present research found that one aspect of Agreeableness (Compassion) was associated with liberalism and
egalitarianism, whereas the other (Politeness) was associated with conservatism and traditionalism. In addition, conservatism
and moral traditionalism were positively associated with the Orderliness aspect of Conscientiousness and negatively with
Openness-Intellect. These findings contribute to a more nuanced understanding of personality’s relation to political attitudes
and values.
Keywords
personality, politics, morality, conservatism, liberalism
Received January 10, 2009; revision accepted September 6, 2009
Although psychologists have long been interested in study-
ing political attitudes and orientations (Adorno, Frenkel-
Brunswik, Levinson, & Sanford, 1950; Eysenck, 1954;
Rokeach, 1973), there has recently been a renewed interest in
studying political behavior from social-cognitive and motiva-
tional perspectives (Jost, 2006; Jost, Glaser, Kruglanski, &
Sulloway, 2003; Jost, Nosek, & Gosling, 2008). These frame-
works argue that the adoption of different political beliefs
satisfies a variety of motivational needs. In particular, politi-
cal conservatism is thought to be a belief system predicated on
resistance to change and the acceptance of inequality, strate-
gies that serve as defenses against the experience of threat
and uncertainty (Jost et al., 2007). Support for this model has
come from both correlational research, in which conservative
beliefs are associated with measures of uncertainty avoid-
ance and acceptance of inequality, and experimental manip-
ulations, in which increasing the salience of a threat leads to
greater support of conservative values (Jost, Fitzsimons, &
Kay, 2004; Landau et al., 2004).
An important feature of these models is that political beliefs
are derived from deeper psychological needs, which suggests
that individuals may be predisposed by their personalities
to adopt particular ideological perspectives. Findings that
political attitudes are heritable and thus genetically influ-
enced (Bouchard et al., 2003; Koenig & Bouchard, 2006)
highlight the possibility of a connection to basic traits. Consis-
tent with this possibility, research has demonstrated that con-
servatives tend to be higher in trait Conscientiousness, whereas
liberals are higher in trait Openness-Intellect (Carney, Jost,
Gosling, & Potter, 2008; Goldberg & Rosolack, 1994; Jost,
2006). Although this combination of personality traits clearly
relates to a preference for tradition, order, and stability
(McCrae & Costa, 1997; Roberts, Chernyshenko, Stark, &
Goldberg, 2005), it fails to explain the second core aspect of
1University of Toronto, Toronto, Ontario, Canada
2University of Minnesota, Minneapolis, MN, USA
Corresponding Author:
Jacob B. Hirsh, University of Toronto, Department of Psychology,
4th Floor, Sidney Smith Hall, 100 St. George Street, Toronto, Ontario,
Canada, M5S 3G3
Email: jacob.hirsh@utoronto.ca

656 Personality and Social Psychology Bulletin 36(5)
conservative ideology, namely, the acceptance of inequality.
It has been argued that conservatives support inequality sim-
ply because it forms part of the status quo and any change
would lead to instability. Although this is a plausible argu-
ment, research suggests that a preference for egalitarian out-
comes might be a motivational system in its own right
(Moskowitz, Gollwitzer, Wasel, & Schaal, 1999). In particu-
lar, individual differences in empathy, communal goals, and
the acceptance of outgroups all appear related to the person-
ality trait of Agreeableness (Ashton, Paunonen, Helmes, &
Jackson, 1998; Graziano, Bruce, Sheese, & Tobin, 2007;
Graziano, Habashi, Sheese, & Tobin, 2007; Graziano &
Tobin, 2002; Nettle, 2007). Furthermore, social dominance
orientation, which predicts a variety of antiegalitarian atti-
tudes, has been related to a more disagreeable disposition
(Lippa & Arad, 1999). It may thus seem a conspicuous
absence that Agreeableness has not emerged as a significant
predictor of political beliefs, despite the fact that attitudes
toward equality are thought to be a core aspect of conserva-
tive versus liberal ideologies (Jost, 2006; Jost et al., 2003).
There are at least two potential explanations for the
abs ence of Agreeableness in models of political belief, neither
of which precludes the other. The first is that the personality
trait of Agreeableness may contain elements that are differ-
entially related to conservative and liberal worldviews. Spe-
cifically, Agreeableness appears to be divisible into subtraits
of Compassion and Politeness (DeYoung, Quilty, & Peterson,
2007). Compassion appears to index the components of Agree-
ableness most closely linked to empathy and interpersonal
concern. Politeness, by contrast, appears to reflect the com-
ponents of Agreeableness that are more closely linked to
norm compliance and traditionalism. It is thus possible that
Compassion would relate to the liberal emphasis on fairness
and equality, whereas Politeness would relate to the conser-
vative emphasis on order and traditionalism.
A second potential explanation for the lack of reported
correlation between Agreeableness and political beliefs is
that it is a consequence of measuring liberalism and conser-
vatism as opposite ends of a single psychological dimension.
Although the unidimensional view is the most commonly
employed, a number of researchers have argued that it does
not accurately reflect the underlying structure of political
psychology (Conover & Feldman, 1981). Alternative models
propose that political values are better represented by two or
more separate psychological dimensions, such as “radicalism”
and “tender-mindedness” (Eysenck, 1954, 1975) or “freedom”
and “equality” (Rokeach, 1973). Such models argue that global
political beliefs emerge from the interaction of multiple
motivational and dispositional systems. Interestingly, there
appears to be a conceptual relation, on one hand, of radicalism
and freedom with a reduced emphasis on order and tradition
and, on the other, of tender-mindedness and equality with a
preference for egalitarian values. It is also worth noting that
similar two-factor models have been used to describe the
motivational bases of prejudiced social attitudes, as captured
in the separate measures of Right Wing Authoritarianism
and Social Dominance Orientation (Duckitt, Wagner, Du
Plessis, & Birum, 2002).
Researchers interested in moral psychology have simi-
larly argued for the existence of distinct motivational systems
underlying different moral values and political affiliations,
such that liberalism emerges out of psychological systems
that are distinct from those that give rise to conservatism.
