Advantaged groups emotional reactions to intergroup inequality the dynamics of pride, guilt, and sympathy

For example, recent research has shown that the perception that one’s ingroup has illegitimate advantage is correlated with the unpleasant feeling of group-based guilt about this inequal

DOI: 10.1177/0146167207309193

2008; 34; 115

Pers Soc Psychol Bull

Nicole Syringa Harth, Thomas Kessler and Colin Wayne Leach

Sympathy Advantaged Group's Emotional Reactions to Intergroup Inequality: The Dynamics of Pride, Guilt, and

http://psp.sagepub.com/cgi/content/abstract/34/1/115

The online version of this article can be found at:

Published by:

http://www.sagepublications.com

On behalf of:

Society for Personality and Social Psychology, Inc

can be found at:

Personality and Social Psychology Bulletin

Additional services and information for

http://psp.sagepub.com/cgi/alerts

Email Alerts:

http://psp.sagepub.com/subscriptions

Subscriptions:

http://www.sagepub.com/journalsReprints.nav

Reprints:

http://www.sagepub.com/journalsPermissions.nav

Permissions:

http://psp.sagepub.com/cgi/content/refs/34/1/115

SAGE Journals Online and HighWire Press platforms):

(this article cites 28 articles hosted on the

Citations

to Intergroup Inequality: The Dynamics of

Pride, Guilt, and Sympathy

Nicole Syringa Harth

Thomas Kessler

Friedrich Schiller University

Colin Wayne Leach

University of Sussex

more and some of us belong to groups that have less.

Research has documented how members of groups that have less may react to their relative deprivation (e.g., Kessler & Mummendey, 2001; for a review, see Simon

& Klandermans, 2001) Collective protest, demonstra-tions, riots—in short, the motivation to challenge inequality—is a consequence of relative deprivation when individuals feel dissatisfied or angry about it Thus, the specific emotions that individuals feel about intergroup inequality seem to have important implica-tions for what they are willing to do about it Although research on relative deprivation has made this point for some time, there is now a growing body of evidence that the emotions individuals feel about their group mem-bership play an important role in explaining their will-ingness to act on behalf of their group or against other groups (e.g., Leach, Iyer, & Pedersen, 2006; Mackie, Devos, & Smith, 2000; for discussions, see E R Smith, 1993; H J Smith & Kessler, 2004)

Authors’ Note: This research was supported by a doctoral fellowship to

the first author in the International Research Training Group (GRK 622) titled “Conflict and Cooperation Between Social Groups” funded by the German Research Foundation (Deutsche Forschungsgemeinschaft) We thank Christopher Cohrs, Mirjam Dolderer, Ilka Gleibs, Natascha de Hoog, Timo Stich, Vincent Yzerbyt, and two anonymous reviewers for their helpful comments and discussions on an earlier draft Corres-pondence should be addressed to Nicole Syringa Harth, Friedrich-Schiller-Universität, International Graduate College, Wildstrasse 1, D-07743 Jena, Germany; e-mail: n.harth@uni-jena.de.

PSPB, Vol 34 No 1, January 2008 115-129

DOI: 10.1177/0146167207309193

© 2008 by the Society for Personality and Social Psychology, Inc.

Three studies establish intergroup inequality to

investi-gate how it is emotionally experienced by the

advan-taged Studies 1 and 2 examine psychology students’

emotional experience of their unequal job situation with

worse-off pedagogy students When inequality is

ingroup focused and legitimate, participants experience

more pride However, when inequality is ingroup

focused and illegitimate, participants experience more

guilt Sympathy is increased when inequality is

out-group focused and illegitimate These emotions have

particular effects on behavioral tendencies In Study 2

group-based pride predicts greater ingroup favoritism in

a resource distribution task, whereas group-based

sym-pathy predicts less ingroup favoritism Study 3

repli-cates these findings in the context of students’

willingness to let young immigrants take part in a

uni-versity sport Pride predicts less willingness to let

immi-grants take part whereas sympathy predicts greater

willingness Guilt is a weak predictor of behavioral

ten-dencies in all studies This shows the specificity of

emo-tions experienced about intergroup inequality.

Keywords: relative advantage; group-based emotions; pride;

guilt; sympathy; behavioral tendencies

Imagine yourself as a person living in an industrialized

country Now imagine another person, your age, your

gender, living in a developing country How do you feel

with this picture in your head? Prideful? Guilty?

