You are on page 1of 9


discussions, stats, and author profiles for this publication at:

We the People

Article in Social Psychological and Personality Science June 2013

DOI: 10.1177/1948550613512511


2 546

3 authors, including:

Jojanneke van der Toorn Jaime Napier

Utrecht University New York University Abu Dhabi


All content following this page was uploaded by Jaime Napier on 03 March 2016.

The user has requested enhancement of the downloaded file.

Social Psychological and Personality Science

We the People: Intergroup Interdependence Breeds Liberalism

Jojanneke van der Toorn, Jaime L. Napier and John F. Dovidio
Social Psychological and Personality Science 2014 5: 616 originally published online 5 December 2013
DOI: 10.1177/1948550613512511

The online version of this article can be found at:

Published by:

On behalf of:
Society for Personality and Social Psychology

Association for Research in Personality

European Association of Social Psychology

Society of Experimental and Social Psychology

Additional services and information for Social Psychological and Personality Science can be found at:

Email Alerts:




>> Version of Record - Jun 2, 2014

OnlineFirst Version of Record - Dec 5, 2013

What is This?

Downloaded from at Yale University Library on June 2, 2014

Social Psychological and
Personality Science
We the People: Intergroup 2014, Vol. 5(5) 616622
The Author(s) 2013
Reprints and permission:
Interdependence Breeds Liberalism
DOI: 10.1177/1948550613512511

Jojanneke van der Toorn1, Jaime L. Napier2, and John F. Dovidio2

Whereas much social psychological research has focused on the conditions that lead to political conservatism, the current
research suggests that instilling a sense of intergroup interdependence can increase political liberalism and, in turn, foster concern
for universal welfare. Using both correlational (Study 1) and experimental (Study 2) methodologies, we find convergent support
for the novel hypothesis that perceived interdependence between groups in society increases peoples support for human rights
because it increases liberalism. In addition to establishing the hypothesized effect, we empirically distinguished the effect of
intergroup interdependence from that of intragroup (or interpersonal) interdependence, which was related to conservatism.
This research presents a novel demonstration of the effect of intergroup interdependence on political attitudes and fills a gap in
the literature on the conditions that lead to liberalism.

intergroup interdependence, liberalism, human rights, ideological shift, policy support

The basic thought that guides these specific means of national The notion that peoples support for political policies
recovery is not narrowly nationalistic. It is the insistence, as a first reflects underlying psychological needs has garnered consider-
consideration, upon the interdependence of the various elements in able empirical support (see Jost, Federico, & Napier, 2009).
all parts of the United States. (U.S. President Franklin D. Roose- The majority of this work has been inspired by a motivated
velt, First Inaugural Address, 1933) social-cognition account of political ideology and has focused
There is no issue of States rights or National rights. There is only on how adverse environmental factors, such as threat or uncer-
the struggle for human rights. (U.S. President Lyndon B. Johnson, tainty, lead individuals to adopt more politically conservative
Voting Rights Act Address, 1965) beliefs (e.g., Jost, Glaser, Kruglanski, & Sulloway, 2003;
Landau et al., 2004; Nail, McGregor, Drinkwater, Steele, &
The policies adopted during the emergence of modern liberal- Thompson, 2009; Thorisdottir & Jost, 2011), presumably
ism in the United Statesstarting with Roosevelts New because these beliefs provide a sense of security, order, and
Deal and peaking under the Johnson administrations Great certainty. Less is known, however, about what conditions
Societynow form the foundations of American civil liber- might foster more liberal attitudes.
ties (Alterman, 2008; Milkis & Mileur, 2005; Shaw, 1988). Recent theoretical perspectives on the moral values of liberals
As illustrated in the above quotations, both of these leaders and conservatives have extended this model of ideological differ-
appealed to Americans sense of societal interdependence to ences in ways that could offer intriguing insights into the motiva-
promote domestic agendas that protected the rights of all tional underpinnings of liberalism. Specifically, Janoff-Bulman
citizens. In the current work, we empirically examine the rela- and colleagues (Janoff-Bulman, 2009; Janoff-Bulman & Carnes,
tionships between societal (or intergroup) interdependence 2013; Janoff-Bulman, Sheikh, & Baldacci, 2008) have proposed
and political policy support. We propose a causal model that an approach/avoidance model to illuminate ideological divisions
might account, at least in part, for why these unprecedented in moral concern: Conservative values tend to be based in
(and since unrivaled) human rights policies were successfully
implemented during times when diverse groups of Americans
might have felt especially interconnectedon the heels of the Leiden University, Leiden, The Netherlands
Great Depression or after the loss of a President. Specifically, Yale University, New Haven, CT, USA
we hypothesize that an increased salience of the dependency
Corresponding Author:
between various groups in society will lead people to embrace Jojanneke van der Toorn, Department of Psychology, Leiden University,
liberal viewpoints that, in turn, increase support for human PO Box 9555, Leiden, 2300 RB, The Netherlands.
rights. Email:

Downloaded from at Yale University Library on June 2, 2014

van der Toorn et al. 617

avoidance-type motivations, and thus focus on protection, emphasizes hierarchy and distinctive roles, predicts greater con-
whereas liberal values are based in approach-type motivations servatism (Janoff-Bulman & Carnes, 2013). Vertical collectivism
and focus on providing. On the collective (or societal) level, con- is also positively related to authoritarianism (Kemmelmeier et al.,
servatives concerns center on protection from threat. Toward that 2003).
end, they value social order and promote notions of in-group We posit that intergroup interdependence will be related to
loyalty and cohesion (see also Graham, Haidt, & Nosek, 2009). political liberalism insofar as liberal policies promote well-
Liberals, by contrast, are concerned about providing for the col- being for all groups in society (see Janoff-Bulman, 2009;
lective; in turn, they value social justice and promote notions Janoff-Bulman & Carnes, 2013), but that conservatives will
of intergroup interdependence and shared responsibility be more inclined toward interpersonal (or intragroup) interde-
(Janoff-Bulman & Carnes, 2013). pendence (see Graham et al., 2009). In two studies, we tested
This model highlights one important (yet understudied) dif- the possibility that liberalism and, in turn, support for policies
ference between liberals and conservativestheir divergent promoting human rights, would be enhanced when people
conceptions of group interdependence. Conservatives tend to perceive intergroup (but not interpersonal) interdependence.
focus on interdependence within their own groups as well as To our knowledge, no research has directly examined how
the boundaries between groups, whereas liberals focus on col- interdependence potentially affects political ideology nor stud-
lectivism and interdependence within (constellations of) ied the implications of this relationship for policy support. In
groups (Janoff-Bulman, 2009; Janoff-Bulman, Sheikh, & Study 1, we conducted a survey to test the mediation hypothesis
Baldacci, 2008). In line with this, research has demonstrated that perceived intergroup interdependence (but not interperso-
that there are ideological differences in the perceived scope nal interdependence) is positively related to participants
of category memberships, such that liberals have a more inclu- liberalism scores and, consequently, with support for universal
sive conception of group (e.g., society) as compared to conser- welfare. In Study 2, we used an experimental paradigm to
vatives (e.g., my ethnic group; Rock & Janoff-Bulman, 2010). establish causality.
The notion that intergroup interdependence is an important
ideological construct is in line with work in political science,
which highlights the fact that left-wing (vs. right wing) ideologies Study 1
tend to espouse a general view that human beings should live
together cooperatively without social distinctions (Magstadt, Method
2010) and emphasize interdependence between groups and Participants. Eighty-seven White American Mechanical Turk
nations (Monaghan, 1984). Consistent with this position, Stel and participants (age, mean [M] 30.17, standard deviation
Harinck (2011) found that a manipulation of interdependence (vs. [SD] 11.17; 67.8% male) completed an online survey in
independence) led to leftist voting behavior. Conservative ideol- exchange for monetary compensation.
ogies, by contrast, tend to emphasize either individualism or a dif-
ferent type of interdependenceinterdependence on others Procedure. Participants first rated their political orientation on a
within ones group. Moral foundations theory (Graham et al., scale ranging from 1 (extremely liberal) to 7 (extremely conser-
2009) proposes that conservatives, to a greater extent than liber- vative), which was reverse coded so that higher scores indicate
als, possess the binding moral value of community, referring greater liberalism (Jost et al., 2009). Participants perceived
to ones perception of connectedness to other in-group members. intergroup interdependence was then assessed with the follow-
In the present research, we examined how the nature of inter- ing statement: Different groups are needed for society to
dependence affects political orientations and ultimately support work (1 extremely disagree; 7 extremely agree). To
for human rights. In social psychological research, much of the measure participants sense of interpersonal interdependence,
research on interdependence has been in the context of work on we used Lu and Gilmours (2007) 21-item interdependence
the relational self, with a focus on how people define themselves scale (1 extremely disagree; 7 extremely agree; a
in terms of their social relationships, group memberships, and .86). This scale was designed to measure constructs at the
social roles (Gardner, Gabriel, & Lee, 1999; Markus & Kitayama, individual level to correspond to individualism-collectivism
1991; Trafimow, Triandis, & Goto, 1991; see Oyserman, Coon, & (IC) at the cultural level (p. 249) and includes items that
Kemmelmeier, 2002 for an overview). However, research in the reflect various aspects of this distinction, including conformity
area of intergroup relations demonstrates the importance of (e.g., I believe that people should behave appropriately
whether such feelings of connectedness occur within or across according to their different social status and roles), cohesion
group boundaries: Whereas perceptions of intergroup interdepen- (e.g., I have a strong identification with people close to
dence reduce bias (Dovidio & Gaertner, 2010; Sherif, Harvey, me), and identification with various in-groups (e.g., I
White, Hood, & Sherif, 1961), perceptions of greater interdepen- believe that family is the source of our self). Finally, to assess
dence on others within ones group promote intergroup bias support for universal welfare, we adapted the fixed resource
(Gaertner, Iuzzini, Witt, & Orina, 2006). In addition, in the cul- distribution task employed by Son Hing, Li, and Zanna
tural psychology literature, greater horizontal collectivism, which (2002; see also Haddock, Zanna, & Esses, 1993), which is
emphasizes greater cooperation with others generally, predicts designed to reduce the influence of social desirability respond-
greater liberalism, whereas greater vertical collectivism, which ing by requiring participants to allocate limited funds between