According to moral foundations theory (Graham, Haidt, &
Nosek, 2009; Haidt & Graham, 2007; Haidt & Joseph, 2007),
human morality can be summarized by five major domains.
Importantly, research using this model has demonstrated that
the relative importance of these domains varies between lib-
erals and conservatives. In particular, liberals tend to be
more concerned about compassion and justice, whereas con-
servatives are more concerned about ingroup loyalty, respect
for authority, and purity. Differences in the strength of these
underlying motivational systems are thought to influence
explicit political attitudes and ideologies.
Multidimensional models such as these imply that indi-
viduals can have dispositional leanings toward conservative
and liberal moral values simultaneously, with political behav-
ior emerging out of the relative strength of the two systems.
Such models allow for the possibility that individuals with a
strong preference for order are not necessarily antiegalitar-
ian. Similarly, individuals who have a relative preference for
novelty are not necessarily more inclined toward equality.
Instead, attitudes toward equality and order may vary inde-
pendently from one another, in accordance with the strength
of their distinct underlying motivational systems. If this is
true, then the use of a single dimension to assess political
attitudes (conservatism vs. liberalism) might mask the asso-
ciation of Agreeableness with liberalism and its associated
moral values.
The two studies described in the current article were
designed to test the two possible reasons for the lack of pre-
vious findings of association of Agreeableness with political
attitudes (a) by examining the two different aspects of
Agreeableness, Compassion, and Politeness and (b) by utiliz-
ing a multidimensional measure of moral values. Study 1
employed a measure of the Big Five personality traits that dis-
tinguishes between the two empirically derived aspects of
Agreeableness to determine whether differential associa-
tions with political values would be observed. Two measures
of political values were employed to examine the robustness
of the effect. It was hypothesized that the Compassion aspect
of Agreeableness would be positively associated with liber-
alism, whereas the Politeness aspect would be negatively
associated with liberalism. Study 2 employed the Moral
Foundations Questionnaire (MFQ), an instrument designed
to assess multiple distinct moral systems thought to influ-
ence outward political behavior (and supporting the notion
that attitudes toward inequality and preferences for order can

Hirsh et al. 657
vary independently). It was hypothesized that higher levels of
Compassion would relate to moral systems associated with
egalitarianism, whereas low Openness-Intellect and high
Conscientiousness would relate to moral systems associated
with order and traditionalism.
Study 1
Method
Participants and design. Participants included 481 members
of the Eugene–Springfield Community Sample (ESCS; 200
male, 281 female), ranging in age from 20 to 85 years (M =
52.5, SD = 12.6). The ESCS is a longitudinal data collection
project in which a number of questionnaires were completed
by community members from Eugene and Springfield, Ore-
gon (Goldberg, 2005). Participants were recruited by mail
from lists of homeowners who agreed to complete question-
naires, delivered by mail, for pay, over a period of many
years, beginning in 1994. The sample spanned all levels of
educational attainment, with an average of 2 years of post-
secondary schooling. Most participants identified them-
selves as White (97%). The remainder were Hispanic, Asian
American, or Native American or did not report their ethnicity.
Because the included measures were spread across a number
of years, any results obtained should reflect the more stable
trait aspects of the constructs under investigation.
Measures
Big Five Aspect Scales (BFAS). The BFAS (DeYoung et al.,
2007) is an empirically derived instrument for measuring the
broad Big Five dimensions of personality as well as the
lower level aspects. Factor analyses of the facet scales from
two personality questionnaires, the Revised NEO Person-
ality Inventory (NEO PI-R; Costa & McCrae, 1992) and
Abridged Big Five Circumplex (Hofstee, De Raad, & Gold-
berg, 1992), revealed two correlated but distinct factors for
each of the Big Five domains (DeYoung et al., 2007). These
factors correspond to factors found in a behavior-genetic
study demonstrating that two distinct genetic factors underlie
the shared variance of the six facet scales composing each of
the Big Five domains in the NEO PI-R (Jang, Livesley,
Angleitner, Riemann, & Vernon, 2002). Each Big Five trait
domain thus appears divisible into two aspects, with distinct
biological sources. The BFAS was designed to assess this
midrange level of personality, between the broad domains
and the narrow facets.
The BFAS features 100 descriptions with which respon-
dents must rate their agreement on a 5-point Likert-type
scale (e.g., “Sympathize with others’ feelings”; “Like to
solve complex problems”). Items were selected from the
International Personality Item Pool (IPIP; Goldberg et al.,
2006) based on their correlations with the 10 aspect-level
factors derived by DeYoung et al. (2007). As a measure of
the broad domains, the BFAS has been validated against
standard Big Five instruments such as the Big Five Inventory
(BFI) and the NEO PI-R, with an average uncorrected corre-
lation of r = .76. The scale also demonstrates internal (mean
a = .83) and test–retest (mean r = .81) reliability. At the
aspect level, the five domains are broken down into Asser-
tiveness and Enthusiasm (Extraversion), Compassion and
Politeness (Agreeab leness), Industriousness and Orderliness
(Conscientiousness), Volatility and Withdrawal (Neuroti-
cism), and Openness and Intellect (Openness-Intellect).
Although the aspects from each domain are correlated with
each other, they are also characterized by reasonable and
meaningful discriminant validity (DeYoung et al., 2007) and
are not correlated so strongly (mean r = .44) as to present a
problem of collinearity when using pairs of aspects as simul-
taneous predictors in multiple regression. The BFAS thus
provides a good assessment of the broad Big Five domains
and provides the additional advantage of assessing an empir-
ically derived aspect level of personality. The IPIP items
(including those that now constitute the BFAS) were admin-
istered to the ESCS between 1994 and 1996.