Sympathetic? Given that human societies are structured

as systems of group-based hierarchies (Sidanius &

Pratto, 1999), some of us belong to groups that have

Although understanding of the antecedents and

con-sequences of relative deprivation has grown recently,

there is sparse knowledge about how people experience

belonging to a group that has more than other groups

However, just as in the case of relative deprivation,

there is good reason to think that how members of

groups with more perceive their position will affect how

they feel and what they are likely to do (Leach, Snider,

& Iyer, 2002) For example, recent research has shown

that the perception that one’s ingroup has illegitimate

advantage is correlated with the unpleasant feeling of

group-based guilt about this inequality (Iyer, Leach, &

Crosby, 2003, Study 1; Leach et al., 2006; Powell,

Branscombe, & Schmitt, 2005; Swim & Miller, 1999)

Additionally, quasi-experimental research has shown

that group-based guilt is increased when members of

real-world groups are made to focus their attention on

their ingroup’s preexisting illegitimate advantage (Iyer

et al., 2003, Study 2; Powell et al., 2005) This suggests

that self-focus and illegitimacy are appraisals of

inter-group inequality that lead to inter-group-based guilt

However, to our knowledge, no prior research has

exam-ined these two dimensions experimentally by creating

an intergroup inequality and manipulating its

illegiti-macy as well as group members’ focus of attention In

addition, no prior experimental research has contrasted

a self-focus on illegitimate intergroup inequality with

other conditions This is important because other

con-ditions of intergroup inequality may elicit different

emotions (Leach et al., 2002) For example, Iyer et al

(2003, Study 2) found that encouraging European

Americans to focus on African Americans’ illegitimate

disadvantage increased these European Americans’

sympathy and decreased their guilt Thus, in this article

we examine the degree to which pride, guilt, and

sym-pathy are distinct emotions about intergroup inequality

We extend previous research by establishing an

inter-group inequality and manipulating legitimacy and focus

To investigate the potentially distinct behavioral

tenden-cies triggered by pride, guilt, and sympathy, we examine

actual ingroup-favoring behavior (Study 2) and the

will-ingness to share resources with an outgroup (Study 3)

Focus and Legitimacy: Differentiating

Emotions About Intergroup Inequality

Individuals become aware of intergroup inequality

by comparing one group with another in terms of

resources, level of success, or other attributes (Tajfel &

Turner, 1986) A downward comparison that

estab-lishes one’s ingroup as better off than an outgroup

establishes the ingroup’s relative advantage and the

out-group’s relative disadvantage (Guimond & Dambrun,

2002; Leach et al., 2002) Recent theorizing suggests

that individuals may experience a variety of emotions about this kind of intergroup comparison and that each emotion will have specific implications for intergroup behavior Combining self-categorization theory (Turner, Hogg, Oakes, Reicher, & Wetherell, 1987) and appraisal theories of emotions (e.g., C A Smith & Ellsworth, 1985), E R Smith (1993) proposed that individuals’ appraisals of their ingroup’s relation to an outgroup determines which emotions they experience These emotions trigger specific behavioral tendencies (e.g., Frijda, Kuipers, & ter Schure, 1989) Based in this general perspective, Leach et al (2002) offered a con-ceptual model that identified four appraisal dimensions

as differentiating between group-based emotions about intergroup inequality In this article, we highlight two of these appraisal dimensions as we think that they are cen-tral to the differentiation of emotions among those who benefit from intergroup inequality: focus and legitimacy When combined, focus and legitimacy suggest the condi-tions under which members of groups that benefit from intergroup inequality will feel pride, guilt, and sympathy Intergroup relations are complex Thus, people usually

do not focus on intergroup relations in their entirety Appraisal theories of emotion (e.g., C A Smith & Ellsworth, 1985) suggest that individuals first appraise whether a situation affects the self or others This distinc-tion allows individuals to focus their attendistinc-tion on the rel-evant party—the self or others As a result, focus of attention may guide subsequent appraisals and emotions

In the context of an intergroup comparison where the ingroup is better off, a self-focus highlights the ingroup’s relative advantage over another group As detailed next, depending on the legitimacy of the intergroup inequality, this self-focus may result in feelings of guilt or pride (Leach et al., 2002) An other-focus in the case of an inter-group inter-group comparison where the ininter-group is better off highlights the outgroup’s relative disadvantage (Leach

et al., 2002) Other-focus should lead to the emotion of sympathy where the inequality is illegitimate (Iyer et al.,

2003, Study 2) Thus, framing the same intergroup inequality either as ingroup advantage or outgroup disad-vantage should lead to different emotions because this framing focuses attention on either the ingroup or the out-group However, it should also be clear that the legitimacy

of the intergroup inequality is an important determinant

of the specific emotions felt about intergroup inequality Legitimacy is known within social identity theory (Tajfel & Turner, 1986) as a sociostructural variable that indicates whether the intergroup relation should change According to this view, an illegitimate inter-group inequality makes both better- and worse-off groups aware of alternatives to the existing relation Other approaches to illegitimacy are similar but tend to emphasize the fairness and deservingness of intergroup

inequality (for a review, see Sidanius & Pratto, 1999).

