Content uploaded by Johanna Ray Vollhardt
Author content
All content in this area was uploaded by Johanna Ray Vollhardt on Mar 05, 2019
Content may be subject to copyright.
Inclusive Altruism Born of Suffering: The Relationship
Between Adversity and Prosocial Attitudes and
Behavior Toward Disadvantaged Outgroups
Johanna R. Vollhardt
Clark University Ervin Staub
University of Massachusetts, Amherst
This article reports the results of 2 studies examining altruism born of suffering (E. Staub
& J. R. Vollhardt, 2008). More specifically, we examined inclusive altruism born of suffer-
ing, which is directed toward members of disadvantaged outgroups. Drawing on and inte-
grating clinical and social psychological theories, we hypothesized that individuals who
had suffered from adverse life events would be more likely to help the outgroups in need
than those who had not suffered. This was demonstrated for helpers who had experienced
various forms of suffering (interpersonal and group-based harm, natural disasters) and
for 2 distinct types of prosocial behavior and attitudes (long-term volunteering and disas-
ter aid) benefiting outgroups within society and from other countries. We also found that
prosocial attitudes toward tsunami victims were highest among those who had suffered in
a similar way (from natural disasters). Additionally, we examined the underlying social
psychological processes and found that empathy and reduced ingroup bias (but not
personal distress) mediated the effect (Study 2). Implications for social justice and an
empowering view of victims as potential helpers in society are discussed.
Interpersonal and intergroup relations are often negatively
impacted by past suffering, for example through cycles of
violence and revenge (e.g., Dodge, Bates, & Pettit, 1990;
Staub, 1998). But past victimization and suffering do not inevi-
tably have destructive effects. On the contrary, occasionally,
individuals appear to be motivated by their own adverse experi-
ences to help others and prevent further suffering. This impor-
tant, yet under researched phenomenon has been referred to
as ‘‘altruism born of suffering’’ (Staub, 2003, 2005; Staub &
Vollhardt, 2008; Vollhardt, 2009).
While there is some limited empirical evidence of altruism
born of suffering in the interpersonal realm and toward mem-
bers of the social ingroup (see reviews in Staub & Vollhardt,
2008; Vollhardt, 2009), it has not been studied explicitly whether
this phenomenon applies to support for members of disadvan-
taged outgroups as well. Some anecdotal evidence and case
studies suggest that this may be the case, such as Holocaust sur-
vivors protesting in the past against My Lai (Lifton, 2003) or in
the present against the genocide in Darfur (Messinger, 2004).
Suffering from ostracism and social exclusion or marginality has
also been reported as a possible reason for engaging in out-
groups activism (Borshuk, 2004; London, 1970). Thus, altruism
born of suffering can also be inclusive (see Staub, 2005; Voll-
hardt, 2009), that is, it may extend beyond the ingroup and to
those who suffer from experiences that differ from one’s own.
Shedding more light on this largely unexplored phenomenon is
extremely important, not only because it provides an empower-
ing view of the potential role of victims in society (Tedeschi,
1999) but also because it shows an additional pathway toward
support for disadvantaged members of society that can contrib-
ute to social justice worldwide.
Theory and research on inclusive forms of altruism born of
suffering can benefit from a consideration of the complementary
perspectives provided by clinical and social psychology. Scholars
of posttraumatic growth have proposed that it may include
increased compassion, empathy, and altruism (Janoff-Bulman,
1992; Tedeschi, Park, & Calhoun, 1998; see also review in Staub
& Vollhardt, 2008). However, this literature has focused largely
on interpersonal helping and helping members of the social
ingroup. Additionally, research on posttraumatic growth has
focused mostly on positive consequences for the self, for exam-
ple, on how helping others can increase coping (Midlarsky,
1991), provide meaning (Janoff-Bulman, 1992), and thereby fos-
ter healing (e.g., Kishon-Barash, Midlarsky, & Johnson, 1999).
Yet, obviously, prosocial outcomes of posttraumatic growth
may also benefit others, and a crucial, unresolved question is
how these beneficial effects may extend to outgroups in need.
These questions can be addressed by drawing on social
psychological theories that examine general processes leading to
helping behavior and that consider the recipient of help, such
as perspective-taking and empathy (Batson & Oleson, 1991),
Correspondence concerning this article should be addressed to
Johanna Ray Vollhardt, Department of Psychology, Jonas Clark Hall,
Clark University, 950 Main Street, Worcester, MA 01610. Electronic
mail may be sent to jvollhardt@clarku.edu.
American Journal of Orthopsychiatry 2011 American Orthopsychiatric Association
2011, Vol. 81, No. 3, 307–315 DOI: 10.1111/j.1939-0025.2011.01099.x
307
similarity with the victim (Dovidio, 1984; Krebs, 1975), and a
perceived common ingroup (Gaertner & Dovidio, 2000).
The present article reports the results of two studies that
tested whether suffering would be associated with increased will-
ingness to help outgroup victims. We studied this question for
different outgroups (disadvantaged groups within society and
from other parts of the world) and for different kinds of proso-
cial attitudes and behavior (volunteering in everyday life, donat-
ing, and volunteering for disaster aid). In addition, we tested
social psychological processes that may mediate the effect
(Study 2).
Altruism Born of Suffering: A Brief Review of
Empirical Evidence
Several studies document increased prosocial attitudes and
behavior among individuals who have suffered, both from natu-
ral causes and harm inflicted by other human beings (see
reviews in Staub & Vollhardt, 2008; Vollhardt, 2009). For exam-
ple, researchers have observed enhanced prosocial behavior
among bereaved spouses (Brown, Brown, House, & Smith,
2008), survivors of strokes (Gillen, 2005), and HIV-positive
individuals (Reeves, Merriam, & Courtnenay, 1999). These stud-
ies primarily report prosocial behavior that benefits family
members, friends, and those suffering from the same illness—in
other words, ingroup members. Similar findings have also been
documented among victims of interpersonal violence, such as
sexual abuse (e.g., Grossman, Sorsoli, & Kia-Keating, 2006)
and among those who survived mass violence such as war (Rab-
oteg-S
ˇaric, Z
ˇuzˇ ul, & Kerestes
ˇ, 1994), terrorism (Kleinman,
1989), and genocide (Kahana, Harel, & Kahana, 1988). Like-
wise, the ‘‘altruistic community’’ after disasters is often
restricted to ingroup members, whereas structurally disadvan-
taged outgroups such as ethnic minorities, the less educated, or
elderly, are less likely to receive help (Kaniasty & Norris, 1995;
Norris, Baker, Murphy, & Kaniasty, 2005).