BFI. The BFI (John & Srivastava, 1999) is a short, reliable
measure of the Big Five featuring 44 Likert-type scale items
ranging from 1 (strongly disagree) to 5 (strongly agree). The
BFI was employed as an additional measure of the Big Five
because multiple peer ratings using this instrument were
available for a subset of the sample (n = 293; 176 female,
117 male). Peer ratings contribute incremental validity over
self-ratings while minimizing the chance that obtained results
are because of rater bias (Fiedler, Oltmanns, & Turkheimer,
2004; Mount, Barrick, & Strauss, 1994). Results using scores
for self- and peer ratings on the BFI were therefore com-
pared to the BFAS results. During the summer of 1998, the
BFI was administered to the ESCS participants and to three
peers who knew the participants well and were asked to rate
them. The peer-rated personality scores were averaged together
to generate composite ratings for each Big Five dimension.
Political party preference. In 2001, 431 members of the sam-
ple completed two 5-point Likert-type scale items regarding
their political orientation: “Politically, I favor the Democratic
Party” and “Politically, I favor the Republican Party.” These
two items were strongly negatively correlated (r = –.87) and
were therefore combined to form a single Republican versus
Democrat variable by subtracting scores on the latter from
scores on the former. Higher scores on the combined vari-
able indicate greater support for the Republican Party. The
sample demonstrated a range of political preferences, with a
slight overall tilt toward preference for the Democratic Party
(M = –0.69, SD = 2.8).
Liberalism. Liberalism was assessed using the IPIP Liber-
alism scale, which is a 10-item scale for assessing liberal
versus conservative values (Goldberg, 1999). Participants
use a 5-point Likert-type scale to rate the extent to which

658 Personality and Social Psychology Bulletin 36(5)
they agree with a variety of political and moral statements
(e.g., “Tend to vote for liberal political candidates,” “Believe
that there is no absolute right or wrong,” “Believe laws
should be strictly enforced” (reversed), “Tend to vote for
conservative political candidates” (reversed)). The scale
demonstrates good internal reliability (Cronbach’s a = .86).
Liberalism and political party preference were strongly cor-
related (r = –.63, p < .01). The IPIP items were administered
to the ESCS between 1994 and 1996.
Results
A regression analysis was first conducted using each of the
broad Big Five domains to predict political party preference.
The overall model significantly predicted party preference,
R2 = .11, F(5, 425) = 10.11, p < .01. As expected, significant
effects were observed for Conscientiousness (b = .23), t(425) =
4.71, p < .01; Openness-Intellect (b = –.16), t(425) = –3.28,
p < .01; and Agreeableness (b = –.17), t(425) = –3.53, p < .01.
Republican supporters thus appear to be higher in Conscien-
tiousness, but lower in Openness-Intellect and Agreeableness.
A secondary series of regression analyses predicting polit-
ical party preference was conducted, with each regression
containing the two aspects of the significant domain-level
predictors. These analyses revealed that the domain-level
effects of Conscientiousness and Openness-Intellect were
driven by Orderliness (b = .26), t(428) = 3.94, p < .01, but
not Industriousness (b = –.03), t(428) = –0.50, p = .62, and
by Openness (b = –.37), t(428) = –6.81, p < .01, but not Intel-
lect (b = .09), t(428) = 1.67, p = .10. For the aspects of
Agreeableness, as hypothesized, preference for the Republi-
can Party was predicted negatively by Compassion (b = –.32),
t(428) = –5.35, p < .01, and positively by Politeness (b = .18),
t(428) = 3.00, p < .05.
Additional regression analyses were conducted to exam-
ine the personality correlates of the IPIP Liberalism scale.
When all Big Five domains were simultaneously entered
into the regression, the model predicted a significant amount
of the overall variance, R2 = .53, F(5, 475) = 36.35, p < .01.
Significant effects were again observed for Conscientious-
ness (b = –.36), t(475) = –8.57, p < .01, and Openness-Intellect
(b = .38), t(475) = 9.07, p < .01, but not for Agreeableness
(b = .00), t(478) = 0.04, p = .97.
The aspect level of personality was then examined in the
same manner as for political party preference. Significant
effects were observed for Orderliness (b = –.48), t(478) =
–8.05, p < .01, but not Industriousness (b = .10), t(478) = 1.7,
p = .09, and Openness (b = .44), t(478) = 9.12, p < .01, but
not Intellect (b = .07), t(478) = 1.38, p = .17. Although the
broad domain of Agreeableness was not a significant predic-
tor, an analysis of its aspect-level components again revealed
the hypothesized pattern of dissociation. Specifically, Com-
passion was positively associated with Liberalism (b = .27),
t(478) = 4.92, p < .01, whereas Politeness was negatively
associated with Liberalism (b = –.30), t(478) = –5.41, p < .01.
The failure of Agreeableness to significantly predict Liberal-
ism thus appears to be because of this divergence in the polit-
ical manifestations of its lower order aspects.
Peer-rated personality and political orientation. A limitation
of the above analysis is that it relies on self-report personal-
ity assessments. The regressions of political party preference
and Liberalism on the Big Five were thus duplicated, using
peer rating scores from the BFI. For political party prefer-
ence, significant effects were once again observed for Con-
scientiousness (b = .24), t(287) = 3.87, p < .01;
Openness-Intellect (b = –.28), t(287) = –4.75, p < .01; and
Agreeableness (b = –.21), t(287) = –3.07, p < .01. Very simi-
lar findings emerged when these traits were entered into a
regression predicting Liberalism, with significant effects
being found for Conscientiousness (b = –.30), t(287) = –5.34,
p < .01, and Openness-Intellect (b = .44), t(287) = 8.30, p <
.01, whereas a near-significant effect emerged for Agreeable-
ness (b = .12), t(287) = 1.91, p = .06. The observed associa-
tions between the Big Five and political orientation thus
remain reasonably stable across alternative measures and
both self and multi-informant ratings. Controlling for demo-
graphic variables (age, gender, education, and ethnicity) did
not influence the obs erved pattern of results. Because the BFI
does not differentiate the aspects of the Big Five, the multirater
data could not be used to examine the more specific aspect
levels of personality.1
Importance of measurement instrument and analytic tech-
nique. It is worth noting that the failure of previous studies to
detect significant effects for Agreeableness appears to be in
part related to the instrument and analysis employed. Because
the ESCS has data from multiple Big Five instruments, it
was possible to compare their ability to predict political out-
comes in a set of post hoc analyses. When using the Liberal-
ism scale as a measure of political ideology, no significant
zero-order correlations were observed for any of the Agree-
ableness measures (including the BFAS, NEO PI-R, and
self- and peer-rated BFI). When using political party prefer-
ence to assess political affiliation, significant effects were
observed only for BFAS (r = –.14, p < .01) and peer-rated
BFI Agreeableness (r = –.14, p < .05); no zero-order effects
were observed for the self-rated BFI (r = –.09, p > .05) or
NEO PI-R (r = –.08, p > .05). When entered into simultane-
ous regressions with the other Big Five factors, however,
both NEO PI-R and self-rated BFI Agreeableness signifi-
cantly predicted party affiliation. The relatively small effects
of domain-level Agreeableness may have been suppressed at
the zero-order level of analysis in previous research, only to
emerge when the variance associated with the other traits is
controlled. In addition, the current analysis of the BFAS sug-
gests that differentiation between Compassion and Politeness
is important for predicting political outcomes; instruments that
do not differentiate the two aspects of Agreeableness should
be less able to reveal the importance of Agreeableness.

