A vengefulness of the impotent” the pain of in group inferiority and schadenfreude toward successful out groups

Individuals’ pain about their in-group’s inferiority explained schadenfreude at the failure of a successful out-group better than dislike of the out-group, interest in the domain of comp

“A Vengefulness of the Impotent”: The Pain of In-Group Inferiority and

Schadenfreude Toward Successful Out-Groups

Colin Wayne Leach

Cardiff University and University of Amsterdam

Nietzsche (1887/1967) suggested that the emotional pain individuals feel about their in-group’s inferi-ority leads them to feel the pleasure of schadenfreude when a successful out-group fails To test this idea,

2 studies examined a fictitious competition between real universities Individuals’ pain about their in-group’s inferiority explained schadenfreude at the failure of a successful out-group better than dislike

of the out-group, interest in the domain of competition, illegitimacy of the out-group’s success, and illegitimacy of the in-group’s inferiority In addition, emotions regarding the out-group’s success (i.e., envy, dislike-based anger, and illegitimacy-based anger) were weaker explanations of schadenfreude than

the pain of in-group inferiority and anger based in this pain (which Nietzsche referred to as ressentiment).

Thus, schadenfreude has more to do with the inferiority of the self than with the success of others As well as providing evidence for a specific form of prejudice grounded in group-based emotions, this research also revives displacement explanations of prejudice toward 3rd parties

Keywords:schadenfreude, emotion, intergroup relations, social identity theory, social status

How one’s in-group compares with out-groups is an important

determinant of one’s psychological experience For example, an

in-group’s comparative inferiority can pose a serious threat to

one’s self-worth (Tajfel & Turner, 1979), producing painful

emo-tions such as shame and frustration (for reviews, see Scheff, 1994;

Tiedens & Leach, 2004) In-group inferiority can lead individuals

to derogate, devalue, or compete against superior out-groups in an

attempt to reverse their fortunes (Spears, Jetten, & Doosje, 2001;

Tajfel & Turner, 1979; Walker & Smith, 2002) However, where

there is objective evidence of one’s in-group being inferior, this

social reality can constrain malevolence toward and competition

with the superior out-group (Sidanius & Pratto, 1999; Spears et al.,

2001; Tajfel & Turner, 1979) Indeed, in a meta-analytic review,

Bettencourt, Dorr, Charlton, and Hume (2001) found that

in-groups that were objectively inferior to an out-group in a particular

domain tended not to devalue the out-group in this domain

Because it is difficult for inferior in-groups to be malevolent

toward comparatively superior out-groups, it has long been

sug-gested that in-group inferiority leads to “displaced” malevolence

toward third parties (e.g., Fromm, 1941; for reviews, see Allport,

1954/1979; Wills, 1981) This perspective is consistent with the

more general view, advocated by Allport, among others, that the self’s subjective experience is an important basis of malevolence toward out-groups (see Brewer, 1979; Tajfel & Turner, 1979) For example, prejudice is associated with individuals’ subjective ex-perience of deprivation relative to other individuals (for a review, see Walker & Smith, 2002) and the threat of individual mortality (e.g., Greenberg et al., 1990)

Although the idea of displacement is central to popular notions of frustration–aggression and “poor White racism” in prejudice, displacement has rarely been observed in research on intergroup malevolence (see Green, Glaser, & Rich, 1998; Sidanius & Pratto, 1999) Partly in response to this lack of evidence, Leach, Spears, Branscombe, and Doosje (2003) iden-tified schadenfreude (i.e., pleasure at the misfortune that hap-penstance causes another party) as a passive form of intergroup malevolence that can result from in-group inferiority In two studies, they showed that objective in-group inferiority led individuals to feel schadenfreude when an otherwise successful third-party out-group failed in the domain of the in-group’s inferiority We extend this research by proposing that objective in-group inferiority leads to schadenfreude toward third parties mainly because of the subjective experience of painful emotions associated with the psychological threat of in-group inferiority (e.g., shame and frustration) Thus, building on the notion of intergroup emotion (E R Smith, 1993), we conceptualize in-tergroup schadenfreude as an unfolding emotional episode whereby unpleasant emotions about the self lead to a pleasant emotion about another party This mechanism by which in-groups address their painful inferiority by taking pleasure in an out-group’s failure revives the classic view of prejudice as a result of displaced feelings about the self

In contrast to our emphasis of the pain of in-group inferiority, most previous schadenfreude research has emphasized other-focused ex-planations For example, previous work has argued that emotional