However, most of the existing studies are not concerned with
relations between group members, and therefore do not explic-
itly distinguish between ingroup and outgroup helping. In addi-
tion, many of these studies were not explicitly designed to test
the altruism born of suffering hypothesis and therefore have
methodological limitations such as small sample sizes and lack
of a control group. Thus, the current research aimed to provide
a first explicit test of altruism born of suffering directed toward
disadvantaged outgroups and to examine underlying social psy-
chological processes that can join the existing clinical theories
and explanations.
Underlying Processes of Altruism Born of
Suffering
The clinical perspective: Helping as coping and as
posttraumatic growth. Research documenting phenomena
relevant to altruism born of suffering has focused primarily on
prosocial behavior as a coping mechanism (e.g., Brown et al.,
2008; Kahana et al., 1988; Wayment, 2004) or as an outcome of
posttraumatic growth. Midlarsky (1991) has proposed several
mechanisms whereby helping others may serve to cope with
stress, for example by distracting from one’s own troubles in a
downward social comparison, increasing self-efficacy and per-
ceived competence, and providing meaning. Most of the litera-
ture on altruism as a manifestation of posttraumatic growth has
focused on the latter, providing meaning after suffering (e.g.,
Gillen, 2005; Tedeschi et al., 1998).
While this perspective provides important insights into gen-
eral underlying processes of altruism born of suffering, it cannot
explain why (and when) altruism born of suffering may be direc-
ted toward disadvantaged outgroup members. This question can
be addressed by drawing on several social psychological theories
that may be applied to altruism born of suffering.
Social Psychological Explanations for Altruism
Born of Suffering
Empathy and perspective-taking. Empathy and per-
spective-taking are important precursors of altruism and proso-
cial behavior (Batson & Oleson, 1991; Krebs, 1975) and
similarity with a victim (in terms of personality traits, attitudes,
etc.) has been shown to increase empathy and, in turn, prosocial
behavior (e.g., Dovidio, 1984; Westmaas & Silver, 2006).
Although only very little research to date has examined similar-
ity in regard to the experienced suffering (see Westmaas &
Silver, 2006), it seems possible that these general processes may
also apply in this case, such that empathy and perspective-tak-
ing with victims may be higher among those who have suffered
in similar ways. Specifically, an individual’s own adverse experi-
ences may make others’ suffering more comprehensible by
increasing the ability to ‘‘put oneself in the other’s shoes.’’
Accordingly, women who reported that they had experienced
rape responded with higher levels of empathy to videos about
rape victims than women who had not suffered from this experi-
ence (Barnett, Tetreault, & Masbad, 1987). However, the limited
empirical evidence is mixed; for example, Mason, Riger, and
Foley (2004) did not find that self-identified rape survivors
blamed rape victims less than other participants. Alternatively,
similarity with a victim could also increase personal distress,
which would only predict nonaltruistic helping (Wayment, 2004)
or even decrease helping when the helper can escape the situa-
tion (Carlson & Miller, 1987). Thus, we believed this question
needs further examination.
Common ingroup identity. Another way in which an
individual’s experience of suffering could increase the willing-
ness to help other victims is through social categorization pro-
cesses. When people are aware of attributes or superordinate
group memberships that are shared with outgroup members,
they may recategorize these outgroup members into a common
ingroup (Gaertner & Dovidio, 2000). Superordinate categoriza-
tion, in turn, has been shown to increase prosocial behavior
toward former outgroup members (e.g., Dovidio, Gaertner,
Validzic, & Matoka, 1997; Levine, Prosser, Evans, & Reicher,
2005). Thus, by extension, when suffering is salient, individuals
who themselves have undergone significant adversity may
include other victims in a common ingroup. This process could
explain reduced bias and increased helping behavior toward out-
group members in need.
308 VOLLHARDT AND STAUB
Potentially inhibiting factors of altruism born of
suffering. Obviously, increased prosocial behavior toward
other victims does not always occur, and destructive, antisocial
responses to victimization are well documented (e.g., Dodge
et al., 1990). Under conditions of ongoing threat, victims may
be so absorbed by their own suffering that they are unable to
perceive or empathize with the suffering of others (Chaitin &
Steinberg, 2008; Lifton, 2003). In addition, many individuals
who have suffered from traumatic life experiences and adversity
may not have the necessary resources to help. Trauma is charac-
terized by a loss of material and psychological resources
(Hobfoll, 1991), and several types of adversity, such as poverty
and structural violence, entail by definition a lack of resources.
In this situation, individuals may not only be unable to assist
others but also perceive appeals to help others as unreasonable,
given their own disadvantage and need. For this reason, helping
behavior was chosen where the potential helpers were not in
conflict over resources and had the opportunity and ability to
help.
In sum, the goals of this research were: (a) to provide a first
direct test of the altruism born of suffering hypothesis with a lar-
ger sample and a control group and to examine whether it
extends to disadvantaged outgroup members, and (b) to compare
the role of empathy and perspective-taking, personal distress,
and reduced ingroup bias as potentially underlying processes.
Study 1
In Study 1, we examined volunteering, a planned helping
behavior in everyday life with a relatively high level of personal
costs and commitment. We hypothesized that individuals who
had suffered would be more likely to volunteer than those who
had not. In addition, based on limited preliminary research, we
expected not only quantitative but also qualitative differences.