Hirsh et al. 659
Indeed, even though NEO PI-R Agreeableness had no relation
to political orientation, the NEO facet Tender-Mindedness
(which reflects having sympathy for others) demonstrated a
positive correlation with Liberalism (r = .19, p < .01) and
reduced preference for the Republican Party (r = –.29, p < .01).
By contrast, the NEO facet Modesty (a tendency to demon-
strate polite humility) demonstrated a correlation with Liber-
alism in the opposite direction (r = –.13, p < .01). These
results are in keeping with the finding that Compassion and
Politeness predict political orientation in opposite directions.
It is worth noting at this point that although there is no defin-
itive list of facets in the Big Five framework, the aspects
measured by the BFAS reflect an empirically derived tax-
onomy of lower order personality traits (DeYoung et al.,
2007). We thus prefer the aspect-level analysis to the facet-
level analysis and believe it provides a clearer indication of
underlying personality processes.
Study 2
Method
Participants. Participants included 146 members of the Uni-
versity of Toronto community (47 male, 99 female), ranging
in age from 18 to 63 years (M = 26.14, SD = 9.14). They
were recruited by flyers posted around campus and messages
on electronic bulletin boards advertising the study. The sam-
ple included a variety of ethnic backgrounds, mostly consist-
ing of Caucasian (61.0%), East Asian (14.0%), and South
Asian (7.5%) participants. The remainder identified them-
selves as Hispanic, African, or Native American or did not
report their ethnicity. Inclusion criteria required that each
participant had voted in at least one government election
(city or municipal, state or provincial, or federal).
Measures
MFQ. The MFQ (Graham, Nosek, et al., 2009) is a
30-item self-report questionnaire designed to assess the five
domains of morality proposed by moral foundations theory
(Graham, Haidt, et al., 2009; Haidt & Joseph, 2007). The
five moral dimensions assessed by the questionnaire are
Harm-Care, Fairness-Reciprocity, Ingroup-Loyalty, Authority-
Respect, and Purity-Sanctity. These variables tend to cluster
into two broad factors with reasonable discriminant validity,
with Harm and Fairness being closely associated with self-
reported liberalism, and Ingroup, Loyalty, and Purity being
more associated with self-reported conservatism (Graham,
Haidt, et al., 2008).
Political orientation. The same liberalism and political party
preference measures were employed in Study 2 as in Study 1.
However, because Study 2 involved a Canadian population,
additional party preference items were included, assessing
participants’ attitudes toward the major Canadian political
parties with a national presence (Conservative Party, Liberal
Party, New Democratic Party, and Green Party). In addition,
we included a single item measure of self-rated political ori-
entation ranging from 1 (very conservative) to 5 (very liberal).
Participants again displayed a range of political orientations,
with a slight tilt toward the liberal end of the scale (M = 3.54,
SD = 0.95).
Personality was assessed using the BFAS, as in Study 1.
Procedure
Participants who expressed interest in the study were e-mailed
instructions for accessing the study materials online. After
completing an informed consent form, participants completed
online versions of a demographics questionnaire, the MFQ,
and the BFAS. Previous research suggests that online ques-
tionnaire-based assessment produces results similar to in-lab
assessments (Chuah, Drasgow, & Roberts, 2006). At the end
of the study, participants received an electronic debriefing
form and were reimbursed for their time with access to a
detailed report on their personality profile.
Results
Alpha reliabilities for each of the MFQ domains were as fol-
lows: Harm-Care = .61, Fairness-Reciprocity = .70, Ingroup-
Loyalty = .70, Authority-Respect = .63, Purity-Sanctity = .77.
A factor analysis using maximum likelihood estimation and
direct oblimin rotation (D = 0) was conducted on the five
moral foundation scores. Factor loadings are presented in
Table 1. Consistent with previous research, the MFQ domains
clustered into two broad dimensions, which can be described
as the moral values related to order-traditionalism (authority,
purity, and ingroup) and egalitarianism (fairness and harm-
care). These two factors were correlated with each other at
r = .16. The eigenvalues of these first two factors were 2.32
and 1.51 (accounting for 76.48% of the total variance), fol-
lowed by smaller factors with eigenvalues of 0.43, 0.40, and
0.34. Factor score estimates (based on the regression method)
were used in regressions to examine the personality predic-
tors of each cluster of moral domains.