Colin Wayne Leach, Department of Psychology, University of Sussex,

Brighton, England; Russell Spears, School of Psychology, Cardiff

Univer-sity, Cardiff, Wales, and Department of Social Psychology, University of

Amsterdam, Amsterdam, the Netherlands

This research was supported by a visiting scholar grant to Colin Wayne

Leach from the Nederlandse Organisatie voor Wetenschappelijk

Onder-zoek We thank Nesrien Abu-Ghazaleh and Roos Knap for their research

assistance We thank Patricia M Rodriguez Mosquera, Richard H Smith,

Ngaire Donaghue, the University of Amsterdam intergroup lab, and the

Leiden University social and organizational psychology research group

Correspondence concerning this article should be addressed to Colin

Wayne Leach, Department of Psychology, Pevensey 1 Building, University

of Sussex, Brighton BN1 9QH, England E-mail: c.w.leach@sussex.ac.uk

1383

reactions to another party’s success, such as envy (e.g., R H Smith et

al., 1996), illegitimacy-based anger (Feather & Sherman, 2002), and

dislike-based anger (Hareli & Weiner, 2002) lead to schadenfreude

when this party subsequently fails These other-focused explanations

of schadenfreude are consistent with a perspective that views an

out-group’s characteristics (e.g., its status and how it is stereotyped) as

the main explanations of prejudice toward them

Given that schadenfreude may result from either self- or

other-focused explanations, we offer a comprehensive model that

ad-dresses both Because of the importance of the self in prejudice, we

expect the pain of in-group inferiority to best explain

schaden-freude However, envy, illegitimacy, and illegitimacy-based anger

may also play a role Thus, in two studies, we compared the power

of self- and other-focused explanations of schadenfreude by

lead-ing students to believe that their in-group was inferior in an

intergroup competition and that an out-group successful in the

competition ultimately failed

Our emphasis on the pain of in-group inferiority as a

self-focused explanation of schadenfreude follows from Nietzsche’s

(1887/1967, p 37) discussion of the “vengefulness of the

impo-tent.” Consistent with contemporary research (Bettencourt et al.,

2001; Sidanius & Pratto, 1999; Spears et al., 2001), he argued that

an in-group’s inferiority to an out-group rendered it impotent to

compete against this out-group Also consistent with research at

the individual level (e.g., R H Smith, Parrott, Ozer, & Moniz,

1994; Tangney, Wagner, Fletcher, & Gramzow, 1992) and the

group level (Scheff, 1994; Walker & Smith, 2002), Nietzsche

argued that (objective) inferiority fuels the (subjective) pain of

shame and frustration about one’s threatened self-worth He

thought the pain of inferiority so aversive that individuals were

sure to take advantage of easy opportunities to feel more

posi-tively; the pleasure of schadenfreude at the failure of third parties

presents one such opportunity Thus, Nietzsche’s approach to schadenfreude presaged the psychodynamic view of prejudice as the displacement of painful feelings about the self, such as in-group inferiority

Leach et al (2003) were the first to empirically confirm the idea that objective in-group inferiority leads to schadenfreude toward third parties who played no role in establishing this inferiority However, Leach et al did not examine the subjective experience

of the pain of in-group inferiority as an explanation of this effect Therefore, in both of the present studies, we led students to believe that their (university) in-group was inferior in a domain as a result

of repeated failure against other universities We then assessed the emotional pain felt about the in-group’s inferiority in this domain

We expected this group-based emotion to be an important expla-nation of schadenfreude toward a successful third party who was subsequently presented as failing in the same domain as the in-group’s inferiority This route to schadenfreude is shown in the black path of the emotional episode illustrated in Figure 1

We have emphasized individuals’ pain about their in-group’s inferiority in a domain as an explanation of schadenfreude Al-though this inferiority is established independently of the success

of the eventual target of schadenfreude, encountering a successful out-group can also imply the in-group’s comparative inferiority (R H Smith, 1991; Tajfel & Turner, 1979) Based in part on social comparison theory, R H Smith et al (1996) suggested that the pain of the inferiority implied by another’s success is an important explanation of schadenfreude However, evidence for this is lack-ing, mainly because the pain of implied inferiority has been con-flated with other-focused explanations of schadenfreude, such as envy and anger about illegitimate success (e.g., R H Smith et al., 1996; van Dijk, Ouwerkerk, Goslinga, Nieweg, & Gallucci, 2006; for discussions, see Leach, 2008; Spears & Leach, 2004) To

Schadenfreude:

Out-group Failure

Pain of Inferiority Implied

Illegitimacy:

Out-group Success

Envy: Out-group Success

Anger: Out-group Success Dislike:

Out-group

In-group Domain Inferiority

Out-group Success

Pain of Inferiority Domain

Out-group Failure

Pathways

Dislike-based anger

Illegitimacy-based anger

Pain of In-group Inferiority

Figure 1. Model of emotion explanations of schadenfreude toward a successful out-group

clarify this issue, we propose a specific route by which the pain of

inferiority implied by another’s success leads to schadenfreude

through a particular type of anger that Nietzsche (1887/1967)

called ressentiment.

Nietzsche (1887/1967) argued that the pain individuals feel

about their in-group’s inferiority in a domain leads them to

inter-pret a third party’s success as implying their comparative

inferi-ority to this out-group Like the broader domain inferiinferi-ority that

feeds it, the inferiority implied by an out-group’s success should be

emotionally painful Nietzsche argued that those suffering implied

inferiority “externalized” their pain in the form of anger at the

successful: ressentiment In this way, ressentiment uses anger at

those who are successful to hide its true source—the pain of

implied inferiority Although there have been no direct

examina-tions of ressentiment, some contemporary work is consistent with

Nietzsche’s thinking Scheff (1994) used archival evidence to

suggest that in-groups who suffer inferiority-based shame are

prone to malicious anger toward out-groups In her clinical work

with individuals, Lewis (1971) observed that shame could be

redirected away from the self into anger at others (see also R H

Smith et al., 1994; Tangney et al., 1992) In a recent review,

McDonald and Leary (2005) confirmed that being devalued or

rejected by others is associated with indirect aggression in

inter-personal relations (see also Wills, 1981) Thus, there is good

reason to expect that the pain of inferiority implied by an

out-group’s success can lead to the externalized anger of ressentiment

and, in turn, the malevolence of schadenfreude In the current

context, ressentiment toward a successful third party should

me-diate the association between the pain of implied in-group

inferi-ority and schadenfreude toward the third party Thus, we expect the

pain of implied inferiority to have no direct effect on schadenfreude

independent of the anger through which it is externalized (i.e.,

res-sentiment) This pathway is shown in solid lines in Figure 1.

Although most previous approaches to schadenfreude have

aimed to offer general explanations of the emotion, most have

focused exclusively on schadenfreude at the interpersonal level

There is general agreement in this work that schadenfreude is felt

in response to the misfortune of a party who is successful in a

domain of some relevance to the self (Spears & Leach, 2004) In

addition, most previous theory has argued that schadenfreude is

best explained by individuals’ reactions to another party’s success,

such as dislike, illegitimacy of the other party’s success, envy, and

anger However, there is great disagreement about which of these

other-focused explanations is most important

Studies at both the interpersonal (R.H Smith et al., 1996) and

the intergroup (Leach et al., 2003) levels have shown preexisting

dislike to be a modest explanation of schadenfreude toward the

disliked party Thus, both of the present studies assessed

preexist-ing dislike of an out-group and then subsequently presented this

party as suffering a failure In addition, Study 1 measured the

perceived illegitimacy of the out-group’s success and Study 2

manipulated the procedural illegitimacy of the out-group’s

suc-cess Since antiquity, it has been argued that the illegitimacy of

another party’s success leads to schadenfreude when this party

suffers a misfortune (R H Smith, 1991) Nevertheless, we expect

the pain of in-group inferiority to better explain schadenfreude

than dislike of the out-group or the illegitimacy of its success

Although early research on schadenfreude suggested envy as an

important explanation (R H Smith et al., 1996), subsequent work

that has differentiated envy from perceived injustice and anger offered mixed evidence of an envy–schadenfreude link (Feather & Sherman, 2002; Hareli & Weiner, 2002; van Dijk, Ouwerkerk, Goslinga, Nieweg, and Gallucci, 2006) Because variation in the way envy has been conceptualized and measured may account for this inconsistency, we measured envy as coveting another party’s success and distinguished it from all of the other possible expla-nations of schadenfreude that have been included within the envy construct (i.e., self’s inferiority implied by the other’s success, perceived illegitimacy of inferiority, perceived illegitimacy of other party’s success, and anger; for a discussion, see Leach, 2008) When the pain of inferiority is distinguished from envy, we expect the former to be a much more potent explanation of schadenfreude than the latter