Downloaded from at Yale University Library on June 2, 2014

618 Social Psychological and Personality Science 5(5)

Table 1. Descriptives and Intercorrelations (Study 1).

Variable M (SD) 1 2 3 4 5

1. Intergroup interdependence 5.41 (1.31)

2. Interpersonal interdependence 4.43 (0.74) .03
3. Liberalism 5.07 (1.60) .23* .25*
4. Support for human rights 45.25 (19.34) .24* .11 .48***
5. Gender 0.36 (0.94) .16 .22* .08 .11
6. Age 30.17 (11.17) .00 .25* .12 .08 .26*
Note. Gender was effect-coded (women 1; men 1).
*p < .05, **p < .01, ***p < .001.

desirable alternatives. Specifically, we told participants: the To examine whether the effect of intergroup interdependence
government is currently making its annual budget and has set on support for human rights was mediated by liberalism, we con-
aside a certain amount of money to be split between Human ducted bootstrapping analyses based on 1,000 bootstrap resam-
Rights Advocate groups and Road Maintenance. Last year, this ples, adjusting for interpersonal interdependence (Preacher &
money was split 50/50. We selected road maintenance as the Hayes, 2004). As predicted, liberalism significantly predicted
alternative to human rights because it is ideologically neutral support for human rights, b 6.18, SE 1.19, p < .001. In addi-
and considered generally valuable. Participants then indicated tion, with liberalism in the model, the direct effect of intergroup
on slider scales (which added up to 100%), what percentage the interdependence on support for human rights was no longer sig-
government should allocate for each issue this year.1 nificant, b 1.63, SE 1.41, p .250, suggesting mediation
(see Figure 1). The bias-corrected confidence interval did not
include zero [0.57; 4.26], confirming that the indirect path was
Results and Discussion Because these data were correlational, we were also able to
Table 1 lists the variable Ms, SDs, and intercorrelations. test the competing hypothesis that perceived intergroup inter-
Intergroup interdependence was not significantly related to inter- dependence would mediate the relationship between liberalism
personal interdependence, r(85) .03, p .782, suggesting that and support for universal welfare. This indirect effect was not
the two types of interdependence are independent constructs. significant [lower bound 0.16, upper bound 1.40]. In
To test our central hypothesis that concern for universal summary, these findings are consistent with our hypothesis that
welfare would be predicted by intergroup interdependence and intergroup interdependence uniquely predicts greater support
liberalism, we conducted two linear multiple regression analy- for human rights through liberalism. Although no support was
ses with intergroup interdependence and interpersonal interde- obtained for the opposite mediation pattern, experimental evi-
pendence predicting (1) liberalism and (2) support for human dence is required to definitively establish a causal relationship.
rights. As hypothesized, perceived intergroup interdependence This was done in Study 2.
significantly predicted political orientation, b .29, standard
error (SE) 0.13, p .022: The more participants believed
different groups are needed for society to work, the more they
identified as liberal. Perceived interpersonal interdependence
Study 2
also significantly predicted political orientation, but in the In this study, we used a word search task to prime either the indi-
opposite direction, b .56, SE 0.22, p .014: As pre- vidual or the collective and then measured peoples political
dicted, the more participants reported valuing conformity, ideology and support for universal welfare. Based on our findings
cohesion, and identification with their in-group, the more con- from Study 1, we predicted that a collective (vs. self-) focused
servative they were. A separate analysis predicting support for mind-set would lead to an increase in liberalism and that this
human rights from intergroup and interpersonal interdepen- effect would be mediated by increased intergroup interdepen-
dence yielded a significant positive association between dence. Because it is conceivable that the collective (vs. self)
perceived intergroup interdependence and support for human focused prime might also increase interpersonal interdependence
rights, b 3.43, SE 1.56, p .030, such that the more inter- (cf. Gardner et al., 1999), in which case we would expect to
group interdependence participants perceived, the more funds observe an increase in conservatism (as opposed to liberalism),
they allocated to human rights advocacy groups. Although the we also included a measure of interpersonal interdependence.
zero-order correlation between perceived interpersonal interde-
pendence and support for human rights was not significant (see
Table 1), the regression analysis that adjusted for intergroup Method
interdependence yielded a marginally significant association Participants. We recruited 223 White Americans from an online
between perceived interpersonal interdependence and support subject pool in exchange for entry into a cash prize lottery.
for human rights, b 2.67, SE 2.77, p .073. Thirteen participants did not complete the manipulation and