For example, in an exploratory survey study among future help-
ing professionals, Krous and Nauta (2005) found that those with
a troubled past were more likely to express the desire to specifi-
cally serve underserved populations. Similarly, we expected that
participants who had suffered would be more likely to report
volunteering activities that benefit and involve personal contact
with those who are disadvantaged—such as the ill, homeless
people, and the elderly and disabled. Past research has shown
that these stigmatized groups are often perceived as social out-
groups (e.g., elderly: Turner, Crisp, & Lambert, 2007; homeless
people: Harris & Fiske, 2006; disabled: Guimond, Dif, & Aupy,
2002).
Method
Participants. One-hundred sixty-three undergraduate stu-
dents (mean age = 19.27, SD = 1.69; 67% female) participated
in this study for research credit. In regard to ethnicity, 79.1% of
the sample reported being European American, 11.7% Asian
American, 6.1% African American, and 3% Latino or Latina.
A modified version of the Traumatic Stress Schedule (Norris,
1990) with 13 items assessed whether participants had experi-
enced any of these traumatic life events (e.g., ‘‘Have you ever
personally experienced severe or repeated physical violence?’’;
‘‘Have you ever suffered injury or property damage because of
fire, flood, earthquake, or hurricane?’’; ‘‘Have you ever person-
ally experienced extreme forms of political violence?’’). Because
of the small number of participants in each category, we only
compared those who had experienced at least one of these trau-
matic events (‘‘suffering,’’ n= 97) with those who reported that
they had not (‘‘no suffering,’’ n= 66).
Measures and procedure. Participants completed a
brief demographic questionnaire and were then asked to list the
organizations they volunteer for on a regular basis, if any. Their
open-ended responses were coded into three broad categories.
The first was volunteer work in organizations, without direct
contact with those who suffer. This included (a) environmental
or animal rights organizations, (b) health organizations, and (c)
charity and fundraising activities. The second category was vol-
unteering that involves contact with ingroup members who are
not (visibly) suffering or disadvantaged. This included (a)
community work (e.g., in the local library), (b) tutoring and
teaching, and (c) other work with youth (e.g., in summer
camps). The third category was volunteer activities that involve
direct contact with those who suffer, namely (a) the ill, (b)
homeless people, and (c) the elderly or disabled.
Volunteering activities that benefit ethnic or racial outgroups
(e.g., tutoring immigrants) were only mentioned by two partici-
pants and therefore were dropped from the analysis.
Results
As hypothesized, the overall percentage of volunteers was
significantly higher, v
2
(1, 162) = 8.07, p< .01, among partici-
pants who had suffered from traumatic life events (63.5%) than
among participants who had not suffered (41.9%). Participants
who had experienced traumatic life events also volunteered for a
greater number of organizations, M= 1.19, SD = 1.19 vs.
M= 0.73, SD = 1.1, t(1, 160) = )2.49, p< .05.
Figure 1 further breaks down this effect, showing the propor-
tion of participants from each group who were involved in spe-
cific types of volunteer work. In support of our second
hypothesis, significant differences were revealed for the compos-
ite measure of volunteer work that involved contact with disad-
vantaged outgroups, such that participants who had suffered
reported, on average, higher involvement with this kind of vol-
unteering, t(1,160) = )2.01, p< .05. In contrast, there were no
significant differences for volunteer work involving contact with
ingroup members who are not disadvantaged, t(1, 160) = )1.38,
p= .17, nor for volunteer work in organizations, t(1,
160) = )0.64, p= 0.52. However, it should be noted that the
descriptive trends were in the same direction, such that those
who had suffered had higher average rates of volunteering
across volunteer categories (nondisadvantaged ingroup mem-
bers: M
suff
=0.33vs.M
nonsuff
= 0.21; organizations: M
suff
=
0.33 vs. M
nonsuff
= 0.27).
Discussion
Study 1 provided initial support for the altruism born of suf-
fering hypothesis. Individuals who had suffered from traumatic
INCLUSIVE ALTRUISM BORN OF SUFFERING 309
life events were overall more likely to volunteer; they also volun-
teered for a greater number of organizations. Moreover, further
breaking down the various forms of volunteering revealed that
those who had suffered were particularly engaged in volunteer-
ing that benefits disadvantaged groups in society, and involves
direct contact with stigmatized outgroups. In other words,
although the general effect of altruism born of suffering was
demonstrated in the overall analysis across different types of
helping, the exploratory analysis of volunteering types suggests
that, additionally, this phenomenon might be even more
nuanced: Those who suffered in the past are not only generally
more likely to help, but appear particularly drawn to helping
alleviate and prevent suffering of those who are currently in
need and disadvantaged (Staub & Vollhardt, 2008).
Building on this exploratory finding, studies should be
designed to specifically test the interaction with different volun-
teering tasks and opportunities. Although the lack of statistical
power in this initial study did not allow for testing the interac-
tion between suffering and volunteer task, the findings suggest
that testing the type of helping as a moderator might be an
important direction for future research.
Future research will also need to systematically test potential
explanations of this phenomenon. For example, those who have
suffered may be more familiar with the situation and less inhib-
ited to be in direct contact with those in need because of their
own experience of suffering; they may also perceive general sim-
ilarities with their own experience and identify with those in
need (see Westmaas & Silver, 2006). It has also been suggested
that those who have suffered may have been recipients of help-
ing themselves and therefore more exposed to role models of
helping those in need (Krous & Nauta, 2005; see also review in
Vollhardt, 2009). Additionally, a downward social comparison
with others who suffer may occur (Midlarsky, 1991) as well as a
need to help those who are in even greater need than oneself
(Mattis et al., 1999).
These qualitative differences in the preference for certain vol-
unteer tasks also suggest other-oriented motives to improve the
condition of those in need, rather than self-enhancement
motives, which could also be satisfied by other forms of volun-
teering (Clary & Snyder, 1999). However, without testing the
underlying, mediating processes directly it cannot be ruled out
that some may have volunteered for self-focused reasons. A fur-
ther limitation of this study was that it relied on self-reports of
volunteering rather than direct observations, and that we did
not assess the extent and duration of volunteering. Importantly,
it was also not clear whether participants actually perceived the
group involved as clear outgroups. These limitations were
addressed in Study 2.