Thus, where a group’s effort or ability determines their

position in an intergroup relation, this position can be

said to be legitimate For example, a group that has

received excellent education and training but is

disad-vantaged in job opportunities (compared to a less

well-educated group) suffers an illegitimate disadvantage

However, if the disadvantaged group made little effort

and was poorly educated, it would be legitimate for its

members to have worse job opportunities Legitimacy

of this sort is an important basis of emotion, whether at

the individual (e.g., C A Smith & Ellsworth, 1985) or

group (e.g., Kessler & Mummendey, 2001) level

Together, the combination of focus and legitimacy

dif-ferentiate the three group-based emotions of guilt,

sym-pathy, and pride about intergroup inequality

Pride

Evidence for group-based pride, and its association

with appraisal, is limited (for a review, see Leach et al.,

2002) For example, Cialdini’s (1976) notion of

“bask-ing in reflected glory” implies that individuals feel pride

in their ingroup’s success over an outgroup However,

none of this work has examined focus or legitimacy

appraisal or has directly assessed pride Although

indi-viduals evaluate themselves more positively when their

ingroup benefits from a legitimate advantage over an

outgroup (for a meta-analytic review, see Bettencourt,

Dorr, Charlton, & Hume, 2001), it is unclear whether

this is self-focused or experienced in terms of pride For

example, in a study that told psychology students that

their job prospects were better than those of other

students, Guimond and Dambrun (2002, Study 2) found

this relative advantage to have no effect on positive

affect or feelings of satisfaction, although it did lead to

more positive evaluation of the ingroup and prejudice

toward outgroups (see also H J Smith & Tyler, 1997)

Nevertheless, theory and research at the individual level

suggest that a self-focused framing of legitimate ingroup

advantage should stimulate group-based pride For

example, research at the individual level has shown that

pride is associated with the achievement of good

out-comes for the individual self (e.g., Rodriguez Mosquera,

Manstead, & Fischer, 2000; C A Smith & Ellsworth,

1985) that are thought to be deserved Thus, at the

group level, focusing on an ingroup’s achievement of a

legitimate advantage over an outgroup should lead to

group-based pride (Leach et al., 2002)

Existential Guilt

Several recent studies show that the perception that

one’s real-world ingroup has an illegitimate advantage

over an outgroup is associated with the feeling of group-based guilt (Iyer et al., 2003, Study 1; Leach et al., 2006; Powell et al., 2005; Schmitt, Behner, Montada, Müller,

& Müller-Fohrbrodt, 2000; Swim & Miller, 1999) When this feeling of guilt is based purely on one’s exis-tence within a group that has illegitimate advantages, it

can be referred to as existential guilt (e.g., Hoffman, 1976;

Schmitt et al., 2000) Existential guilt is experienced with-out individuals or their ingroup having responsibility for illegitimate acts against an outgroup Although previous research has examined the role of legitimacy or self-focus

on group-based guilt, no prior research has examined their combination In addition, as far as we are aware, all prior research has used quasi-experimental designs where the fea-tures of real intergroup inequalities were made more or less salient For example, Miron, Branscombe, and Schmitt (2006, Study 2) manipulated the legitimacy of gender inequality in pay by providing men with research evidence that women were either equal or lower in the abilities required for well-paid jobs When gender inequality was framed as illegitimate, men reported feeling more guilty about it Although this study is one of the only to manip-ulate the legitimacy of intergroup inequality, it did not address the self-focus also shown to be central to group-based guilt In addition, like most other studies of guilt, Miron et al did not contrast the conditions that lead to guilt against those that might lead to other emotions Sympathy

At the interpersonal level, it has been shown that a focus on others’ suffering motivates prosocial emotions and helping behavior (Batson et al., 2003) Sympathy seems to be the prevailing response to others’ misfortunes (Weiner, 1995) In one of the few studies at the inter-group level, Iyer et al (2003) showed that European Americans’ belief that African Americans suffered dis-crimination was associated with sympathy as well as support for greater equality However, like most previ-ous studies of emotion about intergroup inequality, par-ticipants were well aware of the outgroup’s disadvantage before the study Few studies have manipulated the focus of attention in a more controlled setting, where the basis of the inequality can be specified In addition, less is known about how legitimacy appraisals affect sym-pathy at the intergroup level Generally, symsym-pathy seems

to be stronger when others’ disadvantage is perceived to

be illegitimate (e.g., Montada & Schneider, 1989) Overview

Three studies were conducted to compare the group-based emotions of pride, existential guilt, and sympathy

in their distinctiveness for reacting to social inequality

Based in the theoretical model of Leach et al (2002)

