Description
Case Study on Moral Identity and the Expanding Circle of Moral Regard toward Out-Groups
An economic expansion is an increase in the level of economic activity, and of the goods and services available. It is a period of economic growth as measured by a rise in real GDP.[1] The explanation of such fluctuations in aggregate economic activity is one of the primary concerns of macroeconomics.
Case Study on Moral Identity and the Expanding Circle of Moral Regard toward Out-Groups
This article examines moral identity and reactions to out-groups during intergroup conflict. Four studies suggest that a highly self-important moral identity is associated with an expansive circle of moral regard toward out-group members (Study 1) and more favorable attitudes toward relief efforts to aid out-group members (Study 2). Study 3 examines moral identity and national identity influences on the provision of financial assistance to out-groups. Study 4 investigates the relationship between moral identity and (a) the willingness to harm innocent out-group members not involved in the conflict and (b) moral judgments of revenge and forgiveness toward out-group members directly responsible for transgressions against the ingroup. Results are discussed in terms of self-regulatory mechanisms that mitigate in-group favoritism and out-group hostility.
A well-established finding in social psychology is the in-group favoritism and out-group hostility that occurs during intergroup conflict. For many Americans, the September 11th attacks on the World Trade Center and Pentagon had profound psychological effects on in-group national identity. Not surprisingly, these events triggered a variety of hostile reactions toward out-group members perceived as being closely, or even loosely, associated with the
Americus Reed II, Marketing Department, The Wharton School, University of Pennsylvania; Karl F. Aquino, Management Department, College of Business, University of Delaware. Both authors contributed equally to this research. Order of authorship was reverse alphabetical. This study was partially supported by a College of Business and Economics Research Summer Grant awarded to Karl F. Aquino. For either administrative or data collection assistance or for comments on a draft of the article, we thank Alan Ercan Aquino, Mark Forehand, Gavan Fitzsimmons, Nancy Kim, Karen Reed, Christophe Van den Bulte, Patti Williams, and participants of the four-school (Columbia University; New York University; The Wharton School, University of Pennsylvania; and Yale University) Marketing Colloquium held at New York University in May of 2002. Young Lee is especially thanked for assistance with computer programming for Studies 2 and 4. Steve Grover is especially thanked for data collection assistance for Study 3. Correspondence concerning this article should be addressed to Americus Reed II, The Wharton School, University of Pennsylvania, 3730 Walnut Street, 700 Jon M. Huntsman Hall, Philadelphia, Pennsylvania 191046371. E-mail: [email protected]
terrorist attacks. For example, reports of anti-Muslim sentiment expressed by some Americans were evidenced by increased racial and ethnic violence, heightened suspicion and mistrust of immigrants from Middle-East countries, and public support for targeted surveillance of people whose physical characteristics fit the terrorist prototype (cf. American Arab Anti Discrimination Committee: http://www.adc.org/education/advice.htm) One explanation for out-group hostility in the aftermath of September 11th is self-protective motives at either the individual (selfidentity) or group (social identity) level. That is, out-group denigration often upholds the self- or national image in the face of an external threat. Although the motivation to enhance one's self or one's group by negatively evaluating others is unquestionably powerful, some people are less susceptible to this motivation. For example, in contrast to the hostility that some Americans showed toward Muslims and Arab Americans after September 11th, others pleaded for racial tolerance and openly condemned acts of discrimination directed against fellow citizens and even noncitizens (cf. Coalition for World Peace: http://www.actionla.org/peace/#point). This article explores a possible social psychological explanation for the attitudes and behaviors of people who respond less negatively toward out-groups even in the midst of inter- group conflict. This explanation derives from the psychological effects of a self-regulating construct referred to as moral iden- tity that connects the individual (personal identity) to others (social identity) through the evaluative implications of a set of moral associations that define the moral self (Aquino & Reed, 2002). We hypothesize that moral identity can influence out- group hostility because its self-importance alters the psycho- logical boundaries that define in-group membership. When the activation potential of moral identity is high —that is, when moral identity assumes high self-importance—the self/other relation should be characterized by a more expansive concep- tion of the in-group toward which a person feels obligated to
1270
MORAL IDENTITY AND OUT-GROUPS
1271
exhibit moral regard.1 We refer to this premise as the "circle of moral regard" hypothesis: Rather than confining one's in-group to a narrow set of others (e.g., family, kin, fellow citizens), a person whose moral identity has high self-importance should include a larger set of social groups. In the extreme, this psychological boundary might include all of humanity. The goal of this article is to broadly examine the circle of moral regard hypothesis. We begin by briefly reviewing the literature on self-protective mechanisms based on personal (self-affirmation) and social (self-esteem through group status) identities. Next, some key conditions that may generate in-group favoritism and out-group hostility are described. Third, the concept of moral identity is defined and its relationship to out-group delineations and in-group favoritism is highlighted. Finally, the results of four studies that have investigated the connection between moral identity and perceptions and behavior toward out-groups are reported and discussed.