Study 2
Study 2 examined reactions to victims of the tsunami that
struck Southeast Asia in December 2004. We hypothesized that
those who had suffered would show higher levels of prosocial
attitudes and behavior toward tsunami victims—that is, out-
group members for non-Asian students in the United States. An
exploratory question was whether this effect would be mediated
by increased empathy, personal distress, or reduced ingroup
bias. We also explored whether individuals who had experienced
a similar event (natural disasters) would be more likely to help
than those who had suffered in other ways.
Method
Participants. The sample consisted of 146 undergraduate
students (mean age = 19.32, SD = 1.76, 64% female) in the
Northeast United States Participants from the region affected
by the tsunami were excluded from the analysis to ensure that
the recipients of help were clear outgroup members. The
remaining sample consisted of 82.2% European Americans,
8.2% Asian Americans, 6.2% African Americans, 2.1% Lati-
no ⁄as, and 1.4% Native Americans. Eleven percent of the sam-
ple was born outside the United States.
Measures and materials
Assessment of suffering.We used the same modified
version of Norris’s Traumatic Stress Schedule (1990) as in Study
1 to assess past suffering. This time however, we classified par-
ticipants into four categories: those who had experienced (a)
natural disasters (n= 17), (b) interpersonal violence or unspeci-
fied trauma (severe or repeated physical or psychological vio-
lence and ‘‘other’’ traumatic life events; n= 28), or (c) group-
based violence (war, political violence, severe verbal or physical
abuse due to one’s group membership; n= 37), and (d) no suf-
fering when participants reported none of the traumatic life
experiences (n= 64). Participants who indicated several types
of traumatic life events were classified into only one category,
namely the less common event (a or c). As the number of those
reporting several traumatic events was substantial (68% of those
who had suffered), this variable was included as a control vari-
able in the regression analyzes.
Prosocial outcome measures.Participants’ prosocial
attitudes and behavior toward tsunami victims were elicited
after reading an illustrated newspaper article about the tsunami.
Six items assessed participants’ attitudes toward helping tsunami
victims on 6-point rating scales from 1 (strongly disagree)to6
(strongly agree; Cronbach’s a= 0.75; see Table 1). Prosocial
behavioral intentions were assessed by asking participants
0
2
4
6
8
10
12
14
16
18
%
No Suffering
Suffering
Figure 1. Types of volunteer activities among those who have suffered
and those who have not.
310 VOLLHARDT AND STAUB
whether they planned to donate to organizations providing aid
to tsunami victims on a 4-point scale from 1 (definitely not)to4
(definitely yes). A behavioral measure was obtained after the
ostensible end of the study. Participants were given the option
of signing up to receive e-mails with information about local
fundraising events for tsunami victims. They could also sign up
for the first meeting of a group on campus that would organize
such events.
Mediators and control variables.After reading the
news story about the tsunami, participants were asked to write
down their immediate thoughts. Following a classification pro-
vided by Fultz, Schaller, and Cialdini (1988), answers were coded
(by two independent raters who were blind to the participants’
category of suffering) for expressions of empathy and perspective-
taking (e.g., pity and compassion; imagining details of the
victims’ situations). In addition, answers were coded for lack of
empathy (e.g., self-centered thoughts such as ‘‘I am glad this
didn’t happen to me,’’ or focusing on nonhuman aspects of the
disaster, e.g., ‘‘nature is so powerful’’). Personal distress was
coded when participants reported negative affect in response to
the events, such as ‘‘A state of sadness overwhelmed me.’’ The in-
terrater reliability was satisfactory (Cohen’s j= 0.90 for empa-
thy, 0.86 for personal distress, and 0.83 for lack of empathy), and
disagreements between ratings were resolved through discussion.
A measure of ingroup bias in helping tsunami victims was
obtained by computing difference scores between participants’
response to the item ‘‘The U.S. government has the responsibil-
ity to help American tourists who were victims of the tsunami’’
and a parallel item stating that the U.S. government should help
victims from the region (see Table 1). Higher positive difference
scores indicated greater ingroup bias, that is, a greater inclina-
tion to help American victims than outgroup victims.
Finally, demographic information (age, gender, ethnicity,
nationality) was collected as potential control variables.
Procedure. Participants completed the questionnaire in a
lab setting and received research credit for their participation.
After providing demographic information, they read the news
article about the tsunami and were asked to write down their
spontaneous reaction and complete the measures of prosocial
attitudes and behavioral intentions. This was followed by the
items from the Traumatic Stress Schedule. After the ostensible
end of the study, the behavioral measure was obtained by giving
participants a sign-up sheet for a tsunami fundraising event. To
reduce social desirability, the sign-up sheets were placed in enve-
lopes and collected in a box in a corner of the room where the
research assistant could not see it. Before leaving, the partici-
pants were debriefed and provided with real information about
tsunami fundraising events.
Results
Prosocial attitudes. The group means and standard devi-
ations for each scale item measuring prosocial attitudes toward
tsunami victims are shown in Table 1. Planned contrasts were
used to test all three categories of suffering against the group
‘‘no suffering.’’ This analysis revealed that those who had suf-
fered (from either of the assessed traumatic events) scored sig-
nificantly higher on the scale assessing prosocial attitudes
toward tsunami victims, t(3, 142) = 2.01, p< .05. This effect
was driven primarily by the item ‘‘perceived personal responsi-
bility to help,’’ for which the greatest mean differences were
observed, t(3, 141) = 2.73, p< .01.
Additional multiple regression analyses were conducted to
control for potentially confounding variables. Specifically, apart
from demographic variables such as gender and age, we also
controlled for ethnic minority status and place of birth (born
abroad). The rationale for this was twofold: First, norms of
helping strangers vary considerably across cultures (e.g., Knafo,
Schwartz, & Levine, 2009); in addition, it seemed possible that
international students and first generation immigrants might
have identified more with issues related to the world news and
perceived more communalities with people from the regions
Table 1. Means and Standard Deviations of Prosocial Attitudes Toward Tsunami Victims
Item
Interpersonal
(n= 27)
Group-based
(n= 37)
Natural disaster
(n= 17)
Suffering (all)
(n= 81)
No suffering
(n= 65)
1. The U.S. government has pledged $350 million to tsunami
relief funds. Some believe that this amount is too high (R).