and the existing literature, we predicted that self-focus

on a legitimate ingroup advantage would lead to greater

feelings of group-based pride in comparison to

illegiti-mate advantage or other-focus We also predicted that

self-focus on an illegitimate ingroup advantage would

lead to greater feelings of existential guilt whereas

other-focus on an outgroup illegitimate disadvantage

would lead to greater group-based sympathy

The Emotion–Behavior Link

Each emotional experiences of intergroup inequality

may trigger different behavioral tendencies (Mackie

et al., 2000) that either preserve or alter the status

rela-tion We expected group-based pride to be associated

with ingroup-favoring behavioral tendencies because

pos-itive evaluation of the ingroup enhances ingroup favoritism

(e.g., Verkuyten & Hagendoorn, 2002) Moreover, research

on nationalism indicates that nationalistic pride predicts

xenophobia (Cohrs et al., 2004) However, the

rela-tionship between guilt and behavioral tendencies is

more ambiguous Guilt seems to explain support of

abstract goals of compensation, but it does not appear to

motivate concrete forms of political action, such as

orga-nizing demonstrations (Leach et al., 2006) Therefore, we

expected existential guilt to be a relatively weak predictor

for the willingness to act to reduce the intergroup

inequality In contrast, feelings of sympathy have been

found to be a powerful source of helping behavior

(Eisenberg, 2003) Thus, in the context of intergroup

inequality, we expected sympathy to be associated with

behavioral tendencies to help the disadvantaged

PILOT STUDY

Method

Participants and Procedure

A pilot study was carried out to ensure that the

mate-rial would be appropriate to manipulate the appraisals

of focus and legitimacy Eighty-one students of the

University of Jena took part in this experiment, which

consisted of four conditions: ingroup focus (IGF)

mate, IGF illegitimate, outgroup focus (OGF)

legiti-mate, and OGF illegitimate

Written scenarios in the form of fake newspaper

arti-cles were used to establish intergroup inequality and to

manipulate focus and legitimacy Each version of the

article described the job situation for social scientists in

Germany Psychology students were told that their job

opportunities were better than those of social pedagogy

students In the IGF condition the intergroup inequality was framed as an ingroup advantage by stating that psychologists have better job opportunities and an aver-age income of 130% more than social pedagogues (cf Iyer et al., 2003) In the OGF condition the intergroup inequality was framed as an outgroup disadvantage It was said that social pedagogues have worse job oppor-tunities and an average income 70% less than psychol-ogists The manipulation of legitimacy of the inequality referred to quality of education In the legitimate condi-tions the education of psychologists was described as excellent and the education of social pedagogues as poor, and vice versa in the illegitimate condition Measures

A 3-point scale was used to check the focus

manipu-lation (1 = social pedagogues, 2 = both groups, 3 =

psy-chologists) Furthermore, participants had to indicate

the perceived fairness of the inequality between psy-chologists and social pedagogues with three closely related items anchored by a 7-point scale that varied

from 1 (very unfair) to 7 (very fair).

Results

A 2 (focus) × 2 (legitimacy) ANOVA on the focus manipulation check revealed a very large main effect of

focus, MOGF=1.03, MIGF=2.88, F(1, 77) = 1126.53, p <

.001, η²=.936, a very small main effect of legitimacy,

MLegitimate=2.05, MIllegitimate=1.93, F(1, 77) = 6.79, p < 05,

η² = 081, and a nonsignificant interaction, F(1, 77) =

2.89, p = 09 A 2 (focus) × 2 (legitimacy) ANOVA with

perceived fairness as the dependent variable revealed that participants in the legitimate condition found the

situa-tion to be more fair (M = 4.48) than participants in the illegitimate condition (M = 3.08), F(1, 78) = 30.00, p <

.001, η² = 279 Also, participants in the IGF conditions

perceived the situation as more fair (M = 4.50) than par-ticipants in the OGF conditions (M = 3.05), F(1, 78) = 31.99, p < 001, η² = 290 The interaction between focus and legitimacy was not significant, F < 1.

Discussion

We appeared to successfully manipulate participants’ focus of attention When an inequality was framed as an ingroup advantage, participants focused more on their ingroup than on the outgroup Although the legitimacy manipulation led participants to focus slightly more on the outgroup in the illegitimate condition, participants tended to focus on both groups irrespective of legitimacy Our manipulation of the legitimacy of the intergroup inequality also appeared successful Participants perceived