Downloaded from at Yale University Library on June 2, 2014

van der Toorn et al. 619

As in Study 1, we conducted follow-up bootstrapping anal-

yses to test for mediation. As hypothesized, liberalism signifi-
Liberalism cantly predicted support for human rights, b 3.42, SE 0.88,
0.29* (0.13) 6.18*** (1.19) p < .001. In addition, with liberalism in the model, the direct
effect of the prime on support for human rights was no longer
significant, b 3.91, SE 2.69, p .146, suggesting
mediation (Figure 2). Bootstrapping results yielded a signifi-
Intergroup Support for cant indirect effect [lower bound 0.38, upper bound
Interdependence Human Rights 4.11], indicating that the interdependence prime indeed causes
greater support for universal welfare through increased
Figure 1. Mediation model for Study 1 testing the indirect effect of Separate additional bootstrapping analyses demonstrated
intergroup interdependence on support for human rights through that perceived intergroup interdependence (the manipulation
liberalism. check) mediated the relationship between the prime and liber-
Note. All path coefficients represent unstandardized regression alism [lower bound 0.03, upper bound 0.26], and between
weights; *p < .05, **p < .01, ***p < .001. the prime and support for human rights [lower bound 0.13,
upper bound 3.99]. Finally, liberalism mediated the relation-
were excluded from analysis, yielding a final sample of 210 ship between perceived intergroup interdependence and
(age, M 37.59, SD 14.02; 71.4% female). support for universal welfare [lower bound 0.28, upper
bound 1.82]. These findings provide further support for our
Procedure. Participants were randomly assigned to a prime that contention that it is the perception that different groups are
activated either the self (independent) or the collective (inter- needed for society to work that increases policy support
dependent), adapted from Brewer and Gardner (1996). They through liberalism.
read a short paragraph describing a trip to the city and copied
all the pronouns below it. The self-focused prime included
first-person singular pronouns (e.g., I, me, my), while the
General Discussion
collective-focused prime included first-person plural pronouns We now realize, as we have never realized before, our interdepen-
(e.g., we, us, our). The paragraphs were otherwise identical. dence on each other. (Franklin D. Roosevelt, First Inaugural
After rating the pleasantness of the paragraph (1 very unplea- Address, March 1933)
sant, 5 very pleasant), participants responded to the exact
same measures of perceived intergroup interdependence, polit- As Roosevelt began his first term as President of the United
ical orientation, and universal welfare support as in Study 1. States during the depth of the Great Depression, lines between
They also rated their interdependence with the in-group by groups in society were blurred. This heightened sense of
using 1 item from the interpersonal interdependence scale used interdependence among Americans might have been the ideal
in Study 1 (Belonging to a group is important to my self- climate to promote Roosevelts liberal policies that established
identity, or sense of myself) on a 7-point scale. rights for diverse groups in society. Whereas previous research-
ers have speculated that people who are more liberal generally
see groups in society as more positively interdependent
Results and Discussion (Janoff-Bulman, 2009; Monaghan, 1984), the present research
Table 2 lists the variable Ms, SDs, and inter-correlations. There investigated the surprisingly understudied possibility that per-
were no differences in the pleasantness ratings of the para- ceptions of interdependence between groups produce stronger
graph between the two priming conditions, t(208) .54, liberal orientations. Indeed, the current research suggests that
p .591. As predicted, t-tests revealed that participants instilling a sense of intergroup interdependence (even through
primed with interdependence scored higher on liberalism as simple a prime as copying pronouns) can increase political
(M 4.65, SD 1.62) than participants primed with inde- liberalism and, in turn, foster concern for universal welfare.
pendence (M 4.15, SD 1.42), t(191) 2.35, p .020. Using both correlational (Study 1) and experimental (Study
In addition, participants in the interdependence condition 2) methodologies, we find convergent support for the novel
scored higher on support for human rights (M 49.01, hypothesis that perceived intergroup interdependence increases
SD 18.66) than participants in the independence condition peoples support for human rights because it increases liberal-
(M 43.39, SD 20.67), t(207) 2.07, p .040.3 ism. Of course, our findings do not preclude the possibility of
As expected, participants endorsed intergroup interde- bidirectionality in which we acknowledge that increases in lib-
pendence to a greater degree when primed with interdepen- eralism may also produce stronger feelings of intergroup
dence (M 5.49, SD 1.31) than independence (M 5.01, interdependence.
SD 1.48), t(208) 2.48, p .014; the manipulation did not In addition to establishing the hypothesized effect, we build
affect participants ratings of interpersonal interdependence, on moral foundations theory (Graham et al., 2009) and the
t(208) .39, p .698. approach/avoidance model of morality (Janoff-Bulman,

Downloaded from at Yale University Library on June 2, 2014

620 Social Psychological and Personality Science 5(5)

Table 2. Descriptives and Intercorrelations (Study 2).

Variable M (SD) 1 2 3 4 5

1. Interdependence manipulation 0.46 (0.50)

2. Intergroup interdependence 5.23 (1.42) .17*
3. Liberalism 4.38 (1.53) .16* .24***
4. Support for human rights 45.96 (19.92) .14* .23*** .28***
5. Gender 0.43 (0.91) .03 .01 .04 .15*
6. Age 37.59 (14.02) .03 .01 .09 .17* .12
Note. Interdependence manipulation was coded 0 for independence and 1 for interdependence. Gender was effect-coded (women 1; men 1). *p < .05,
**p < .01, ***p < .001.