5.07 (1.07) 5.08 (0.95) 5.12 (1.05) 5.09 (1.00) 4.66 (1.14)
2. The U.S. government has the responsibility to help the
inhabitants of the region affected by the tsunami.
4.63 (1.33) 4.51 (1.35) 4.88 (1.11) 4.63 (1.23) 4.18 (1.2)
3. The U.S. government has the responsibility to help
American tourists who were victims of the tsunami.
4.89 (0.97) 4.95 (1.27) 5.29 (0.77) 5.00 (1.08) 4.95 (0.87)
4. If a regional warning system had been in place in the Indian
Ocean on the day of the tsunami, many thousands of
people could have been saved. Such an early warning
system for the Indian Ocean would cost $30 million and
could go into operation by mid-2006. Some believe the U.S.
should cofinance this project.
4.04 (1.22) 4.32 (1.11) 4.47 (0.87) 4.26 (1.10) 4.26 (1.11)
5. Americans (private households) should donate money to
help the victims of the tsunami.
4.59 (0.80) 4.43 (1.14) 4.76 (0.66) 4.56 (0.95) 4.28 (1.01)
6. I feel a personal obligation to donate money to help victims
of the tsunami.
4.19 (1.24) 4.08 (1.4) 4.53 (1.07) 4.21 (1.28) 3.61 (1.22)
Scale mean and standard deviation (Cronbach’s a= 0.75) 4.57 (0.73) 4.56 (0.83) 4.84 (0.55) 4.62 (0.75) 4.32 (0.72)
INCLUSIVE ALTRUISM BORN OF SUFFERING 311
affected by the tsunami than those born and raised in the
United States would.
Table 2 reports the final models (including only significant
control variables). Suffering remained a significant predictor of
prosocial attitudes when controlling for gender and birth place
(born abroad) and of perceived personal responsibility when
controlling for gender and age. Moreover, suffering was the
strongest predictor of prosocial attitudes toward outgroup
victims.
Measures of the number of experienced traumatic events and
of the self-reported severity of participants’ suffering were also
entered in the regression models as potential control variables.
However, they were neither significant predictors of prosocial
attitudes (number of events: b=)0.06, t=)0.47, p= 0.64;
severity of suffering: b=)0.13, t=)1.41, p= .16) nor of
perceived personal responsibility to help (number of events:
b= 0.00, t=)0.002, p= 0.998; severity of suffering: b=
)0.07, t=)0.77, p= 0.44). As adding these coefficients did
not change the results, they were dropped from the final models
reported in Table 2.
As the degree of prosocial attitudes differed for the three types
of suffering, and the effects appeared to be driven by those who
experienced similar suffering, a natural disaster (see Table 1), the
regression analyses were repeated with each type of suffering as an
independent predictor. In this analysis, when controlling for place
of birth and gender, indeed only similar fate (natural disaster)
remained a significant predictor of overall prosocial attitudes
(b= 0.23, t= 2.64, p< .001). The perceived personal responsi-
bility to help was predicted by all three categories of suffering,
controlling for age and gender, although similar fate was a stron-
ger predictor (b= 0.25, t= 2.92, p< .01) than the experience
of interpersonal violence (b= 0.18, t= 2.03, p< .05) or group-
based violence (b= 0.18, t= 2.06, p< .05).
Prosocial behavioral intentions and behavior. In
support of the basic hypothesis, a planned contrast revealed that
those who had suffered indicated, on average, a higher intention
of donating to tsunami relief, M= 2.14, SD = 0.57 vs.
M= 2.45, SD = 0.64; t(3, 93) = )2.40, p< .05. In multiple
regression analysis, this effect remained significant (b= 0.21,
t= 2.01, p< .05) after controlling for demographic predictors
that were significant in the first step, gender (Step 2: b= 0.31,
t= 3.12, p< .01) and minority group status (Step 2: b= 0.15,
t= 1.41, p= .16).
A composite measure was created for prosocial behavior
by assigning a score of 0 to those who did not sign up for any
helping opportunity, 1 for requesting fundraising information
only, 2 for signing up to participate in the fundraising group,
and 3 to those who signed up for both. Participants who had
suffered scored significantly higher on this measure (M= 0.59,
SD = 0.95; t=)2.22, p< .05) than those who had not expe-
rienced any traumatic life event (M= 0.28, SD = 0.72). How-
ever, when controlling for gender (b= 0.15, t= 1.9, p= .059)
and minority group status (b= 0.22, t= 2.67, p< .01), suffer-
ing was only a marginally significant predictor (b= 0.15,
t= 1.8, p= .07), presumably because of the correlation
between minority group status and suffering, r= 0.30, p<
.001. One reason for this correlation might be that one of the
assessed forms of suffering, group-based violence, was reported
more frequently among ethnic minority group members (55.6%)
than among European Americans (15.5%).
Mediational analysis
Empathy and perspective-taking. A content analysis
of the open-ended responses to the newspaper article revealed
that participants who had suffered were significantly more likely
to express empathy and imagine the perspective of the victims
than those who had not suffered, 37% vs. 20%; v
2
(1,
146) = 5.04, p< .05. In addition, those who had not suffered
were more likely to report nonempathic thoughts, 40% vs.
21%; v
2
(1, 146) = 6.27, p= .01. Using Baron and Kenny’s
(1986) steps, we found that empathy and perspective-taking
mediated the effects of suffering on prosocial attitudes (see Fig-
ure 2a) as well as on prosocial behavior (see Figure 2b).
Personal distress.Participants who had suffered were
also more likely to report personal distress after reading the
newspaper article than those who had not suffered (60.5% vs.
46.2%). However, this difference was only marginally signifi-
cant, v
2
(1, 146) = 2.99, p= .059, and personal distress was
not a mediator, as it predicted neither prosocial attitudes
(b= 0.07, t= 0.88, p= 0.38) nor behavior (b= 0.01,
t= 0.06, p= 0.95).