as more fair the inequality made legitimate by an

“excellent” education compared to the inequality made

illegitimate by a “poor” education The perceived

fair-ness of the intergroup inequality was also affected by

our manipulation of focus Consistent with previous

research (e.g., Iyer et al., 2003, Study 2; for a review, see

Mikula, 1993), participants appeared to justify the

inequality when it was IGF Perceiving intergroup

inequality as more fair when IGF seems likely to be a

defensive reaction to belonging to an ingroup that is

undeservedly advantaged (e.g., Leach et al., 2006; for a

review, see Leach et al., 2002) Indeed, people in

advan-taged positions may benefit from justifying inequality

(Sidanius & Pratto, 1999) The implication of this

apparent justification is that our manipulations of focus

and legitimacy may less easily provoke guilt If an IGF

leads participants to view the inequality as more fair,

this will undermine the effect of illegitimacy on guilt in

the IGF condition Thus, our manipulations may

gener-ate a conservative test of the basis of group-based guilt

Of course, if IGF leads participants to justify inequality,

our manipulations will more easily provoke pride The

following studies address this issue

STUDY 1

We investigated the group-based emotions of pride,

existential guilt, and sympathy among members of an

ingroup that is better off than an outgroup To expand

on previous studies that tested the separate effects of

focus and legitimacy, we investigated the joint effect of

these two variables Thus, we used the design validated

in the pilot study

Method

Design and Participants

This experiment reproduced exactly the four conditions

examined in the pilot study: IGF legitimate, IGF

illegiti-mate, OGF legitiillegiti-mate, and OGF illegitimate

Under-graduate psychology students from the Friedrich-Schiller

University, Jena (N = 48, 85% females, M age = 21 years,

range = 19–28 years) were randomly assigned to one of

the four conditions We chose the 2 × 2 design for the

sake of completeness, but we are mainly interested in the

conditions IGF legitimate, IGF illegitimate, and OGF

ille-gitimate, as these are the conditions for group-based

pride, guilt, and sympathy For the OGF legitimate

con-dition, we had no clear prediction and thus investigated

this condition in a more exploratory manner A focus on

legitimately disadvantaged others may trigger negative

feelings, such as disdain (Leach et al., 2002)

Procedure The questionnaires were administered after psychol-ogy lectures First, participants were asked to read one

of the fake newspaper articles pretested in the pilot study They were then asked to complete the measures detailed next Having completed the questionnaires, participants were debriefed and thanked for their par-ticipation with a chocolate bar

Measures

Intergroup inequality and identification We checked

our establishment of an intergroup inequality based on

a single item “Regarding the job situation, psychologists are relatively advantaged compared to social

peda-gogues” using a 7-point scale (1 = absolutely not, 7 =

absolutely) To check whether participants identified

with the ingroup, four items measured identification with psychology students (e.g., “I identify with psychol-ogy students,” α = 83) using a 7-point Likert-type scale These items were taken from Doosje, Ellemers, and Spears (1995)

Emotions Immediately after reading the fake

news-paper article, participants were asked to indicate their emotional state (“When thinking about the described

situation I feel ”) using a 9-point scale (1 = not at all,

9 = very intense) Three items measured pride (proud,

successful, happy), three items measured guilt (guilty, have a bad conscience, ashamed1), and two items mea-sured sympathy (sympathy, compassion) To explore the effects of the OGF legitimate condition, three items measured disdain (contemn, disdain, disgusted) The order of the emotion items was varied randomly within the experimental conditions

Results Preliminary Analyses

Recognition of inequality Participants’ perception of

intergroup inequality was significantly above the scale

midpoint (M = 5.21), t(47) = 10.16, p < 001 In fact, all

individuals perceived their ingroup as better off than social pedagogues Hence, the intergroup inequality we created was recognized A 2 (focus) × 2 (legitimacy) ANOVA revealed that the manipulations did not influ-ence psychology students’ perception of the intergroup

inequality (F < 1).

Identification Identification with psychologists was

well above the scale midpoint (M = 5.22), t(47) = 7.10,

p < 001 Moreover, there were no significant differences

between the experimental conditions in identification

(Fs < 2.19), indicating that randomization of

partici-pants was successful and that the manipulations had no

significant effect on identification

Factor analyses of group-based emotions The 11

emotion items were submitted to a principal-axis factor

analysis with oblimin rotation A four-factor solution

was obtained accounting for 60.31% of the common

variance The first factor was indicated by the disdain

items, the second factor was indicated by the sympathy

items, the third factor was indicated by the pride items,

and the fourth factor was indicated by the guilt items We

aggregated the items of each factor into four emotion

scales, which showed good internal consistencies: disdain

(α = 82), sympathy (α = 83), pride (α = 70), and guilt

(α = 76) Sympathy correlated with guilt (r = 26, p =

.07), but not with disdain (r = –.12, p = 41) or pride (r =

.07, p = 64) Disdain correlated with pride (r = 276, p =

.06) and guilt (r = 374, p = 009), and pride and guilt

were also positively correlated (r = 34, p = 02).