condition that moved people in a liberal direction. Regardless

of the direction of the effect, the mediation analysis does con-
Liberalism firm that greater perceived intergroup interdependence leads to
liberalism (vs. conservatism), which is an important and novel
0.50* (0.21) 3.42*** (.88)
contribution to our understanding of the underpinnings of polit-
ical ideology.
It is conceivable that priming the collective we could
Intergroup Support for have differential effects depending on the contextfor
Interdependence Human Rights instance, it might be the case that if ideology (or some other
group) is first made salient, conservatives will think more
about their interdependence on an interpersonal level,
whereas liberals will think more about their collective, inter-
Figure 2. Mediation model for Study 2 testing the indirect effect of
group interdependence. Although more work is needed to
intergroup interdependence on support for human rights through fully understand how intergroup interdependence can be con-
liberalism. sistently evoked, our studies show that to the extent that peo-
ple do hold an appreciation for a diverse, multigroup society,
2009) by empirically distinguishing the effect of intergroup the more they will lean toward political liberalism and con-
interdependence from that of interpersonal interdependence, cern for human rights.
which instead was related to conservatism (Study 1). In Study Although a sense of interdependence may be elicited under
2, we used a simple we versus I prime to induce inter- various conditions, our work demonstrates the importance of
group interdependence versus independence. We confirmed more specifically examining peoples responses in times of
that the prime uniquely affected intergroup interdependence societal upheaval. While the conservative shift observed
and did not affect interpersonal interdependence, in our study. following 9/11 was likely due to the immense sense of threat
It should be noted, however, that this might not always be that was associated with it, our findings suggest that we may
the case. Research by Brewer and Gardner (1996) showed that well have seen increased liberalism should intergroup interde-
priming of we versus priming of they or it led partici- pendence have been activated instead. It remains an open ques-
pants to have more inclusive self-concepts and to generate tion whether priming of intergroup interdependence makes one
more interpersonal and collective self-descriptions. No com- generally open-minded or whether its effects are restricted to
parison was made to an I condition. However, although the facilitating liberal attitudes in areas related to human welfare.
authors did not test it for significance, participants in the we Future research might examine its effects on other attitudes and
condition generated more collective (18%) than interpersonal in situations where concern for human welfare is at odds with
(7%) self-descriptions (Brewer & Gardner, 1996, Study 3). other liberal stances (e.g., when protecting the environment is
Gardner, Gabriel, and Lee (1999) found that priming we at odds with more immediate human welfare concerns).
(vs. I) led to greater interdependent self-construals and In conclusion, the current research is scientifically innova-
social values, but they did not distinguish between interperso- tive in that it fills a gap in the literature on political attitude for-
nal versus collective self-construals. mation. Whereas previous research has mainly focused on the
One limitation of Study 2 is that we did not have a neutral circumstances that lead individuals to adopt more politically
control condition and thus cannot conclusively say whether conservative beliefs, the current work offers a novel perspec-
participants are in fact becoming more liberal in the interdepen- tive by demonstrating the effect of intergroup interdependence
dent condition, or whether they are becoming more conserva- on liberal attitudes. Beyond its theoretical contribution, the cur-
tive in the independent condition. However, the fact that rent research is of practical significance as well. One intriguing
participants independent self-concept scores were unrelated implication of these findings is that leveraging peoples sense
to their political orientation, as found in Study 1 (see Note of intergroup interdependence could impact social and political
1), is in line with the assumption that it is the interdependent attitudes. Framing a message that highlights the shared