Reduced ingroup bias.The difference scores between
the perceived responsibility of the U.S. government to help
American victims of the tsunami versus victims of the region
were significantly lower among those who had suffered than
those who had not, )0.37 (SD = 1.10) vs. )0.77 (SD = 1.13),
t(1, 144) = )2.15, p< .05. In other words, ingroup bias in
helping was reduced among those who had suffered; and it med-
iated the effect of suffering on prosocial attitudes toward tsu-
nami victims (see Figure 2c), explaining 19% of the variance.
However, reduced ingroup bias did not mediate the effect of
suffering on prosocial behavior (bias was not a significant pre-
dictor of the behavioral measure: b=)0.10, t=)1.21,
p= 0.23).
Discussion
The results of this study support the hypothesis that indi-
viduals who have suffered are more likely to exhibit prosocial
attitudes and behaviors toward outgroup members in need—
specifically, victims of a natural disaster in a different part of
the world—than those who have not experienced significant
Table 2. Predicting Prosocial Attitudes Toward Tsunami Victims
Prosocial attitudes
Perceived personal
responsibility
B(SE)BB(SE)b
Age ns.ns.)0.11 (0.06) )0.16
†
Female 0.14 (0.13) 0.09 0.35 (0.22) 0.13
Born abroad 0.35 (0.20) 0.15
†
ns.ns.
Suffering 0.25 (0.13) 0.17* 0.64 (0.20) 0.25**
Adjusted R
2
5% 8.4%
*p< .05. **p< .01.
†
p< .10
312 VOLLHARDT AND STAUB
adversity. Exploratory analyses also revealed that prosocial
attitudes were higher among those who had experienced a simi-
lar event type (natural disaster), compared with those who had
suffered for other reasons.
Further analysis demonstrated that empathy mediated the
effects of suffering on increased prosocial attitudes and behav-
ior, whereas personal distress in response to the events did not.
This finding strengthens the case for altruistic motivations that
may emerge from the experience of suffering when confronted
with others in need.
In addition, reduced ingroup bias among those who had suf-
fered mediated the effects of suffering on prosocial attitudes.
This suggests that when confronted with suffering, participants
with adverse life experiences may have included other (out-
group) victims in a common ingroup, based on similarity of
experiences. However, as perceived similarity was not assessed
explicitly, future research is needed to test this process more
directly by including measures of perceived similarity and com-
monality with other victims and victim groups (Vollhardt,
2009).
Despite overall supporting evidence, several limitations of this
study must be noted. These include rather small effect sizes,
unequal numbers of participants in the different categories of
adverse life events, and the reliance on self-reports. In addition,
the assessment of suffering was very broad and future research
will have to use more detailed measures of the time point and
severity of suffering as well as measures of posttraumatic stress
disorder (PTSD), resilience, and psychological well-being as
potential moderators of these effects.
General Discussion and Conclusion
Overall, the two studies presented in this paper provide sup-
port for the altruism born of suffering hypothesis across differ-
ent contexts and benefiting various outgroups, both within
society and in other countries. We found consistently for differ-
ent measures that prosocial responses toward outgroups in need
were higher among those who had suffered (from various types
of adverse life events) than among those who had not. This
effect occurred for different types of prosocial attitudes and
behaviors, including volunteering in everyday life and short-
term emergency helping. In addition, we found that increased
empathy and perspective-taking with other victims as well as
decreased ingroup bias toward victims in a different part of the
world, are underlying processes that can explain this effect at
least in part. Overall, the findings demonstrate the fruitfulness
of integrating theories from clinical and social psychology.
Implications for Future Research
A central question for future research on this topic is whether
and when prosocial behavior among those who have suffered is
actually motivated by altruism. Although the term we use
implies an altruistic motivation, the present studies do not
explicitly assess individuals’ motivation to help. As prosocial
behavior can also serve as a coping mechanism (Midlarsky,
1991), the motivations among participants to help or express
prosocial opinions may have been self-focused rather than
other-focused. However, we found that spontaneous expressions
of empathy were higher among those who had suffered and,
moreover, that empathy mediated the effect of suffering on pro-
social attitudes and behavior, whereas personal distress did not.
This suggests the presence of altruistic motives (Batson &
Oleson, 1991). Future research should directly compare the
prevalence of nonaltruistic and altruistic motivations to help
among those who have suffered and those who have not.
As a result of the correlational nature of this research, it is
not possible to make claims of causality. Although suffering
obviously cannot be manipulated experimentally, future
research could, however, manipulate different ways of framing
the groups’ experience, for example in regard to perceived simi-
larity or a common ingroup identity. In this context, it will also
be important to examine when altruism born of suffering is
focused primarily on those experiencing the same kind of event,
and when it extends to other types of suffering as well. It will be
important to study how perceptions of similarity and shared
suffering that may lead to increased helping behavior can be
promoted between individuals who suffered in distinct ways and
not from similar events.
To further extend the generalizability of the altruism born
of suffering hypothesis, it will be important to replicate the
findings in noncollege populations with lower levels of educa-
tion and lower socio-economic status and among less resilient
individuals. It also is possible that individuals with more
severe experiences of trauma and adversity would have shown
different responses. Thus, it is important to assess the bound-
ary conditions of altruism born of suffering as well as how it
may be fostered among those who are more affected by
adversity.
Experience
of suffering
Empathy
.19* .35***
.20*
(.13)
Prosocial
attitudes
R2= .16
z= 2.02, p< .05
Experience
of suffering
Empathy
Prosocial
attitudes
Experience
of suffering
Empathy
.19* .27**
.20*
(.15)
Signing up to
help
R2= .11
z= 1.96, p< .05
Experience
of suffering
Empathy
Signing up to
help
Experience of
suffering
Ingroup bias in
helping victims
-.18* -.39***
.20*
(.13)
Prosocial
attitudes
R
2
= .19
z= 2.15, p< .05
Experience of
suffering
Ingroup bias in
helping victims
Prosocial
attitudes
(a)
(b)
(c)
Figure 2. (a) Mediation analysis of suffering, empathy ⁄perspective-
taking, and prosocial attitudes. (b) Mediation analysis for suffering,
empathy ⁄perspective-taking, and prosocial behavior. (c) Mediation
analysis for suffering, reduced ingroup bias, and prosocial attitudes.