Personality and moral values. In the first regression analysis,
each of the Big Five personality traits was used to predict the
Table 1. Factor Loadings of Moral Foundations
Egalitarianism Order-traditionalism
Harm .775 .045
Fairness .751 -.037
Ingroup .005 .744
Authority -.081 .799
Purity .092 .807

660 Personality and Social Psychology Bulletin 36(5)
order-traditionalism factor. The overall model predicted a
significant amount of variance, R2 = .13, F(5, 140) = 4.27,
p < .01. Specific personality effects were observed for Con-
scientiousness (b = .31), t(140) = 3.78, p < .01, and
Openness-Intellect (b = –.23), t(140) = –2.48, p < .01. Moral
values related to order-traditionalism were thus associated
with higher levels of trait Conscientiousness but lower levels
of Openness-Intellect. To examine the lower level traits
associated with these values, the analytic procedure of Study 1
was repeated, using regressions containing the two aspects
from each of these traits. These regressions demonstrated,
once again, that the personality predictors of these values
were specific to the Orderliness aspect of Conscientiousness
(b = .27), t(140) = 3.23, p < .01, but not Industriousness (b =
.08), t(140) = 0.96, p = .34. For Openness-Intellect, the pre-
dictive variance appears to have been split between Open-
ness (b = –.11), t(140) = –1.28, p = .20, and Intellect (b = –.09),
t(140) = –1.01, p = .31. Both of these aspects displayed simi-
lar relationships with order-traditionalism when analyzed
using zero-order correlations (r = –.15 for Openness, r = –.13
for Intellect), suggesting that neither aspect was a stronger
predictor. When looking at the role of the Agreeableness
aspects, a significant effect was observed for Politeness (b =
.18), t(140) = 1.98, p < .05, but not Compassion (b = –.11),
t(140) = –1.20, p = .23.
This procedure was repeated to examine the personality
predictors of egalitarian values. Again, the overall model
predicted a significant amount of the variance in this cluster
of moral domains, R2 = .17, F(5, 140) = 5.58, p < .01. Of the
broad personality domains, Agreeableness was the only sig-
nificant predictor (b = .33), t(140) = 4.20, p < .01, with more
agreeable individuals reporting a greater endorsement of egal-
itarian values. When looking at the lower order aspect level,
this effect appeared to be driven primarily by the Compassion
aspect of Agreeableness (b = .29), t(140) = 3.45, p < .01, with
less influence of Politeness (b = .13), t(140) = 1.58, p = .12.
In addition to the two broad factors extracted above, which
have previously been related to political ideology, the rela-
tionship between the more differentiated moral foundation
scales and each of the BFAS dimensions was examined
while controlling for the other personality traits. Results of
these analyses are reported in Table 2. As can be seen, the
relationship between personality and moral values was largely
preserved at the level of specific moral foundations. Of par-
ticular note is the fact that Compassion was most strongly
associated with the Harm-Care and Fairness-Reciprocity
scales. Politeness, in contrast, was more strongly associated
with Authority-Respect, but not with Fairness-Reciprocity.
Both of these findings are in keeping with the notion that
Compassion more closely reflects egalitarianism, whereas
Politeness is more closely related to order-traditionalism.
Moral values and political orientation. The next analyses exam-
ined whether the moral foundations described above contrib-
uted uniquely to political orientation. Indeed, according to
previous research, attitudes toward equality and support for
order-traditionalism should reflect two distinct motivational
factors that contribute to outward political behavior (Jost,
2006; Jost et al., 2003). To examine this question, a compos-
ite index of political orientation was derived by conducting a
factor analysis with maximum likelihood estimation on the
available measures of political behavior described previ-
ously (Liberalism scale, support for American and Canadian
political parties, and self-rated political orientation). The
scree plot suggested that the first factor was an appropriate
cutoff point, accounting for considerably more shared vari-
ance than the other factors. The eigenvalue of the first factor
Table 2. Associations Between Personality and the Moral Foundations Questionnaire
Harm-Care Fairness-Reciprocity Ingroup-Loyalty Authority-Respect Purity-Sanctity
Neuroticism .03 .13 .02 .00 -.02
Volatility .01 .04 .00 .10 -.07
Withdrawal .03 .12 .02 -.11 .08
Extraversion .10 .10 -.06 .03 .04
Assertiveness -.04 .00 -.09 -.01 -.07
Enthusiasm .15 .11 .00 .04 .10
Openness-Intellect .13 .14 -.12 -.25* -.20*
Intellect .03 -.04 -.14 -.26* -.25*
Openness .12 .17 -.02 -.06 -.01
Agreeableness .38* .21* -.02 .06 .08
Compassion .32* .20* -.10 -.13 -.06
Politeness .16 .06 .07 .19* .15
Conscientiousness -.03 -.02 .25* .29* .27*
Industriousness -.03 -.05 .06 .10 .11
Orderliness .00 .09 .22* .24* .21*
*p < .05.

Hirsh et al. 661
was 3.15 (39.4% of the total variance), followed by smaller
factors with eigenvalues of 1.21, 0.96, 0.77, 0.75, 0.50, 0.40,
and 0.27.
The broad moral foundation factors of egalitarianism and
order-traditionalism were simultaneously entered into a regres-
sion predicting this political outcome measure. The overall
model predicted a significant amount of the variance in polit-
ical orientation, R2 = .18, F(2, 143) = 16.13, p < .01. Further-
more, political preferences were predicted independently by
the egalitarianism factor (b = .27), t(143) = 3.44, p < .01, and
the order-traditionalism factor (b = –.40), t(143) = –5.12, p <
.01. In a secondary analysis, the five moral foundation scales
were entered simultaneously into a regression predicting
political orientation. Significant effects were observed for
Fairness-Reciprocity (b = .29), t(140) = 3.09, p < .01, as well
as a near-significant trend for Authority-Respect (b = –.20),
t(140) = –1.93, p = .06. By contrast, no independent effects
were observed for Harm-Care (b = –.01), t(140) = –0.09, p =
.93; Ingroup-Loyalty (b = –.14), t(140) = –1.34, p = .18; or
Purity-Sanctity (b = –.09), t(140) = –0.87, p = .39. Note,
however, that all MFQ domains except Harm-Care significantly
predicted political orientation when employing zero-order
correlations (Harm-Care = .10, Fairness-Reciprocity = .26,
Ingroup-Loyalty = –.29, Authority-Respect = –.34, Purity-
Sanctity = –.27), suggesting that much of the predictive vari-
ance is shared amongst the specific scales.