Numerous authors have argued that anger at successful parties can explain schadenfreude at their subsequent failure However, there are competing views regarding the basis of this anger (for a discussion, see Leach, 2008) Dislike of a successful party (Hareli

& Weiner, 2002) and the illegitimacy of their success (Feather & Sherman, 2002) have both been suggested as bases of anger that lead to schadenfreude We build on Hareli and Weiner (2002) by examining the explanatory process they implied: preexisting dis-like of out-group 3 anger at out-group’s success 3 schaden-freude In addition, we examine the pathway implied by Feather and Sherman (2002): out-group’s illegitimate success 3 anger at out-group’s success 3 schadenfreude

As outlined above, we expect the two routes of the pain of in-group inferiority to provide more potent explanations of schadenfreude toward third parties than the other-focused expla-nations suggested in the literature As emotional reactions to another party’s success are not based in the serious threat to the self that is posed by in-group inferiority, we expect the explanatory power of these other-focused explanations of schadenfreude to be relatively modest We also expect the two routes of the pain of in-group inferiority to explain schadenfreude better than the self-focused explanation of interest in the domain of competition (examined in Leach et al., 2003)

To test these ideas, both studies led students to believe that their senior peers were engaged in an interuniversity competition This allowed us to present false feedback of inferior in-group perfor-mance against a pool of out-groups in this domain The competi-tion also allowed us to establish independently a relevant out-group as successful in the domain This out-out-group was presented as successful through their repeated victories against a pool of uni-versities different from those who competed against the in-group However, the third party out-group was presented as ultimately failing in the final of the competition This gave in-group members

an opportunity for schadenfreude toward a successful out-group who played no role in establishing the in-group’s inferiority

Study 1 Creating a fictitious competition between two pools of univer-sities in the Netherlands allowed us to establish in-group inferior-ity independent of third party out-group success in the same domain We took advantage of this to manipulate the causal antecedents of schadenfreude In a baseline condition, we estab-lished in-group inferiority and then out-group success and assessed related feelings before the schadenfreude opportunity presented by

the out-group’s eventual failure Thus, both self- and other-focused

explanations could cause schadenfreude We expected the

self-focused explanations to have the biggest effects on schadenfreude,

supporting our argument that schadenfreude is based mainly in

concern for the self, rather than the other

In a second condition, we established out-group success and

as-sessed related feelings and perceptions (i.e., dislike, perceived

illegit-imacy of the success, envy, illegitillegit-imacy-based anger, and

dislike-based anger) before the schadenfreude opportunity Thus, this

condition examined the independent effects of the other-focused

ex-planations of schadenfreude Although this condition advantaged

these other-focused explanations, we expected them to have only

modest effects In contrast, we expected the pain of the inferiority

implied by the out-group’s success and its attendant anger (i.e., the

We also used the second condition to rule out the possibility that

schadenfreude and the pain of domain inferiority have a recursive

relationship If the pain of domain inferiority and schadenfreude

are recursively related, greater schadenfreude at an out-group’s

failure should lead to greater pain about in-group domain

inferi-ority in the condition in which this causal order is possible

However, evidence of a recursive relationship would undermine

our view that the pain of domain inferiority is an important cause

of schadenfreude, as either feeling could lead to the other Thus,

we expect a nonrecursive relation such that the pain of domain

inferiority leads to schadenfreude, but not vice versa

Method Participants and Procedure

Participants were 102 psychology students at the University of

Amsterdam (34 men and 68 women) They were told that we were

interested in their evaluation of an interuniversity quiz that had

been in existence for a number of years, but had never been

publi-cized We told participants that a television station was now

consid-ering making the annual competition into a regular program, but

wanted to know if the target audience of students was sufficiently

interested Thus, we presented our study as market research

After explaining the ostensible purpose of the research, we

presented participants with a 1-page description of the

interuni-versity quiz competition called “IQ.” They were told that

univer-sities throughout the country competed each year in two pools of

seven teams each At the end of each year, the winner of each pool

met in a final match to determine the national champion The

two-pool system enabled us to independently establish the

inferi-ority of the in-group (because of very poor performance in Pool A)

and the success of the out-group (because of very good

perfor-mance in Pool B)

In-Group Inferiority

Participants were presented with a table that gave their

in-group’s performance in Pool A It showed the participant’s

uni-versity to have lost to Open, Delft, Leiden, Groningen, and Tilburg

Universities, but to have won against Eindhoven Because they

were reported to have lost five of the six matches they played, the

in-group was objectively inferior in the domain This also made it

clear that the in-group’s inferiority in their pool prevented them from competing in the final against the top team in Pool B

Out-Group Success

A second table gave the out-group’s performance in Pool B The out-group was presented as successful in the domain by virtue of the fact that they won five of the six matches they played in their pool Thus, the VU University Amsterdam lost against Utrecht, but won against Maastricht, Wageningen, Nijmegen, Rotterdam, and Twente Universities It was then explained that the VU University’s success put them in the final against the winner of Pool A, the University of Groningen Thus, participants were in the role of passively observing the out-group fail in the final of a competition As the in-group had already been eliminated from the competition, there could be no material gain from the outcome of the final match