Note. Standardized regression coefficients are reported. The zscores
indicate whether the mediation effect is significant (Sobel test).
*p< .05. **p< .01. ***p< .001.
INCLUSIVE ALTRUISM BORN OF SUFFERING 313
Implications for Intervention and Prevention
Given the findings on the positive effects of helping behavior
on coping and perceived self-efficacy (e.g., Midlarsky, 1991),
facilitating altruism born of suffering could also be an impor-
tant element in interventions for individuals who have suffered
(see Canale & Beckley, 1999). Likewise, given the tendency to
favor the ingroup in the distribution of resources and helping
behavior, the possibility that one’s own suffering can increase
the ability to understand others’ adversity, and possibly even
lead to the inclusion of outgroup victims in a common ingroup,
could have important implications for social justice. In other
words, inclusive altruism born of suffering can replace cycles of
violence with solidarity and mutual care.
In sum, the present research demonstrates that altruism
born of suffering can be a pathway to inclusive caring (Staub,
2005), that is to prosocial attitudes and behavior that benefit
various outgroup members and disadvantaged groups in soci-
ety. This important phenomenon should be studied more, and
interventions should be designed that foster altruism born of
suffering that is directed toward outgroups in need. For exam-
ple, interventions could involve direct contact with other
victimized groups and the exposure to their narratives, under-
standing the systemic nature of oppression that may create
bonds with members of other victimized groups (Duncan,
1999), or education and learning about conflicts in other parts
of the world (Staub, Pearlman, Gubin, & Hagengimana,
2005). Providing structured opportunities to help other victims
is also crucial.
To conclude, increasing our knowledge about altruism born
of suffering and its inclusive nature not only has empowering
implications for victims’ self-concept, but may also foster har-
monious intergroup relations and contribute to social justice.
Keywords: victims of trauma; suffering; altruism; prosocial behav-
ior; volunteering; ingroup bias; outgroup helping; tsunami
References
Barnett, M., Tetreault, P., & Masbad, I. (1987). Empathy with a rape
victim: The role of similarity of experience. Violence and Victims,2,
255–262.
Baron, R., & Kenny, D. (1986). The moderator-mediator variable dis-
tinction in social psychological research: Conceptual, strategic, and
statistical considerations. Journal of Personality and Social Psychol-
ogy,51, 1173–1182.
Batson, C., & Oleson, K. (1991). Current status of the empathy-altruism
hypothesis. In M. S. Clark (Ed.), Prosocial behavior (pp. 62–85).
Newbury Park, CA: Sage.
Borshuk, C. (2004). An interpretive investigation into motivations for
outgroup activism. The Qualitative Report,9, 300–319. Retrieved
from http://www.nova.edu/ssss/QR/QR9-2/borshuk.pdf.
Brown, S., Brown, M., House, J., & Smith, D. (2008). Coping with
spousal loss: potential buffering effects of self-reported helping behav-
ior. Personality and Social Psychology Bulletin,34, 849–861.
Canale, J., & Beckley, S. (1999). Promoting altruism in troubled youth:
Considerations and suggestions. North American Journal of Psychol-
ogy,1, 95–102.
Carlson, M., & Miller, N. (1987). Explanation of the relation between
negative mood and helping. Psychological Bulletin,102, 91–108.
Chaitin, J., & Steinberg, S. (2008). ‘‘You should know better’’: Expres-
sions of empathy and disregard among victims of massive social
trauma. Journal of Aggression, Maltreatment and Trauma,17, 197–
226.
Clary, E., & Snyder, M. (1999). The motivations to volunteer: Theoreti-
cal and practical considerations. Current Directions in Psychological
Science,8, 156–159.
Dodge, K., Bates, J., & Pettit, G. (1990). Mechanisms in the cycle of
violence. Science,250, 1678–1683.
Dovidio, J. (1984). Helping behavior and altruism: An empirical and
conceptual overview. In L. Berkowitz (Ed.), Advances in experimental
social psychology (Vol. 17, pp. 361–427). San Diego, CA: Academic
Press.
Dovidio, J., Gaertner, S., Validzic, A., & Matoka, K. (1997).
Extending the benefits of recategorization: Evaluations, self-disclo-
sure, and helping. Journal of Experimental Social Psychology,33,
401–420.
Duncan, L. (1999). Motivation for collective action: Group conscious-
ness as mediator of personality, life experiences, and women’s rights
activism. Political Psychology,20, 611–635.
Fultz, J., Schaller, M., & Cialdini, R. (1988). Empathy, sadness, and dis-
tress: Three related but distinct vicarious affective responses to
another’s suffering. Personality and Social Psychology Bulletin,14,
312–325.
Gaertner, S., & Dovidio, J. (2000). Reducing intergroup bias: The com-
mon ingroup identity model. Philadelphia, PA: Psychology Press.
Gillen, G. (2005). Positive consequences of surviving a stroke. American
Journal of Occupational Therapy,59, 346–350.
Grossman, F., Sorsoli, L., & Kia-Keating, M. (2006). A gale force wind:
Meaning making by male survivors of childhood sexual abuse.
American Journal of Orthopsychiatry,76, 434–443.
Guimond, S., Dif, S., & Aupy, A. (2002). Social identity, relative group
status and intergroup attitudes: When favourable outcomes change
intergroup relations. . . for the worse. European Journal of Social
Psychology,32, 739–760.
Harris, L., & Fiske, T. (2006). Dehumanizing the lowest of the low.
Psychological Science,10, 847–853.
Hobfoll, S. (1991). Traumatic stress: A theory based on rapid loss of
resources. Anxiety Research,4, 187–197.
Janoff-Bulman, R. (1992). Shattered assumptions: Towards a new
psychology of trauma. New York, NY: Free Press.