Personality and political orientation. Repeating the regression
analyses from Study 1 also confirmed that the same person-
ality traits were able to predict political behavior. In particu-
lar, liberal political attitudes were again associated with the
broad traits of Conscientiousness (b = –.15), t(140) = –1.72,
p < .05, and Openness-Intellect (b = .24), t(140) = 2.56, p <
.05. At the aspect level, political attitudes were associated
with Orderliness (b = –.16), t(143) = –1.86, p < .05, but not
Industriousness (b = .02), t(143) = 0.20, p = .84, as well as
Openness (b = .18), t(143) = 1.97, p < .05, but not Intellect
(b = .08), t(143) = 0.85, p = .40. For the aspects of Agree-
ableness, the expected dissociation was again observed, with
Compassion being associated with liberal political attitudes
(b = .23), t(143) = 2.57, p < .05, and Politeness being associ-
ated with conservative political attitudes (b = –.20), t(143) =
–2.26, p < .05.
To directly test the possibility that these personality traits
influence political attitudes through their association with tra-
ditionalism and egalitarian values, mediation analyses among
the BFAS, the MFQ, and the political outcome measures were
conducted. The product of coefficients method recommended
by MacKinnon, Lockwood, Hoffman, West, and Sheets (2002)
was employed to determine whether or not significant medi-
ation effects were observed. As expected, egalitarian values
(but not traditionalist values) significantly mediated the rela-
tionship between Compassion and liberal political attitudes
(z' = 1.81, p < .01). Similarly, traditionalist values (but not
egalitarian values) mediated the relationships between
Politeness (z' = –1.82, p < .01), Orderliness (z' = –3.00, p <
.01), and Openness-Intellect (z' = 1.69, p < .01) and political
conservatism.
General Discussion
A growing awareness has developed in recent years of the
psychological factors underlying political behavior and affil-
iation (Jost et al., 2008). In particular, the adoption of differ-
ent political practices and beliefs are thought to reflect an
individual’s psychological needs (Jost, 2006; Jost et al., 2003).
The current findings extend this idea by demonstrating spe-
cific personality traits associated with liberal and conserva-
tive ideologies. In previous research, resistance to change
and the acceptance of inequality were identified as the two
main components of conservative values (Jost et al., 2007).
What the current findings suggest is that these two motives
are indeed relevant to political orientation and that they
reflect the functioning of distinct personality systems. Spe-
cifically, resistance to change appears to stem from high lev-
els of Orderliness and Politeness and a low level of
Openness-Intellect, whereas acceptance of inequality stems
from a low level of Compassion. Importantly, although pre-
vious studies have identified Conscientiousness and Open-
ness-Intellect as personality predictors of political beliefs
(Carney et al., 2008; Goldberg & Rosolack, 1994; Jost,
2006), this is the first set of studies to identify the role of
Agreeableness in political ideology. Past failure to identify
the importance of Agreeableness appears primarily due to
the fact that the two aspects of Agreeableness diverge in
their associations with political ideology. Although the Com-
passion aspect of Agreeableness was associated with greater
concern for egalitaria nism (associated with liberal political
attitudes), the Politeness aspect of Agreeableness was associ-
ated with greater concern for order-traditionalism (associated
with conservative political attitudes).
Importantly, it has been previously suggested that atti-
tudes toward inequality may derive from attitudes toward the
status quo (cf. Jost et al., 2003). That is, if an individual sup-
ports the status quo in a nonegalitarian society, any move-
ment toward greater equality is evaluated negatively because
it is inherently disruptive of the current order and tradition.
There are two lines of evidence in the current studies that
suggest that these two ideological dimensions are in fact sep-
arable from one another. First, as in previous research look-
ing at political values (Eysenck, 1954, 1975; Rokeach, 1973),
the value domains of egalitarianism and order-traditionalism
were empirically distinct from one another, suggesting that
they can vary independently within any given individual. In
addition, these two value domains were found to significantly
predict different aspects of the overall variance in political
orientation rather than being statistically redundant (as would
be predicted by the theory that one domain is derived from the
other). Second, the differential association of these value

662 Personality and Social Psychology Bulletin 36(5)
domains with the Big Five personality traits further supports
the two-factor model of political ideology. In particular,
Openness-Intellect, Orderliness, and Politeness are all con-
ceptually and empirically related to the value domain of
order-traditionalism. Compassion, meanwhile, is empirically
and conceptually related to the value domain of egalitarian-
ism. The Big Five is a well-validated taxonomy of dispositional
variables, and the differential associations with egalitarian-
ism and order-traditionalism support the discriminant valid-
ity of these dimensions.
With regard to the aspects of Openness-Intellect, Study 1
found that Openness was a stronger predictor of political
behavior than was Intellect. These findings were partially
supported in Study 2, although the broad Openness-Intellect
domain appeared to be the better predictor of political out-
comes in this sample. It is possible that the relative strength
of association between each of the two aspects and political
behavior varies in different populations, with a somewhat
stronger role for Openness. Based on the current findings,
however, the most parsimonious explanation is that the entire
domain of Openness-Intellect is negatively associated with
order-traditionalism and political conservatism.
According to the current results, an individual’s overall
political orientation is likely to result from the relative bal-
ance of motivational needs for order and traditionalism on
one hand (as reflected by Openness-Intellect, the Orderliness
aspect of Conscientiousness, and the Politeness aspect of
Agreeableness) and for equality and fairness on the other (as
reflected by the Compassion aspect of Agreeableness).
Indeed, Study 2 found that the relationship between person-
ality and political orientation was mediated by the subjective
importance of these moral values. Personality traits thus appear
to influence political behavior by influencing the strength of
these two underlying motivational systems. Individuals who
have high needs for order but low needs for equality are
likely to score at the high ends of conservative ideology.
Conversely, individuals with low needs for order but high
needs for equality are likely to score at the high ends of lib-
eral ideology. If, by contrast, both of these needs are rela-
tively balanced, a more moderate political outlook is likely
to be observed. Although the term bleeding-heart liberal is
often used pejoratively, the current findings suggest that
liberals do indeed tend to have higher levels of compassion.