Order Manipulation

In one experimental condition, we first established the in-group’s domain inferiority and assessed its perceived illegitimacy and the pain felt about it In a second step, we established the third-party out-group as successful and assessed the perceived illegitimacy and envy of this success, anger about this success, and pain of the inferiority implied by the out-group’s success In a third step, we gave participants an opportunity for schadenfreude by having the third party out-group fail In this condition, both self-and other-focused explanations could cause schadenfreude In the second condition, however, the out-group’s success was estab-lished first and its perceived illegitimacy and envy, anger, and the pain of the inferiority implied by the out-group’s success were assessed before the schadenfreude opportunity Only after this schadenfreude opportunity was the in-group’s domain inferiority established and the pain about it assessed Thus, in this condition, the pain of domain inferiority could not cause schadenfreude How-ever, schadenfreude could cause the pain of domain inferiority

Measures

performance information, we assessed the degree to which each group was perceived as “strong,” “successful,” having “the needed capabilities,” “unlikely to be successful” (reversed), and likely to lose in the future (reversed) These items were presented after the appropriate information regarding each group and accompanied by

Likert-type scales that ranged from 1 (strongly disagree) to 7 (strongly agree) This measure of perceived performance proved

internally consistent regarding the in-group (␣ ⫽ 78) and the out-group (␣ ⫽ 82)

as-sessed the perceived illegitimacy of the in-group’s performance immediately after the manipulation check items (␣ ⫽ 56) Partic-ipants were asked whether the in-group’s performance was “fair” (reversed), “unfair,” “illegitimate,” and “acceptable” (reversed)

using a Likert-type scale that ranged from 1 (strongly disagree) to

7 (strongly agree).

establishing the in-group’s inferiority in their pool of the compe-tition, participants were asked the degree to which they felt

“in-ferior,” “frustrated,” “threatened,” and “ashamed” (␣ ⫽ 80).

These emotion labels suggest subjective pain about inferiority

(Nietzsche, 1887/1967; Scheff, 1994; R H Smith, 1991)

about the inferiority implied by the out-group’s success, we asked

how much they felt “inferior,” “threatened,” “frustrated,” and

“ashamed” immediately after the information establishing the

out-group’s success in their pool of the competition (␣ ⫽ 83)

compe-tition just after our general description thereof Thus, we presented

participants with Likert-type response scales ranging from 1 (strongly

disagree) to 7 (strongly agree) and asked whether they “might like to

participate” and whether the competition “would make a nice addition

to TV,” “seems exciting,” “seems dull” (reversed), and “would be

nice to see on TV” (␣ ⫽ 83) Leach et al (2003) suggested that

domain interest may be an alternative self-focused explanation of

schadenfreude However, we expected it to have little effect where the

pain of domain inferiority is also assessed

(Free University) and dislike of the out-group who defeated this

target (University of Groningen), as in Leach et al (2003)

as-sessed the perceived illegitimacy of the out-group’s performance

immediately after the manipulation check regarding this

perfor-mance (␣ ⫽ 77) Participants were asked whether the out-group’s

performance was “fair” (reversed), “unfair,” “illegitimate,” and

“acceptable” (reversed) using a Likert-type scale that ranged from

1 (strongly disagree) to 7 (strongly agree).

and Sherman (2002), we included two items that operationalized

envy as coveting the out-group’s success: “I would like us to be

like the Free University team” and “I want us to have what the Free

University team has” (␣ ⫽ 85) Responses were given on a scale

ranging from 1 (not at all) to 7 (very much).

Sherman (2002), and Hareli and Weiner (2002), we asked partic-ipants to what degree they felt “angry,” “irritated,” and “hostile”

(␣ ⫽ 91) Responses were given on a scale ranging from 1 (not at all) to 7 (very much).

partici-pants were told that the out-group’s success in their pool enabled them to proceed to the competition final, where they failed against the University of Groningen This provided the opportunity for schadenfreude As in Leach et al (2003, Study

2), happy, three synonyms of satisfied, and the Dutch word for

schadenfreude (␣ ⫽ 92) Responses were given on a scale

ranging from 1 (not at all) to 7 (very much).

Results

We first examined the effectiveness of our manipulation of each group’s performance in the competition Participants viewed their

in-group as performing somewhat poorly (M ⫽ 3.99, SD ⫽ 0.88), but perceived the out-group as performing well (M ⫽ 5.47, SD ⫽ 0.69), t(101) ⫽ 21.74, p ⬍ 001 Thus, we effectively established in-group

domain inferiority and out-group success Participants tended to somewhat disagree that their in-group’s inferiority was illegitimate and to moderately disagree that the out-group’s success was illegiti-mate (see Table 1) However, there was sufficient variance in these two measures of perceived illegitimacy to potentially predict schaden-freude

Gender was not examined in the analyses reported here, as in prior analyses gender never produced any main or interaction

Table 1

Means and Intercorrelations of Schadenfreude and Its Potential Explanations: Study 1

In-group inferiority first

Out-group success first

aResponse scale ranged from 1 (strongly disagree) to 7 (strongly agree). bResponse scale ranged from 1 (not at all) to 7 (very much).

ⴱp ⬍.05

effects.1As a first step, we performed an analysis of covariance to

be sure that the pain of domain inferiority’s moderate-sized effect

on schadenfreude was not moderated by the perceived illegitimacy

of in-group inferiority or the perceived illegitimacy of out-group

success This was indeed the case.2

Conceptual Model

To examine the full set of self- and other-focused explanations

identified in our conceptual model, we used EQS 6.1 to estimate

covariance structure models in each experimental condition Table 1

reports the descriptive statistics and intercorrelations of the

vari-ables in each experimental condition

Figure 2 reports the standardized solution in the condition in

which the in-group’s inferiority and the out-group’s success were

established before the schadenfreude opportunity This model

ac-counted for 62% of the variance in schadenfreude That the model

fit the data well was also shown in the nonreliable ␹2(19, N ⫽

51) ⫽ 24.83, p ⫽ 17 In addition, a variety of fit (non-normed fit

index [NNFI] ⫽ 920, comparative fit index [CFI] ⫽ 972,

incre-mental fit index [IFI] ⫽ 976) and residual (standardized

root-mean-square residual [SRMR] ⫽ 085, root-root-mean-square error of

approximation [RMSEA] ⫽ 078) indices showed good model fit

(for a review, see Hu & Bentler, 1999)

As hypothesized, the pain of domain inferiority was a highly

reliable, moderate-sized explanation of schadenfreude The

per-ceived illegitimacy of the in-group’s domain inferiority was not

associated with the pain felt about it Also as expected, the pain of

domain inferiority was strongly linked to the pain of implied

inferiority In support of the ressentiment pathway, the pain of

implied inferiority had an indirect effect on schadenfreude through

anger at the out-group’s success (i.e., the total indirect effect, the

product of the two involved paths, was 252) Consistent with full

mediation, the model fit well without estimating a direct path

between the pain of implied inferiority and schadenfreude Indeed,

the LaGrange multiplier modification index showed that adding

this path would not improve model fit ( p ⬎ 10).