Downloaded from at Yale University Library on June 2, 2014

van der Toorn et al. 621

responsibility of the many diverse groups in society should Dovidio, J. F., & Gaertner, S. L. (2010). Intergroup bias. In S. T. Fiske,
enhance support for human rights and related liberal policies. D. Gilbert, & G. Lindzey (Eds.), Handbook of social psychology
(Vol. 2, 5th ed., pp. 10841121). New York, NY: Wiley.
Declaration of Conflicting Interests Gabriel, S., & Gardner, W. L. (1999). Are there his and hers
types of interdependence? The implications of gender differences
The author(s) declared no potential conflicts of interest with respect to
in collective versus relational interdependence for affect, behavior,
the research, authorship, and/or publication of this article.
and cognition. Journal of Personality and Social Psychology, 77,
Funding Gaertner, L., Iuzzini, J., Witt, M. G., & Orina, M. M. (2006). Us with-
The author(s) received no financial support for the research, author- out them: Evidence for an intragroup origin of positive in-group
ship, and/or publication of this article. regard. Journal of Personality and Social Psychology, 90,
Notes Gardner, W., Gabriel, S., & Lee, A. (1999). I value freedom but
we value relationships: Self-construal priming mirrors cultural
1. We also included an item asking endorsement of the statement, I
differences in judgment. Psychological Science, 10, 321326.
feel dependent on other groups in society. This measure, which
Graham, J., Haidt, J., & Nosek, B. A. (2009). Liberals and conserva-
focuses on an individuals personal connection to other groups was
tives rely on different sets of moral foundations. Journal of Person-
moderately correlated with the measure of perceived interdepen-
ality and Social Psychology, 96, 10291046.
dence between groups in society, r(85) .43, p < .001, and not
Haddock, G., Zanna, M. P., & Esses, V. M. (1993). Assessing the
correlated with participants political orientation, r(85) .12,
structure of prejudicial attitudes: The case of attitudes toward
p .282, nor with their support for human rights, r(85) .11,
homosexuals. Journal of Personality and Social Psychology, 65,
p .327. Because this measure focused more on perceptions of
personal dependence on other groups rather than on the interdepen-
Janoff-Bulman, R. (2009). To provide or protect: Motivational bases
dence between groups, we did not combine it with our item asses-
of political liberalism and conservatism. Psychological Inquiry,
sing intergroup interdependence. Using the mean of this item and
20, 120128.
our intergroup interdependence item slightly weakens the results;
Janoff-Bulman, R., & Carnes, N. C. (2013). Surveying the moral land-
for example, its bivariate relationships with liberalism and support
scape: Moral motives and group-based moralities. Personality and
for human rights are marginally significant, r(85) .20, p .070
Social Psychology Review, 17, 219236.
and r(85) .20, p .078, respectively. We also collected partici-
Janoff-Bulman, R., Sheikh, S., & Baldacci, K. (2008). Mapping moral