Triggers of Out-Group Hostility
Out-group hostility is often triggered when out-group members pose a realistic threat to in-group physical and psychological well-being (e.g., Quillian, 1995) or symbolically threaten the stability of in-group values and beliefs (Esses, Haddock, & Zanna, 1993) or when encounters with out-group members cause anxiety as a result of condemnation and rejection by those out-group members (e.g., Stephan & Stephan, 1993). Moreover, if particular circumstances (e.g., a vivid or dramatic event) make intergroup differences highly salient, then broad-based stereotypes of outgroup members can evoke unfavorable expectations of conflictladen interactions (e.g., Stephan, Ybarra, Martinez, Schwarzwald, & Tur-Kaspa, 1998). The attacks on the World Trade Center and Pentagon created many of the conditions that give rise to hostility toward outgroups. As a result, for some Americans, the image of Arab Americans and Muslims now begins and ends with a terrorist group prototype. This image lays the psychological foundation for acts of prejudice and discrimination directed toward the target out-group (Deaux, 2002). In the wake of strong situational forces invoked in the aftermath of September 11th, an increase in hostility directed toward out-group members fitting the terrorist prototype would seem inevitable. Yet, as mentioned earlier, many Americans have openly condemned acts of discrimination and some have gone out of their way to provide comfort and protection for their Arab and Muslim neighbors. For example, when one suburban mosque was vandalized after the September 11th terrorist attacks, a group of concerned American citizens volunteered at considerable personal risk to accompany Islamic women to the grocery store, to medical appointments, or on other errands after discovering that these Muslim women were afraid to go out in public (Dart, 2001). We theorize that the attitudes and behaviors of people who assisted these out-group members may be partially explained by a person's conception of a moral self (Aquino &Reed, 2002) that defines the self's relation to others and establishes the psychological boundaries between in-groups (us) and out- groups (them).