These higher levels of compassion likely contribute to the
liberal’s preference for fairness and equality. In contrast,
the term compassionate conservative may be something of
an oxymoron. It is true that individuals with a more balanced
personality profile may endorse both conservative and lib-
eral values, but conservatism as a political orientation
appears to be negatively associated with compassion. This
does not mean there are no compassionate conservatives,
but it suggests that the extent to which conservatives are
compassionate may reflect the extent to which they pos-
sess the underlying motivation driving the liberal value of
egalitarianism.
Although certain personality characteristics are often val-
ued over others, the results of these two studies do not indicate
the moral superiority of either the liberal or the conservative
personality profile. Instead, the analyses support the notion that
liberals and conservatives are motivated by distinct domains
of morality (Graham, Haidt, et al., 2009; Haidt & Graham,
2007). The fact that a range of personality profiles has evolved
suggests that no single profile is optimal across all contexts
(Nettle, 2006). Each personality profile involves a distinct
mixture of motivations and cognitive tendencies, so various
costs and benefits will be associated with any particular com-
bination of traits. The conservative profile described above
may, for instance, encourage greater personal responsibility,
self-discipline, and social stability, whereas the liberal profile
may encourage greater innovation and social inclusiveness.
Understanding political behavior in terms of underlying
psychological systems has been a long-standing research goal
(Adorno et al., 1950; Eysenck, 1954; Rokeach, 1973). Apply-
ing contemporary models from social and personality psycho-
logy appears to provide a valuable extension of this classic
work. Importantly, an individual’s political orientation appears
to be reflected not only in a distinct pattern of psychological
motives (Jost et al., 2003) but also in a distinct personality
profile. Given the increasing polarization of the political
sphere, this topic remains a vital and timely point of inquiry.
Acknowledgment
We thank Lewis R. Goldberg for his generosity in making data
available from the Eugene–Springfield Community Sample.
Declaration of Conflicts of Interest
The authors had no conflicts of interest with respect to the author-
ship or the publication of this article.
Funding
The authors received the following financial support for the research
and/or authorship of this article: research support provided by the
Social Sciences and Humanities Research Council of Canada.
Note
1. Soto and John (2009) developed a method for scoring two fac-
ets within each of the Big Five scales in the Big Five Inventory
(BFI). However, these two facets do not in every case map
on cleanly to the aspect-level traits. For example, one of the
BFI Agreeableness facets is Altruism, which is associated
equally with Compassion and Politeness (DeYoung, Quilty,
& Peterson, 2007).
References
Adorno, T. W., Frenkel-Brunswik, E., Levinson, D. J., & Sanford, R. N.
(1950). The authoritarian personality. New York, NY: Harper
& Row.
Ashton, M., Paunonen, S., Helmes, E., & Jackson, D. (1998). Kin
altruism, reciprocal altruism, and the Big Five personality fac-
tors. Evolution and Human Behavior, 19(4), 243-255.

Hirsh et al. 663
Bouchard, T., Segal, N., Tellegen, A., McGue, M., Keyes, M., &
Krueger, R. (2003). Evidence for the construct validity and heri-
tability of the Wilson–Patterson Conservatism Scale: A reared-
apart twins study of social attitudes. Personality and Individual
Differences, 34(6), 959-969.
Carney, D., Jost, J., Gosling, S., & Potter, J. (2008). The secret lives
of liberals and conservatives: Personality profiles, interaction
styles, and the things they leave behind. Political Psychology,
29(6), 807-840.
Chuah, S. C., Drasgow, F., & Roberts, B. W. (2006). Personality
assessment: Does the medium matter? No. Journal of Research
in Personality, 40, 359-376.
Conover, P., & Feldman, S. (1981). The origins and meaning of
liberal/conservative self-identifications. American Journal of
Political Science, 25(4), 617-645.
Costa, P. T., & McCrae, R. R. (1992). Revised NEO Personality
Inventory and NEO Five-Factor Inventory professional manual.
Odessa, FL: Psychological Assessment Resources.
DeYoung, C. G., Quilty, L. C., & Peterson, J. B. (2007). Between
facets and domains: Ten aspects of the Big Five. Journal of Per-
sonality and Social Psychology, 93(5), 880-896.
Duckitt, J., Wagner, C., Du Plessis, I., & Birum, I. (2002). The
psychological bases of ideology and prejudice: Testing a dual
process model. Journal of Personality and Social Psychology,
83(1), 75-93.
Eysenck, H. J. (1954). The psychology of politics. London, UK:
Routledge.
Eysenck, H. J. (1975). The structure of social attitudes. British
Journal of Social and Clinical Psychology, 14, 323-331.
Fiedler, E. R., Oltmanns, T. F., & Turkheimer, E. (2004). Traits
associated with personality disorders and adjustment to military
life: Predictive validity of self and peer reports. Military Medi-
cine, 169(3), 207-211.
Goldberg, L. R. (1999). A broad-bandwidth, public domain, per-
sonality inventory measuring the lower-level facets of several
five-factor models. In I. Mervielde, I. J. Deary, F. De Fruyt, &
F. Ostendorf (Eds.), Personality psychology in Europe (Vol. 7,
pp. 7-28). Tilburg, Netherlands: Tilburg University Press.
Goldberg, L. R. (2005). The Eugene–Springfield Community
Sample: Information available from the research participants
(Report No. 45). Eugene: Oregon Research Institute.
Goldberg, L. R., Johnson, J. A., Eber, H. W., Hogan, R., Ashton, M. C.,
Cloninger, C. R., & Gough, H. G. (2006). The International Per-
sonality Item Pool and the future of public-domain personality
measures. Journal of Research in Personality, 40, 84-96.
Goldberg, L. R., & Rosolack, T. K. (1994). The Big Five factor struc-
ture as an integrative framework: An empirical comparison with
Eysenck’s PEN model. In C. F. Halverston, Jr., G. A. Kohnstamm,
& R. P. Martin (Eds.), The developing structure of temperament
and personality from infancy to adulthood (pp. 7-35). Mahwah,
NJ: Lawrence Erlbaum.
Graham, J., Haidt, J., & Nosek, B. A. (2009). Liberals and conser-
vatives rely on different sets of moral foundations. Journal of
Personality and Social Psychology, 96(5), 1029-1046.