Testing Hypotheses: Contrast Analyses

To investigate our hypotheses about the combined

effect of focus and legitimacy on group-based pride,

existential guilt, and sympathy, we conducted three

planned contrast analyses Contrast analyses are the

appropriate method to test the predictions outlined in

the introduction Given that our focus manipulation

affected perceived fairness in the pilot study, using

planned contrasts rather than omnibus interaction tests

also better isolates the effects of our manipulations

There were three facets to our analyses

First, following Rosnow and Rosenthal (1996), we

treated our 2 × 2 design as a 1 × 4 design to test our three

main hypotheses For each of the three theoretical

predic-tions, a contrast was created that described the

hypothe-sized rank order of means regarding one group-based

emotion (A > B = C = D) This is represented in the focal

contrast with the coefficients 3-1-1-1 For example,

feel-ings of pride should be greatest when the focus is on the

ingroup and the inequality is legitimate (compared to the three other conditions) Thus, the condition in which we expected one of the three group-based emotions to be most intense was weighted with +3 and was compared to the other three conditions which were weighted with –1 Second, to check whether there is systematic variance other than that predicted, orthogonal contrasts were computed

in addition to the focal contrast (Abelson & Prentice, 1997) Orthogonal contrasts are important because they reveal whether there is residual variance that is not explained by the focal contrast If the hypothesis repre-sented in the focal contrast is correct, the focal contrast should be significant, and ideally, the orthogonal contrast should not be significant Given that there were four

exper-imental conditions, we had 2 df to compute two

orthogo-nal contrasts (001-1 and 0-211) Third, to account for the intercorrelations between the emotions, and thus general emotionality, we included the nonfocal emotions as covari-ates in the contrast analyses Means and standard devia-tions for the emotion measures are reported in Table 1

Pride The 3-1-1-1 contrast for group-based pride

had a significant effect, F(1, 40) = 11.61, p = 002, η² =

.225 As expected, participants in the IGF legitimate condition reported more pride than participants in the other three conditions However, the second orthogonal

contrast (0-211) was significant, F(1, 40) = 6.15, p =

.018, indicating that the variance was not fully explained by the focal contrast

Existential guilt The -13-1-1 contrast computed for

existential guilt had a significant effect, F(1, 40) = 5.09,

p = 03, η² = 113 As hypothesized, participants in the

IGF illegitimate condition reported greater existential

guilt than participants in the other three conditions.

Neither of the orthogonal contrasts was significant

Sympathy The -1-1-13 contrast computed for

group-based sympathy had a significant effect, F(1, 40) = 4.29,

p = 045, η² = 097 As expected, feelings of sympathy

were higher in the OGF illegitimate condition None of the orthogonal contrasts were significant

TABLE 1: Means and Standard Deviations for Study 1

Legitimate Illegitimate Legitimate Illegitimate

NOTE: IGF = ingroup focus; OGF = outgroup focus.

Disdain.A 2 (focus) × 2 (legitimacy) ANOVA was

con-ducted on disdain to investigate how participants would

experience an OGF legitimate inequality The ANOVA

showed that neither focus, F(1, 44) = 2.04, p = 16, nor

legitimacy, F(1, 44) = 2.26, p = 14, nor their interaction

(F < 1) had an effect on disdain for the outgroup.

Discussion

Participants identified strongly with the ingroup of

psychology students and acknowledged their

experimen-tally established advantage over the social pedagogy

students Considering identification as a precondition,

the stage was set for group-based emotions.2More

specif-ically, this study supports our hypotheses regarding

dis-tinct emotional experiences of intergroup inequality

Pride was greatest when participants were focused on a

legitimate ingroup advantage whereas existential guilt

was greatest when participants were focused on an

ille-gitimate ingroup advantage Sympathy was most elicited

when participants were made to focus on the outgroup’s

illegitimate disadvantage It should be noted, however,

that one orthogonal contrast for pride was also

signifi-cant According to Abelson and Prentice (1997), this

indi-cates that additional explanation is needed to fully

account for the pattern of results However, this need for

additional explanation in no way undermines the

empir-ical support for our hypothesized explanation of pride

Disdain did not seem to be a typical emotional

response to an OGF legitimate inequality This may

have been due to the relative obscurity of these emotion

words In any case, in the following studies we

concen-trate on our three main hypotheses and leave out the

OGF legitimate condition

STUDY 2

The first aim of Study 2 was to replicate the effects for

group-based pride, existential guilt, and sympathy In

con-trast to Study 1, we used a one-factorial design with three

conditions to investigate the joint effect of focus and

legiti-macy Thus, we excluded the OGF legitimate condition that

yielded little in Study 1 The second aim was to investigate

the link between emotion and behavioral tendencies To

assess behavioral tendencies toward the ingroup and

out-group, a resource distribution task was used Although we

expected overall ingroup bias in the distribution of resources

to the two groups, we expected pride, existential guilt, and

sympathy to differentially predict resource distribution As

argued in the introduction, group-based pride, in contrast

to guilt and sympathy, should predict behavioral tendencies

that favor the ingroup In contrast, existential guilt

and sympathy should predict less ingroup favoritism in the

distribution of resources

Method Participants, Design, and Procedure Psychology students of the Universities of Dresden and Jena were recruited during lectures and took part

in the study immediately after the lectures (N = 61, 90% female, M age =21 years, range = 18–30 years).