None of the other-focused explanations of schadenfreude

pro-duced notable effects Thus, envy (direct effect ⫽ 09),

dislike-based anger (indirect effect ⫽ 004), and illegitimacy-dislike-based anger

(indirect effect ⫽ 049) had relatively small effects on

schaden-freude compared with that of the pain of domain (or implied)

in-group inferiority That the model fit well without specifying

direct paths to schadenfreude from dislike of the out-group or

perceived illegitimacy of their success suggests that these

other-focused explanations were unimportant Indeed, LaGrange

multi-plier modification indices showed that adding these paths would

not improve model fit (both ps ⬎ 05).

Figure 3 reports the standardized solution for the condition in

which only the out-group’s success was established before the

schadenfreude opportunity This model accounted for 57% of the

variance in schadenfreude That the model fit the data well was

also shown in the nonreliable ␹2(19, N ⫽ 51) ⫽ 21.00, p ⫽ 34 In

addition, a variety of fit (NNFI ⫽ 948, CFI ⫽ 982, IFI ⫽ 986)

and residual (SRMR ⫽ 090, RMSEA ⫽ 047) indices showed

good model fit

Although the other-focused explanations were advantaged in

this experimental condition, their effects were similar to the

base-line condition Thus, envy, dislike-based anger (.090), and

illegitimacy-based anger (.105) had small, nonreliable effects In

contrast, the ressentiment pathway was a potent explanation of

schadenfreude (indirect effect ⫽ 563) Consistent with full medi-ation, the model fit well without estimating a direct path between the pain of implied inferiority and schadenfreude Indeed, the LaGrange multiplier modification index showed that adding this

path would not improve model fit ( p ⬎ 10).

Discussion

When the in-group’s inferiority in a domain was established before an opportunity for schadenfreude, the pain of this inferiority was a potent cause of schadenfreude toward a successful third-party out-group As this out-group played no role in establishing the in-group’s inferiority, schadenfreude at their failure seems especially malicious and prejudiced As expected, the pain of domain of inferiority had a much bigger effect on schadenfreude than other self-focused explanations, like interest in the domain of competition and the perceived illegitimacy of the in-group’s do-main inferiority In addition, other-focused explanations, such as preexisting dislike of the third party and victorious out-groups had little effect on schadenfreude Thus, as Nietzsche (1887/1967) suggested, schadenfreude appeared to very much be “a vengeful-ness of the impotent.”

In the condition in which the out-group’s success was estab-lished immediately before the schadenfreude opportunity, other-focused explanations were advantaged as potential explanations of schadenfreude Yet, they played little role Thus, the perceived illegitimacy of the out-group’s success, illegitimacy-based anger, and envy had little effect on schadenfreude In this condition, it was the pain of the inferiority implied by the out-group’s success

1Although there is reason to suspect that men may be more prone to show schadenfreude in the domain of international sport competition, this is largely because men tend to be more interested in this domain As domain interest was treated as a covariate in this analysis, it is likely to have provided a more psychological account for the possible effects of gender

2Dislike of the out-group, dislike of the victor, and domain interest were treated as covariates, none of which were statistically reliable (all

ps ⬎ 22) Treating these variables as covariates was supported by the fact that other analyses showed them not to moderate the effects reported Thus, pain of inferiority (centered), perceived illegitimacy of out-group success (centered), perceived illegitimacy of in-group inferiority (centered), and order (in-group inferiority first vs out-group success first) were treated as

factors that could interact Order had no main effect on schadenfreude, F(1, 99) ⫽ 0.192, p ⫽ 66, partial ␩2⬍.002 The perceived illegitimacy of

in-group inferiority had no main effect on schadenfreude, F(1, 99) ⫽ 0.012, p ⫽ 92, partial ␩2⬍.001 However, the perceived illegitimacy of

out-group success, F(1, 99) ⫽ 3.99, p ⫽ 05, partial ␩2⫽.05, and the pain

of in-group inferiority increased schadenfreude, F(1, 99) ⫽ 43.62, p ⬍

.001, partial ␩2⫽.35 The pain of inferiority was not moderated by the

perceived illegitimacy of this inferiority, F(1, 99) ⫽ 0.083, p ⫽ 77, partial

␩2⫽.001, or by the perceived illegitimacy of out-group success, F(1, 99) ⫽ 2.16, p ⫽ 15, partial ␩2⫽.03 However, order was a marginal

moderator of the pain of inferiority, F(1, 99) ⫽ 3.35, p ⫽ 07, partial ␩2⫽ 04 The pain of inferiority was a stronger predictor of schadenfreude when

the in-group’s inferiority was established first (␤ ⫽ 0.670, p ⬍ 001, vs.

␤ ⫽0.511, p ⫽ 001) The order manipulation did not moderate the effects

of perceived illegitimacy of in-group inferiority, F(1, 99) ⫽ 1.25, p ⫽ 27,

partial ␩2⫽.02, or the perceived illegitimacy of the out-group’s success,

F (1, 99) ⫽ 2.37, p ⫽ 13, partial ␩2⬍.03

that provided the strongest basis of the anger that was a potent

explanation of schadenfreude This is consistent with the

More important, in the condition in which it was possible,

schadenfreude did not cause greater pain of domain inferiority

This suggests that the relation between pain of domain

inferi-ority and schadenfreude is nonrecursive—the pain of domain

inferiority causes schadenfreude, but not vice versa Such evi-dence is important to our causal claims given that both con-structs were measured

The absence of a schadenfreude 3 pain of domain inferiority causal link is also important evidence against the idea that schadenfreude somehow serves to reduce the pain of domain inferiority Although some might expect schadenfreude to have