Kahana, B., Harel, Z., & Kahana, E. (1988). Predictors of psychological
well-being among survivors of the Holocaust. In J. Wilson, Z. Harel
& B. Kahana (Eds.), Human adaptation to extreme stress: From the
Holocaust to Vietnam (pp. 171–192). New York, NY: Plenum.
Kaniasty, K., & Norris, F. (1995). Mobilization and deterioration of
social support following natural disasters. Current Directions in
Psychological Science,4, 94–98.
Kishon-Barash, R., Midlarsky, E., & Johnson, D. R. (1999). Altruism
and the Vietnam War veteran: The relationship of helping to symp-
tomatology. Journal of Traumatic Stress,12, 655–662.
Kleinman, S. (1989). A terrorist hijacking: Victims’ experiences initially
and 9 years later. Journal of Traumatic Stress,2, 49–58.
Knafo, A., Schwartz, S., & Levine, R. (2009). Helping strangers is lower
in embedded cultures. Journal of Cross-Cultural Psychology,40, 875–
879.
Krebs, D. (1975). Empathy and altruism. Journal of Personality and
Social Psychology,32, 1134–1146.
Krous, T. M. D., & Nauta, M. M. (2005). Values, motivations, and
learning experiences of future professionals: Who wants to serve
underserved populations? Professional Psychology: Research and
Practice,36, 688–694.
Levine, M., Prosser, A., Evans, D., & Reicher, S. (2005). Identity and
emergency intervention: How social group membership and inclusiveness
314 VOLLHARDT AND STAUB
of group boundaries shape helping behavior. Personality and Social
Psychology Bulletin,31, 443–453.
Lifton, R. (2003). Super power syndrome. America’s apocalyptic confron-
tation with the world. New York, NY: Nation Books.
London, P. (1970). The rescuers: Motivational hypotheses about Chris-
tians who saved Jews from the Nazis. In J. Macaulay & L. Berkowitz
(Eds.), Altruism and Helping Behavior (pp. 241–250). New York, NY:
Academic Press.
Mason, G., Riger, S., & Foley, L. (2004). The impact of past sexual
experiences on attributions of responsibility for rape. Journal of Inter-
personal Violence,19, 1157–1171.
Mattis, J. S., Hammond, W. P., Grayman, N., Bonacci, M., Brennan, W.,
Cowie, S. A., Ladyzhenskaya, L., & So, S. (1999). The social produc-
tion of altruism: Motivations for caring action in a low-income urban
community. American Journal of Community Psychology,43, 71–84.
Messinger, R. (2004, July 12). Genocide by any other name is still geno-
cide: Crisis in Darfur, Sudan. American Jewish World Service.
Retrieved from http://www.ajws.org/index.cfm?section_id=8&sub_
section_id=14&page_id=265.
Midlarsky, E. (1991). Helping as coping. In M. Clark (Ed.), Prosocial
behavior (pp. 238–264). Thousand Oaks, CA: Sage.
Norris, F. (1990). Screening for traumatic stress: A scale for use in the
general population. Journal of Applied Social Psychology,20, 1704–
1718.
Norris, F., Baker, C., Murphy, A., & Kaniasty, K. (2005). Social sup-
port mobilization and deterioration after Mexico’s 1999 flood: Effects
of context, gender, and time. American Journal of Community Psy-
chology,36, 15–28.
Raboteg-S
ˇaric, Z., Z
ˇuzˇ ul, M., & Kerestes
ˇ, G. (1994). War and children’s
aggressive and prosocial behaviour. European Journal of Personality,
8, 210–212.
Reeves, P., Merriam, S., & Courtnenay, B. (1999). Adaptation to HIV
infection: The development of coping strategies over time. Qualitative
Health Research,9, 344–361.
Staub, E. (1998). Breaking the cycle of genocidal violence: Healing and
reconciliation. In J. Harvey (Ed.), Perspectives on loss (pp. 231–238).
Washington, D.C.: Taylor and Francis.
Staub, E. (2003). The psychology of good and evil: Why children, adults,
and groups help and harm others. MA: Cambridge University Press.
Staub, E. (2005). The roots of goodness: The fulfillment of basic human
needs and the development of caring, helping and nonaggression,
inclusive caring, moral courage, active bystandership, and altruism
born of suffering. In G. Carlo & C. Edwards (Eds.), Nebraska Sympo-
sium on Motivation (Vol. 51). Moral motivation through the life span:
Theory, research, applications (pp. 34–72). Lincoln: University of
Nebraska Press.
Staub, E., Pearlman, L., Gubin, A., & Hagengimana, A. (2005).
Healing, reconciliation, forgiveness and the prevention of violence
after genocide or mass killing: An intervention and its experimental
evaluation in Rwanda. Journal of Social and Clinical Psychology,24,
297–334.
Staub, E., & Vollhardt, J. (2008). Altruism born of suffering: The
roots of caring and helping after experiences of personal and political
victimization. American Journal of Orthopsychiatry,78, 267–
280.
Tedeschi, R. (1999). Violence transformed: Posttraumatic growth in
survivors and their societies. Aggression and Violent Behavior,4, 319–
341.
Tedeschi, R., Park, C., & Calhoun, L. (1998). Posttraumatic growth:
Positive changes in the aftermath of crisis. Mahwah, NJ: Lawrence
Erlbaum.
Turner, R., Crisp, R., & Lambert, E. (2007). Imagining intergroup con-
tact can improve intergroup attitudes. Group Processes Intergroup
Relations,10, 427–441.
Vollhardt, J. R. (2009). Altruism born of suffering and prosocial behav-
ior following adverse life events: A review and conceptual integration.
Social Justice Research,22, 53–97.
Wayment, H. (2004). It could have been me: Vicarious victims and
disaster-focused distress. Personality and Social Psychology Bulletin,
30, 515–528.
Westmaas, J., & Silver, R. (2006). The role of perceived similarity in
supportive responses to victims of negative life events. Personality and
Social Psychology Bulletin,32, 1537–1546.
j
INCLUSIVE ALTRUISM BORN OF SUFFERING 315