Participants were randomly assigned to one of the three key conditions examined in Study 1: IGF legiti-mate, IGF illegitilegiti-mate, and OGF illegitimate The par-adigm and the fake newspaper articles were the same

as in Study 1

Measures

Manipulation checks Recognition of inequality and

group identification (α = 74) were measured in the same way as in Study 1

Emotions Guilt (α = 76) and sympathy (α = 70)

were measured exactly as in Study 1 For pride, how-ever, we substituted the item superior for happy As such, the pride scale now more narrowly focuses on the experience of pride associated with success and superi-ority (i.e., proud, successful, superior; α = 75) The eight emotion items were submitted to a principal-axis factor analysis with oblimin rotation A three-factor solution was obtained, which accounted for 57.20% of the common variance The first factor was indicated by the guilt items, the second factor was indicated by the pride items, and the third factor was indicated by the

sympathy items Guilt correlated with sympathy (r = 286, p = 03) but not with pride (r = –.114, p = 38); pride and sympathy did not correlate (r = –.133,

p = 30).

Behavioral tendencies: resource distribution To

assess participants’ behavioral tendency to distribute resources to the ingroup and outgroup, we instructed them to imagine they could influence the financial com-pensation of social scientists on the job market Their task was to distribute 100 monetary units between psy-chologists and social pedagogues They recorded their distribution in spaces provided in the questionnaire

Results Preliminary Analyses

Recognition of intergroup inequality Participants’

acknowledgment of group advantage was significantly

above the scale midpoint (M = 5.57), t(60) = 13.10, p <

.001 Moreover, a one-factorial ANOVA revealed that the acknowledgment was not significantly influenced by

the manipulation, F(2, 58) = 1.84, p = 17.

Identification Identification with the ingroup of

psy-chology students was significantly above the scale

mid-point (M = 4.94), t(60) = 6.20, p < 001 Importantly,

the manipulation did not affect identification with the

ingroup, F(2, 58) = 1.16, p = 32.

Group-Based Emotions

To test our hypotheses about the combined effect of

focus and legitimacy on the three group-based

emo-tions, we conducted planned contrast analyses with the

contrast coefficients 2-1-1 Given that we had three

experimental groups, there was only 1 df to compute

one orthogonal contrast Again, the two nonfocal

emo-tions were included as covariates in the analyses Means

and standard deviations are reported in Table 2

Pride The focal contrast for pride (2-1-1) was

signif-icant, F(1, 56) = 10.72, p = 002, η² = 161 As expected,

participants in the IGF legitimate condition reported

greater pride than in the other two conditions

Existential guilt The focal contrast for guilt (-12-1)

was significant, F(1, 56) = 4.52, p = 038, η² = 075 In

line with our prediction, participants in the IGF

illegiti-mate condition indicated higher guilt than participants

in the other two conditions

Sympathy The focal contrast for sympathy (-1-12)

was significant, F(1, 56) = 23.15, p < 001, η² = 292 As

expected, participants in the OGF illegitimate condition

indicated stronger sympathy than participants in the

other two conditions

In all three cases, the orthogonal contrast analyses

were not significant, indicating that there was no

sys-tematic residual variance left unexplained (all Fs < 2.9).

Hence, the predicted patterns were the most reasonable

and parsimonious This is an improvement over the

findings for pride in Study 1

Behavioral Tendencies: Resource Distribution Overall, participants gave more monetary units to the

ingroup (M = 55.97) than to the outgroup (M = 44.03),

t(60) = 6.27, p < 001 Thus, participants showed ingroup

favoritism in resource distribution As participants were asked to distribute 100 units between the ingroups and outgroups, the amount distributed to one group comple-mented perfectly that given to the other group Thus, we used distribution to the ingroup as the dependent mea-sure A contrast analysis with the coefficients 2-1-1 and the mean of the ingroup allocation as dependent variable should test whether participants in the IGF legitimate con-dition showed more ingroup-favoring behavior in the resource distribution task than participants in the other two conditions In fact, participants in the IGF legitimate condition distributed more resources to their ingroup

(M = 60.0) than did participants in the IGF illegitimate (M = 54.4) or OGF illegitimate (M = 53.6) conditions,

F(1, 58) = 9.89, p = 003, η² = 146 The difference

between the IGF illegitimate and OGF illegitimate

condi-tions was not significant (F < 1).