Pain: Domain inferiority

Pain: Implied Inferiority

Envy: Out-group Success

.46*

.18 36*

Illegitimacy:

Inferiority

.72*

.35* .09 .70*

.14 17

−.22

Anger: Out-group Success

.14 48*

Illegitimacy:

Outgroup Success

.01

Schadenfreude

at Out-group Failure

Dislike of Out-group

R 2 = 62

Figure 2. Structural model: In-group inferiority and out-group success established before schadenfreude opportunity, Study 1

Pain: Domain inferiority

Envy: Out-group Success

.07 36*

.75*

.12

Illegitimacy:

Inferiority

.14 14

−.22

−.01

.75*

Illegitimacy:

Outgroup Success

R 2 = 57

Anger: Out-group Success

.38*

.47*

.07

Schadenfreude

at Out-group Failure

Pain: Implied Inferiority

−.08

Dislike of Out-group

Figure 3. Structural model: Out-group success established before schadenfreude opportunity, Study 1

this sort of self-enhancement function, this is not consistent with

Nietzsche’s (1887/1967) approach As he put it,

The desire is to deaden, by means of a more violent emotion of

any kind, a tormenting, secret pain that is becoming unendurable, and

to drive it out of consciousness at least for the moment [italics

added] It goes without saying that a “medication” of this kind, a

mere affect medication, cannot possibly bring about a real cure of

sickness (pp 127–128)

Thus, Nietzsche suggested that schadenfreude serves only as a

temporary anesthetic for the pain of domain inferiority As

schadenfreude toward a third party does not address the in-group’s

broader inferiority in a domain, the pain of this domain inferiority

should be unaffected by schadenfreude toward a third party This

is what we observed in the experimental condition in which

schadenfreude could cause the pain of domain inferiority

Study 2 Study 2 was designed to corroborate Study 1 and to extend it in

at least two ways First, we altered the paradigm to manipulate the

illegitimacy of both in-group inferiority and out-group success

This move from measured to manipulated illegitimacy was

espe-cially important to the examination of the illegitimacy of the

out-group’s success Although there was sufficient variance in the

perceived illegitimacy of the out-group’s success to allow this

variable to predict schadenfreude (through anger), participants

tended to disagree that the out-group’s success was illegitimate in

Study 1 Thus, we thought it important to be able to compare

out-group success perceived as legitimate to that perceived as

illegitimate

Manipulating the illegitimacy of out-group success was also

important because perceived illegitimacy may have a differential

association with anger than does objective illegitimacy (Leach,

2008; R H Smith et al., 1994) Indeed, Feather and Sherman

(2002) suggested that it is objective illegitimacy that leads to anger

and thus schadenfreude In contrast, our model proposes that

whether it is objective or subjective, the illegitimacy of (in- or

out-)group performance is only a very modest explanation of

schadenfreude

In a second major improvement over Study 1, we recruited a

large sample of participants This was done for three main reasons

First, we wanted to perform a confirmatory factor analysis on our

emotion measures Second, we wanted to have ample statistical

power to detect some of the small effects shown in Study 1 Third,

we wanted to improve our structural model by using latent, rather

than measured, variables In this way, we could account for

mea-surement error

Method Participants and Procedure

Participants were 412 psychology students at the University of

Amsterdam (291 women and 121 men), who participated in

medium-sized groups in a classroom setting

Design

In a 2 ⫻ 2 factorial design, we manipulated the illegitimacy (i.e.,

legitimate vs illegitimate) of the in-group’s inferiority and the

illegitimacy of an out-group’s independent success in the same domain This was accomplished by adding information about the procedure of the interuniversity competition used in Study 1 As in the baseline condition of Study 1, we first established the in-group’s inferiority, then established the out-group as successful, and finally presented the opportunity for, and assessed, schaden-freude Thus, emotions regarding both in-group inferiority and out-group success could cause schadenfreude It is important to note that participants had no reason to suspect that they would have an opportunity for schadenfreude before one was actually presented

As in Study 1, the in-group’s domain inferiority was established

in a table showing them to have lost five of the six matches they played However, also contained in the table was a column indi-cating which university faculty had chosen the questions used in the match For legitimate domain inferiority, four of the five times that the in-group was defeated the questions were chosen by the faculty of a university different from the winning university Thus, the procedure of the matches was fair, establishing the in-group’s domain inferiority as legitimate (see Lind & Tyler, 1988) For illegitimate domain inferiority, four of the five times that the in-group was defeated the questions were chosen by the faculty of the winning group and were thus procedurally unfair and illegiti-mate We used the procedural legitimacy of the out-group’s matches to manipulate the illegitimacy of their success in a similar way, and their success was established as in Study 1

Manipulation Checks

re-garding each group’s performance, the five items from Study 1 assessed participants’ view of each group as strong and successful

in the quiz domain This measure was internally consistent regard-ing the in-group (␣ ⫽ 82) and the out-group (␣ ⫽ 78) Confirm-ing that participants understood the performance information, they

viewed the in-group as performing somewhat poorly (M ⫽ 4.03,

SD ⫽0.83), but viewed the out-group as performing fairly well

(M ⫽ 5.14, SD ⫽ 0.77; p ⬍ 001).

we asked participants questions regarding the procedure that deter-mined their group’s domain inferiority immediately after this in-formation was presented We asked the same questions regarding the procedure that determined the out-group’s success.3All items were

presented with a Likert-type scale that ranged from 1 (strongly dis-agree) to 7 (strongly dis-agree), reversed to indicate perceived

illegiti-macy Three items assessed perception of the procedure as “legiti-mate,” “right,” and “fair.” These scales were internally consistent regarding the in-group (␣ ⫽ 56) and the out-group (␣ ⫽ 77)

3Mixed in with the manipulation check items were the four items used in Study 1 to assess the perceived illegitimacy of the out-group’s success (␣ ⫽ 71) Participants were asked whether it was “fair” (reversed), “unfair,” “ille-gitimate,” and “acceptable” (reversed), using a Likert-type scale that ranged

from 1 (strongly disagree) to 7 (strongly agree) Our manipulation of the

procedural illegitimacy of the out-group’s success operated in a way similar to our measure of perceived illegitimacy in Study 3 Thus, procedurally ilmate out-group success led to greater perceived illegitimacy than did

legiti-mate success (b ⫽ 707, SE ⫽ 098, p ⬍ 001).