Graham, J., Nosek, B. A., Haidt, J., Iyer, R., Koleva, S., &
Ditto, P. H. (2009). Broadening and mapping the moral domain:
The development and validation of the Moral Foundations
Questionnaire. Manuscript submitted for publication.
Graziano, W., Bruce, J., Sheese, B., & Tobin, R. (2007). Attrac-
tion, personality, and prejudice: Liking none of the people most
of the time. Journal of Personality and Social Psychology,
93(4), 565-582.
Graziano, W., Habashi, M., Sheese, B., & Tobin, R. (2007). Agree-
ableness, empathy, and helping: A person X situation perspective.
Journal of Personality and Social Psychology, 93(4), 583-599.
Graziano, W., & Tobin, R. (2002). Agreeableness: Dimension of
personality or social desirability artifact? Journal of Personal-
ity, 70(5), 695-728.
Haidt, J., & Graham, J. (2007). When morality opposes justice:
Conservatives have moral intuitions that liberals may not recog-
nize. Social Justice Research, 20(1), 98-116.
Haidt, J., & Joseph, C. (2007). The moral mind: How 5 sets of innate
intuitions guide the development of many culture-specific vir-
tues, and perhaps even modules. In P. Carruthers, S. Laurence,
& S. Stich (Eds.), The innate mind: Foundations and the future
(Vol. 3, pp. 367-391). New York, NY: Oxford University Press.
Hofstee, W. K. B., De Raad, B., & Goldberg, L. R. (1992). Integra-
tion of the big five and circumplex approaches to trait structure.
Journal of Personality and Social Psychology, 63(1), 146-163.
Jang, K. L., Livesley, W. J., Angleitner, A., Riemann, R., &
Vernon, P. A. (2002). Genetic and environmental influences
on the covariance of facets defining the domains of the five-
factor model of personality. Personality and Individual Dif-
ferences, 33(1), 83-101.
John, O. P., & Srivastava, S. (1999). The Big Five trait taxonomy: His-
tory, measurement, and theoretical perspectives. In L. A. Pervin &
O. P. John (Eds.), Handbook of personality: Theory and research
(pp. 102-138). New York, NY: Guilford.
Jost, J. T. (2006). The end of the end of ideology. American Psy-
chologist, 61(7), 651-670.
Jost, J. T., Fitzsimons, G., & Kay, A. C. (2004). The ideological ani-
mal: A system justification view. In J. Greenberg, S. L. Koole, &
T. Pyszczynski (Eds.), Handbook of experimental existential psy-
chology (pp. 263-283). New York, NY: Guilford.
Jost, J. T., Glaser, J., Kruglanski, A. W., & Sulloway, F. J. (2003).
Political conservatism as motivated social cognition. Psycho-
logical Bulletin, 129(3), 339-375.
Jost, J. T., Napier, J. L., Thorisdottir, H., Gosling, S. D., Palfai, T. P., &
Ostafin, B. (2007). Are needs to manage uncertainty and threat asso-
ciated with political conservatism or ideological extremity? Person-
ality and Social Psychology Bulletin, 33(7), 989-1007.
Jost, J. T., Nosek, B. A., & Gosling, S. D. (2008). Ideology: Its
resurgence in social, personality, and political psychology. Per-
spectives on Psychological Science, 3(2), 126-136.
Koenig, L. B., & Bouchard, T. J., Jr. (2006). Genetic and envi-
ronmental influences on the traditional moral values triad—
Authoritarianism, conservatism and religiousness—as assessed
by quan titative behavior genetic methods. In P. McNamara (Ed.),

664 Personality and Social Psychology Bulletin 36(5)
Where God and science meet: How brain and evolutionary
studies alter our understanding of religion (Vol. 1, pp. 31-60).
Westport, CT: Praeger.
Landau, M., Solomon, S., Greenberg, J., Cohen, F., Pyszczynski, T.,
Arndt, J., . . . Cook, A. (2004). Deliver us from evil: The effects
of mortality salience and reminders of 9/11 on support for Presi-
dent George W. Bush. Personality and Social Psychology Bul-
letin, 30(9), 1136-1150.
Lippa, R., & Arad, S. (1999). Gender, personality, and prejudice:
The display of authoritarianism and social dominance in inter-
views with college men and women. Journal of Research in
Personality, 33, 463-493.
MacKinnon, D., Lockwood, C., Hoffman, J., West, S., & Sheets, V.
(2002). A comparison of methods to test mediation and other inter-
vening variable effects. Psychological Methods, 7(1), 83-104.
McCrae, R. R., & Costa, P. T. (1997). Conceptions and correlates of
openness to experience. In R. Hogan, J. A. Johnson, & S. R. Briggs
(Eds.), Handbook of personality psychology (pp. 825-847). San
Diego, CA: Academic Press.
Moskowitz, G. B., Gollwitzer, P. M., Wasel, W., & Schaal, B.
(1999). Preconscious control of stereotype activation through
chronic egalitarian goals. Journal of Personality and Social Psy-
chology, 77(1), 167-184.
Mount, M. K., Barrick, M. R., & Strauss, J. P. (1994). Validity of
observer ratings of the Big Five personality factors. Journal of
Applied Psychology, 79(2), 272-280.
Nettle, D. (2006). The evolution of personality variation in humans
and other animals. American Psychologist, 61(6), 622-631.
Nettle, D. (2007). Empathizing and systemizing: What are they,
and what do they contribute to our understanding of psycho-
logical sex differences? British Journal of Psychology, 98(2),
237-256.
Roberts, B. W., Chernyshenko, O. S., Stark, S., & Goldberg, L. R.
(2005). The structure of conscientiousness: An empirical inves-
tigation based on seven major personality questionnaires. Per-
sonnel Psychology, 58(1), 103-140.
Rokeach, M. (1973). The nature of human values. New York, NY:
Free Press.
Soto, C. J., & John, O. P. (2009). Ten facet scales for the Big Five
Inventory: Convergence with NEO PI-R facets, self-peer agree-
ment, and discriminant validity. Journal of Research in Person-
ality, 43, 84-90.