To recap, we expected pride to predict greater ingroup-favoring behavior in the resource distribution task whereas we expected guilt and sympathy to lead to less ingroup-favoring behavior A multiple regression analysis was conducted The three group-based emotions were simultaneously entered as predictors of ingroup allocation As expected, higher pride about intergroup inequality predicted greater allocation of resources to a

participant’s ingroup (β = 45, p < 001) Guilt did not

predict allocation behavior in the distribution task (β =

–.13, p > 10) Sympathy, however, led to marginally less ingroup-favoring allocations (β = –.22, p = 065).

Mediation Analysis: Group-Based Pride

In line with our prediction, IGF legitimate advan-tage led to the experience of pride as well as to

TABLE 2: Means and Standard Deviations for Studies 2 and 3

Study 2

Study 3

NOTE: IGF = ingroup focus; OGF = outgroup focus.

ingroup-favoring behavior in the resource distribution

task Thus, we used multiple regression analyses to test

whether the relation between IGF legitimate advantage

and ingroup-favoring behavior was mediated by pride

As existential guilt and sympathy did not predict the

behavioral tendency, these emotions did not meet the

requirements of mediation and thus were not analyzed

as such (see Baron & Kenny, 1986) However, we did

account for guilt and sympathy in the mediation models

by including them as covariates

First, a regression analysis with the IGF legitimate

con-dition, in contrast to IGF illegitimate and OGF illegitimate

conditions, as predictor (2-1-1) and the ingroup allocation

of the resource distribution task as a criterion showed that

both variables were correlated (β = 38, p < 002) Second,

IGF legitimate (2-1-1) led to greater pride (β = 40, p =

.003) and less sympathy (β = –.21, p = 120), and did not

reliably affect guilt (β = –.11, p = 450) Third, the

alloca-tion measure was regressed on experimental condialloca-tion,

pride, guilt, and sympathy simultaneously Only the

rela-tionship between pride and the behavioral measure was

significant (β = 39, p = 002); the relationship between the

experimental condition and the allocation behavior

became nonsignificant (β = 17, p = 170) Neither guilt

(β = 001, p > 10) nor sympathy (β = –.18, p = 130)

inde-pendently predicted the resource distribution The

boot-strap3 confidence interval indicated the indirect effect of

pride was significant (.036 to 3.58) This provides

evi-dence for pride as a mediator between legitimate IGF

advantage and ingroup-favoring behavior The confidence

intervals of guilt (–.208 to 229) and sympathy (–.101 to

.854) included zero; thus, their indirect effects were not

significant Figure 1 depicts this mediation

Discussion

The results of Study 2 replicated those of Study 1

Again, an IGF on legitimate advantage led to the

great-est pride, whereas an IGF on illegitimate advantage led

to the greatest guilt In contrast, an OGF on illegitimate

disadvantage led to the greatest sympathy The results

also shed light on the relation between group-based

emotions and behavioral tendencies Most notably, the

more participants expressed pride, the stronger was

their ingroup favoritism in the distribution of resources

In fact, pride mediated the effect of IGF and legitimacy

on this behavioral tendency In contrast, existential guilt

did not predict how participants acted in the resource

distribution task Although it was not a mediator,

sym-pathy had the opposite effect of pride, leading to less

ingroup favoritism The relatively minor role of

sympa-thy is likely due to the fact that this emotion should

better explain help giving than ingroup favoritism This

issue was addressed in the following study

STUDY 3

This study differed from both previous studies in three respects First, we aimed at testing whether the same results would occur in a different intergroup relation Thus, students of the University of Jena were told they are relatively advantaged compared to young ethnic German immigrants4 living in Jena with regard to sport opportunities Although this advantage was perceived to

be just as great as in the previous studies, the sport domain is likely less important than jobs and thus is likely

to be a less strong basis of emotion As such, Study 3 may provide a more subtle test of our hypotheses Second, a different kind of manipulation was used Instead of manipulating focus through the wording of a text that also included information about legitimacy, focus was realized through direct instruction This method should improve the orthogonality of the focus and legitimacy manipula-tions Third, we examined more concrete behavioral tendencies that allowed participants to help the outgroup

or to deny them equal opportunities

Method Participants and Design

Jena University students (N = 84, 76% female, M age =

21.30 years, range = 18–36 years), who participated under circumstances similar to that of the first two studies, were randomly allocated to the three experimental conditions: IGF legitimate, IGF illegitimate, and OGF illegitimate Procedure and Material

A cover story was developed describing the situation between students living in Jena and young ethnic German immigrants living in Jena Participants read about a project supported by a federal ministry called

Figure 1 Mediation model of the association between ingroup (IG)

focus on legitimate advantage and the resource distribu-tion mediated by group-based emodistribu-tions, Study 2.

*p < 05 **p < 01.