Emotion Measures

We assessed pain of domain inferiority (␣ ⫽ 80), envy (␣ ⫽ 89),

anger about the out-group’s success (␣ ⫽ 90), the pain of implied

inferiority (␣ ⫽ 88), and schadenfreude (␣ ⫽ 92) as in Study 1

Given that schadenfreude, and the four emotional explanations of

it, was assessed in the same way in Studies 1 and 2, we combined

these data to perform a confirmatory factor analysis Our

hypothe-sized measurement model specified pain of domain inferiority, envy,

anger, pain of implied inferiority, and schadenfreude as latent

vari-ables Each item was allowed to indicate only its corresponding latent

variable, and no errors were allowed to correlate However, we did

allow all of the latent variables to correlate with each other

Given the large N, it is not surprising that the sample-size–

dependent chi-square was large and statistically reliable, ␹2(125,

N ⫽ 412) ⫽ 543.38, p ⬍ 001 However, in the case of

measure-ment models, other indices provide a better indication of fit

Indeed, the hypothesized model fit the data well according to a

variety of fit (NFI ⫽ 924, CFI ⫽ 940, IFI ⫽ 940) and residual

(SRMR ⫽ 043, RMSEA ⫽ 081) indices In terms of item

loadings, the pain of domain inferiority (.694 –.830), envy (.792–

1.00), anger (.842–.887), pain of implied inferiority (.741–.879),

and schadenfreude (.749 –.917) all appeared unitary, with all

indi-cators producing statistically reliable and strong loadings on their

latent factor Pain of implied inferiority and anger were most

highly intercorrelated (r ⫽ 88, p ⬍ 001) This is consistent with

our conceptualization of ressentiment The pain of domain

inferi-ority was highly correlated with the pain of implied inferiinferi-ority

(r ⫽.77, p ⬍ 001) and moderately correlated with anger (r ⫽ 66,

p ⬍.001) Envy had only low correlations with the other emotions

(rs ⫽ 14 –.29, p ⬍ 05) That the emotion measures are distinct

despite these intercorrelations was further suggested by our

com-parison of the hypothesized model to alternatives.4

Alternative Explanations

As in Study 1, we assessed dislike of the out-group, dislike of

the victor, and interest in the domain of competition (␣ ⫽ 83) as

alternative explanations of schadenfreude Separate analyses

showed these variables not to moderate our hypothesized

expla-nations Excluding them from the analyses below did not alter the

pattern of results Prior analyses showed gender to have no main or

interaction effects, thus it was excluded here

Results Perceived Procedural Illegitimacy

We first checked the effectiveness of our manipulations of the

procedural illegitimacy of the in-group’s and the out-group’s

per-formance Thus, we included both manipulations in a multivariate

analysis of variance with perceived procedural illegitimacy of the

in-group’s domain inferiority and the group’s success as

out-comes Both manipulations had statistically reliable multivariate

effects (both ps ⬍ 001, both ␩2s ⬎ 15)

At the univariate level, the procedural illegitimacy of in-group

performance affected only the perceived procedural illegitimacy of

in-group domain inferiority, F(1, 416) ⫽ 59.67, p ⬍ 001

Illegit-imate domain inferiority was perceived as more illegitIllegit-imate (M ⫽

4.30, SD ⫽ 1.40) than procedurally legitimate in-group inferiority

(M ⫽ 3.30, SD ⫽ 1.21) In addition, the procedural illegitimacy of

out-group performance affected only the perceived procedural

illegitimacy of out-group success, F(1, 416) ⫽ 72.05, p ⬍ 001.

Illegitimate out-group success was perceived as more illegitimate

(M ⫽ 4.54, SD ⫽ 1.47) than legitimate out-group success (M ⫽ 3.57, SD ⫽ 1.21) Taken together, these results suggest that we

were able to manipulate successfully the procedural illegitimacy of both groups’ performance

Schadenfreude

As in Study 1, an analysis of covariance showed the pain of domain inferiority to have a reliable effect of moderate magni-tude on schadenfreude that was not moderated by the illegiti-macy of the in-group’s domain inferiority or the out-group’s success.5 However, to examine our full conceptual model, we estimated a covariance structure model with latent variables The means and intercorrelations of these variables are reported

in Table 2

4To further confirm the adequacy of the hypothesized measurement model, we compared it with four more parsimonious alternatives First, we compared the hypothesized model with one that specified all the items as indicating a single, general-emotion latent variable This first alternative fit the data much less well than our hypothesized model, ⌬␹2(10) ⫽ 2,111.13,

p ⬍.001 Indeed, this model fit the data poorly in an absolute sense, as shown in a variety of fit (NFI ⫽ 626, CFI ⫽ 638, IFI ⫽ 639) and residual (SRMR ⫽ 111, RMSEA ⫽ 192) indices Second, we compared the hypothesized model with one that combined the items of pain of implied

inferiority and anger into one variable (i.e., ressentiment) but left the others

untouched This alternative fit less well than the hypothesized model,

⌬␹2(4) ⫽ 159.15, p ⬍ 001 It also fit the data poorly in an absolute sense,

as shown in a variety of fit (NFI ⫽ 901, CFI ⫽ 917, IFI ⫽ 918) and residual (SRMR ⫽ 052, RMSEA ⫽ 094) indices Third, we compared the hypothesized model with one that combined the items of pain of domain inferiority and pain of implied inferiority into one variable but left the others untouched This alternative fit less well than the hypothesized model, ⌬␹2(6) ⫽ 219.53, p ⬍ 001 It also fit the data poorly in an absolute

sense, as shown in a variety of fit (NFI ⫽ 893, CFI ⫽ 909, IFI ⫽ 909) and residual (SRMR ⫽ 053, RMSEA ⫽ 099) indices Fourth, we com-pared the hypothesized model with one that combined the items of pain of domain inferiority and schadenfreude into one variable but left the others untouched This alternative fit less well than our hypothesized model,

⌬␹2(6) ⫽ 598.81, p ⬍ 001 It also fit the data poorly in an absolute sense,

as shown in a variety of fit (NFI ⫽ 839, CFI ⫽ 854, IFI ⫽ 855) and residual (SRMR ⫽ 105, RMSEA ⫽ 125) indices