pants personal independence ratings using Lu and Gilmours
motives: Approach, avoidance, and political orientation. Journal of
(2007) 21-item scale (a .82), but had no specific predictions for
Experimental Social Psychology, 44, 10911099.
them. Personal independence did not significantly correlate with
Jost, J. T., Federico, C. M., & Napier, J. L. (2009). Political ideology:
any of the other variables, all rs < .15. These analyses suggest that
Its structure, functions, and elective affinities. Annual Review of
intergroup interdependence, our variable of primary interest, is
Psychology, 60, 307333.
conceptually distinct from forms of personal (in)dependence.
Jost, J. T., Glaser, J., Kruglanski, A. W., & Sulloway, F. (2003). Polit-
2. As previous research indicated gender and age differences in inter-
ical conservatism as motivated social cognition. Psychological
dependent self-construal (e.g., Gabriel & Gardner, 1999; Watkins,
Bulletin, 129, 339375.
Mortazavi, & Trofimova, 2000), we also ran our analyses adjusting
Kemmelmeier, M., Burnstein, E., Krumov, K., Genkova, P.,
for participant gender and age. The pattern of results for intergroup
Kanagawa, C., Hirshberg, M. S., . . . Noels, K. A. (2003). Indi-
and interpersonal interdependence is nearly identical when adjusting
vidualism, collectivism, and authoritarianism in seven societ-
for these demographics, and the focal effects remain significant;
ies. Journal of Cross-Cultural Psychology, 34, 304322.
gender only marginally significantly predicted support for human
Landau, M. J., Solomon, S., Greenberg, J., Cohen, F., Pyszczynski, T.,
rights, b 4.17, SE 2.29, p .073, suggesting that women were
Arndt, J., . . . Cook, A. (2004). Deliver us from evil: The effects of
somewhat more in favor of human rights advocacy than men were.
mortality salience and reminders of 9/11 on support for President
All other ts < 1. We also tested whether gender interacted with inter-
George W. Bush. Personality and Social Psychology Bulletin,
group and interpersonal interdependence in predicting liberalism and
30, 11361150.
support for human rights, but it did not, all ts < 1.
Lu, L., & Gilmore, R. (2007). Developing a new measure of indepen-
3. Both tests had significant Levenes tests for variance equality, so
dent and interdependent views of the self. Journal of Research in
degrees of freedom were downward adjusted.
Personality, 41, 249257.
Magstadt, T. M. (2010). Understanding politics: Ideas, institutions,
References and issues. Belmont, CA: Thomson/Wadsworth.
Alterman, E. (2008). Why were liberals. New York, NY: Viking Markus, H. R., & Kitayama, S. (1991). Culture and the self: Implica-
Adult. tions for cognition, emotion, and motivation. Psychological
Brewer, M. B., & Gardner, W. (1996). Who is this we? Levels of Review, 20, 568579.
collective identity and self representations. Journal of Personality Milkis, S. M., & Mileur, J. M. (2005). The great society and the high
and Social Psychology, 71, 8393. tide of liberalism. Amherst: University of Massachusetts Press.