5The continuous measures of dislike and domain interest were treated as covariates, whereas the procedural illegitimacy of in-group inferiority, the procedural illegitimacy of out-group success, and the pain of inferiority were treated as factors that could interact None of the covariates were

statistically reliable (all ps ⬎ 10) However, the procedural illegitimacy of success produced a reliable main effect of small magnitude, F(1, 412) ⫽ 4.03, p ⫽ 01, partial ␩2⫽.01 Greater schadenfreude was shown toward

the out-group when their success was procedurally illegitimate (M ⫽ 2.73,

SD ⫽ 1.40) rather than legitimate (M ⫽ 2.45, SD ⫽ 1.42) As expected, the pain of inferiority produced a much larger effect on schadenfreude, F(1, 412) ⫽ 132.63, p ⬍ 001, partial ␩2⫽.25 The procedural illegitimacy of

in-group inferiority did not moderate the pain of inferiority, F(1, 412) ⫽ 2.94, p ⫽ 09, partial ␩2⫽.007 In addition, the procedural illegitimacy of

the out-group’s success did not moderate the pain of inferiority, F(1, 412) ⫽ 2.44, p ⫽ 12, partial ␩2⫽.006 No other interaction effects were

noteworthy (all ps ⬎ 80).

Given the large N, it is not surprising that the sample-sensitive

chi-square was large and highly reliable, ␹2(323, N ⫽ 412) ⫽

728.13, p ⬍ 001 However, a variety of fit (NNFI ⫽ 930, CFI ⫽

.941, IFI ⫽ 941) and residual (SRMR ⫽ 049, RMSEA ⫽ 056)

indices showed the model to fit well Indeed, the model accounted

for 44% of the variance in schadenfreude The model, with

stan-dardized parameter estimates, is shown in Figure 4

As hypothesized, the pain of domain inferiority was a

statisti-cally reliable explanation of schadenfreude with a medium-sized

effect The perceived illegitimacy of domain inferiority was not

associated with the pain felt about it Also as expected, the pain of

domain inferiority was strongly linked to the pain of implied

inferiority In support of the ressentiment pathway, the pain of

implied inferiority had a notable indirect effect on schadenfreude

through anger at the out-group’s success (.400) Consistent with

full mediation, the model fit well without estimating a direct path

between pain of implied inferiority and schadenfreude Indeed, the

LaGrange multiplier modification index showed that adding this

path would not improve model fit ( p ⬎ 10).

None of the other-focused explanations of schadenfreude

pro-duced notable effects Thus, envy (direct effect ⫽ 080),

dislike-based anger (indirect effect ⫽ 028), and illegitimacy-dislike-based anger

(indirect effect ⫽ 056) had very small effects on schadenfreude

However, given the high statistical power in this study, the two

paths involved in the illegitimacy-based anger pathway were

sta-tistically reliable here despite being of similar magnitude and

nonreliable in Study 1 That the model fit well without specifying

direct paths to schadenfreude from dislike of the out-group or

perceived illegitimacy of their success suggests that these

other-focused explanations were unimportant Indeed, LaGrange

multi-plier modification indices showed that adding these paths would

not improve model fit (both ps ⬎ 05).

Discussion

Where an out-group’s success was procedurally illegitimate, it

led to greater schadenfreude at their subsequent failure This is

consistent with the justice model implied in Feather and Sherman

(2002) More important, our structural model offers the first direct

support for the notion that illegitimate success leads to

schaden-freude because it promotes an illegitimacy-based anger However,

this very small effect was only detected because of the high

statistical power in this study

Envy that was narrowly defined as coveting another’s success

did not explain schadenfreude This is consistent with the findings

of Hareli and Weiner (2002) and Feather and Sherman (2002) Thus, where envy is distinguished from the pain of implied infe-riority, the perceived illegitimacy of another party’s success, and the anger that envy is often conflated with (e.g., Smith et al., 1996; van Dijk et al., 2006), envy explains little schadenfreude That envy was not linked to the other-focused explanations of schaden-freude also suggests against envy being indirectly linked to schadenfreude through feelings like anger (see Leach, 2008)

In contrast to the explanations that focus on individuals’ emo-tional responses to another party’s success, those involving the pain of inferiority were much more potent explanations of schadenfreude As in Study 1, the pain of in-group domain infe-riority was by far the strongest explanation of schadenfreude The pain of domain inferiority led to schadenfreude even though this pain was assessed before participants had any reason to expect an opportunity for schadenfreude Thus, it was clear that the pain of domain inferiority caused schadenfreude In contrast, the proce-dural illegitimacy of the in-group’s inferiority did not cause schadenfreude and did not moderate the effect of the pain of domain inferiority This corroborates and extends Study 1 by using

a manipulation of the objective illegitimacy of in-group inferiority rather than a measure of perceived illegitimacy

Study 2 also offered further support of Nietzsche’s (1887/1967)

notion of ressentiment As in Study 1, individuals’ pain about their

in-group’s independently established domain inferiority led to pain about the inferiority implied by a third-party out-group’s success This pain of implied inferiority was strongly linked to anger, a

potent predictor of schadenfreude The ressentiment pathway

ex-plained much more schadenfreude than the anger based in the procedural illegitimacy of the out-group’s success or dislike of the out-group Thus, whether as a direct explanation or as an indirect

explanation through the ressentiment pathway, the pain of in-group

inferiority determined the vast majority of the schadenfreude shown at the third-party out-group’s failure

General Discussion

A wide variety of psychological theory and research has sug-gested that inferiority is a painful emotional experience that poses

a serious threat to the self-concept Indeed, inferiority is so painful that people seem to seek emotional succor for it by taking advan-tage of others’ misfortunes to feel the pleasure of schadenfreude Discussing national and religious groups’ experience of inferiority, Nietzsche (1887/1967) argued that the pain of inferiority makes people prone to feel pleasure at the failure of a successful

out-Table 2

Means and Intercorrelations of Schadenfreude and Its Potential Explanations: Study 2

aResponse scale ranged from 1 (strongly disagree) to 7 (strongly agree). bResponse scale ranged from 1 (not at all) to 7 (very much).

ⴱp ⬍.05