Downloaded from at Yale University Library on June 2, 2014

622 Social Psychological and Personality Science 5(5)

Monaghan, R. R. (1984). Political communication. Etc, 41, 416422. Thorisdottir, H., & Jost, J. T. (2011). Motivated closed-mindedness
Nail, P. R., McGregor, I., Drinkwater, A., Steele, G., & Thompson, A. mediates the effect of threat on political conservatism. Political
(2009). Threat causes liberals to think like conservatives. Journal Psychology, 32, 785811.
of Experimental Social Psychology, 45, 901907. Trafimow, D., Triandis, H. C., & Goto, S. G. (1991). Some tests of the
Oyserman, D., Coon, H. M., & Kemmelmeier, M. (2002). Rethinking distinction between the private self and the collective self. Journal
individualism and collectivism: Evaluation of theoretical assump- of Personality and Social Psychology, 60, 649655.
tions and meta-analyses. Psychological Bulletin, 128, 372. Watkins, D., Mortazavi, S., & Trofimova, I. (2000). Independent and
Preacher, K. J., & Hayes, A. F. (2004). SPSS and SAS procedures interdependent conceptions of self: An investigation of age,
for estimating indirect effects in simple mediation models. gender, and culture differences in importance and satisfaction rat-
Behavior Research Methods, Instruments, and Computers, 36, ings. Cross-Cultural Research, 34, 113134.
Rock, M., & Janoff-Bulman, R. (2010). Where do we draw our lines? Author Biographies
Politics, rigidity, and the role of self-regulation. Social Psychologi-
cal and Personality Science, 1, 2633. Jojanneke van der Toorn is an assistant professor in the Department of
Shaw, G. K. (1988). Keynesian economics: The permanent revolution. Psychology at Leiden University. Her research focuses on the social psy-
Aldershot, England: Edward Elgar. chological mechanisms implicated in social change and resistance to it.
Sherif, M., Harvey, O. J., White, B. J., Hood, W. R., & Sherif, C. W.
Jaime L. Napier is an assistant professor in the Department of
(1961). Intergroup conflict and cooperation: The Robbers Cave
Psychology at Yale University. Her research examines the social and
experiment. Norman: University of Oklahoma Book Exchange.
psychological antecedents and consequences of societal inequality.
Son Hing, L. S., Li, W., & Zanna, M. P. (2002). Inducing hypocrisy to
reduce prejudicial responses among aversive racists. Journal of John F. Dovidio is a professor in the Department of Psychology at
Experimental Social Psychology, 38, 7178. Yale University. His research interests are in stereotyping, prejudice,
Stel, M. A., & Harinck, F. (2011). Being mimicked makes you a pro- and discrimination; social power and nonverbal communication; and
social voter. Experimental Psychology, 58, 7984. altruism and helping.

Downloaded from at Yale University Library on June 2, 2014

View publication stats