Self-Protection Through Self-Affirmation of Personal and Group Identity
Several studies have shown that self-image threat can lead to negative evaluations of out-groups (Brown & Gallagher, 1992; Gibbons & Gerrard, 1991). One reason this occurs is because information that threatens perceivers' sense of self-worth evokes the need to restore a positive self-image. This goal can be achieved by negatively evaluating out-groups (cf. Steele, 1988; Steele &Liu, 1983; Steele, Spencer, &Lynch, 1993). Unless other internal motivations are activated, such as a goal of accountability (Tetlock, 1983), social desirability and/or egalitarian motives (Dovidio & Gaertner, 1991), or self-image bolstered with a self-affirmation procedure (Fein &Spencer, 1997; Study 1), people may find the denigration of out-groups to be an effective way of maintaining "an image of self-integrity" (Steele et al., 1993, p. 885). Social identity theory (Tajfel, 1959; Tajfel & Turner, 1979), an influential theoretical explanation for in-group favoritism and outgroup denigration at the group level, makes a similar assertion. Like the self-affirming processes at the individual level, positive evaluations of in-groups and unfavorable responses toward outgroups are presumed to serve self-protective and self-enhancing functions. One way people maintain or increase their self-esteem is through group membership (Turner, Hogg, Oakes, Reicher, &Wetherell, 1987). Consequently, it is beneficial for people to ascribe favorable characteristics to the groups to which they belong (Hewstone, Rubin & Willis, 2002). Once the concept of "us" is established, however, self-esteem can also be enhanced by negatively evaluating groups to which one does not belong. For example, a person may selectively search for intergroup differences that favor their group and dismiss information that favors the out-group (e.g., Brewer & Kramer, 1985; Devine, 1989; Mikulincer & Shaver, 2001). This can lead them to view out-group members as inferiors. Self-serving causal explanations for ingroup outcomes (e.g., Hamilton & Trolier, 1986) and the use of linguistic labels to differentiate in-groups and out-groups also increase the salience of intergroup differences (e.g., Dovidio & Gaertner, 1993), further encouraging in-group favoritism. Several conditions seem to reliably invoke these kinds of in-group/outgroup effects. The more prominent are reviewed briefly as follows.
The Role of a Moral Self-Conception
It has been argued that a defining characteristic of a person with a legitimate moral identity is that he or she extends feelings of sympathy and affiliation toward a larger segment of humanity than someone whose moral identity is less important (Hart, Atkins & Ford, 1998; Younis & Yates, 1999). If this assertion is valid, then compared with those whose moral identity has low selfimportance, people whose moral identity has high self-importance should be relatively less likely to exhibit in-group favoritism and out-group hostility even during times of intergroup conflict. One of the problems with testing this prediction is that there have been few attempts to provide a definition of moral identity that corresponds to a psychometrically sound measure of its self-importance.
1 Following the definition of morality proposed by moral philosophers (e.g., Hume, 1751/1996) and psychologists (e.g., Eisenberg, 2000; Rest, 1979), moral regard is defined in this study as showing concern for the needs and welfare of others.
1272
REED AND AQUINO
Recent work by Aquino and Reed (2002) addressed this issue by offering both a theoretical framework and a methodological approach for testing the relationship between moral identity and ingroup/out-group effects.
gence of a spontaneous self-concept organized around moral themes (cf. Hart & Fegley, 1995), self-reported volunteering, and the actual donation of food to the needy.
Moral Identity: A Theoretical Framework
Aquino and Reed (2002) proposed a trait-based conceptualization of moral identity and presented empirical evidence supporting the construct and predictive validity of a measure of its selfimportance. They defined moral identity as "a self-conception organized around a set of moral traits." On the basis of social cognitive theories of the self (cf. Kihlstrom & Klein, 1994), Aquino and Reed (2002) argued that moral identity is linked to specific moral traits (cf. Lapsley & Lasky, 2001; Walker & Pitts, 1998) and can, therefore, be invoked by tapping a subset of traits that are associatively linked to other traits that form part of the person's unique moral self-conception. They further proposed that because moral identity is amenable to a distinct mental image of what a moral person is likely to think, feel, and do (Kihlstrom & Klein, 1994), it can be associated with a social referent that could be a real membership group (e.g., fellow Peace Corp volunteers), an abstracted ideal (e.g., God), a known individual (e.g., a big brother in a mentoring program), an unknown individual (e.g., Mother Teresa), or any social construction (cf. Reed, 2002). Consequently, moral identity is one of the many possible social identities that comprise a person's social self-schema (cf. Markus, 1977). Because the traits comprising this identity have moral content, however, its self-importance should be a stronger predictor of moral affect, cognitions, and behaviors than other identities (e.g., occupation, gender) that are less likely to be imbued with moral meaning (cf. Aquino & Reed, 2002, 2003; Reed, 2002; Reed, Aquino, & Ray, 2003). Building on their trait-based conception, Aquino and Reed (2002) developed an instrument that taps two aspects of identity as defined by Erikson (1964): identity as being rooted in the very core of one's being and as being true to one's self in action. Procedurally, Aquino and Reed's (2002) instrument uses trait stimuli to make a moral identity schema salient and then asks respondents several questions to measure the self-importance of their unique moral identities along the two dimensions (see Appendix).2 In six studies using diverse samples, Aquino and Reed (2002) showed that the items tap the two dimensions of identity that are consistent with Erikson's (1964) conception of an identity as being rooted in the very core of one's being and as being true to oneself in action. Aquino and Reed (2002) labeled these dimensions internalization and symbolization. Internalization reflects the degree to which a set of moral traits is central to the self-concept, whereas Symbolization reflects the degree to which these traits are expressed publicly through the person's actions in the world.3 This latter dimension corresponds to Laughlin's (1970) definition of symbolization as the process "through which an external object becomes the disguised outward representation for another internal and hidden object, idea, person, or complex" (p. 414). Aquino and Reed demonstrated that the instrument they developed to measure these dimensions of moral identity had a stable factor structure; was internally consistent; possessed nomological validity; and predicted a variety of morally relevant outcomes, including the emer-
Present Research
In this research we examine whether the self-importance of moral identity as defined by the two dimensions of internalization and symbolization (Aquino & Reed, 2002) is associated with an expansive circle of moral regard. Does a highly self-important moral identity diminish the psychological boundary between ingroups and out-groups and extend the web of sympathy toward
2 Following Blasi (1984), Aquino and Reed (2002) made several assumptions about moral identity that are also assumed in this article. First, they acknowledge that the specific content of people's moral identities is likely to vary. So whereas one person may see being compassionate as critical to his or her moral identity, another person might emphasize being just and fair. Even though there may be some nonoverlapping moral traits that compose each person's unique moral identity, a common set of traits appears to be central to many individuals' moral self-definitions (Aquino &Reed, 2002; Lapsley & Lasky, 2001; Walker & Pitts, 1998). Second, Aquino and Reed assume that being a moral person may be, but need not be, a part of a person's overall self-definition. This assumption is consistent with the notion of the self as comprising multiple identities (Brewer & Gardner, 1996) that are hierarchically ordered (Stryker, 1980). Consequently, the self-importance of moral identity may change as a function of place and time (Hart et al., 1998). Last, Aquino and Reed (2002) do not claim that having a highly self-important moral identity makes a normative or prescriptive statement about a person's moral character or beliefs. This last point is important because there are many different views of morality and what the ideal moral character should be. Aquino and Reed's conception of moral identity does not presume to reflect or endorse any particular moral view or position about what moral self-representation a person ought to possess (cf. Aquino & Reed, 2003). 3 One way to view Aquino and Reed's (2002) instrument, or any instrument that asks people to rate themselves relative to any dimension of the self, is from an impression management perspective (Schlenker, 1985). In other words, one could argue that the moral identity instrument (see Appendix) captures merely the importance of maintaining and conveying a socially constructed image of a moral person. Such motivation could conceivably be purely self-presentational. However, the conceptual definition of moral identity offered by Aquino and Reed and the underlying dimensions linked to Erikson's notion of identity assume that the operationalization of moral identity using Aquino and Reed's instrument captures (at least along the internalization dimension) the importance of how critical that social image is a part of a person's objective sense of who they are (i.e., their actual self). Although the instrument can be thought of as tapping merely into the projected image of a social construction, Aquino and Reed have presented empirical evidence that the instrument taps into a deeper sense of self-relevance. For example, in Study 2 of their article, they show a strong correlation between the internalization dimension and an Implicit Association Test (Greenwald et al., 1998; Greenwald & Farnham, 2000) modified to include the moral traits used as salience induction stimuli in their instrument. To the extent that this implicit measure reflects aspects of self that are not amenable merely to self-presentational projected image concerns, then there is at least some preliminary evidence that the measure really captures aspects beyond the social desirability to adhere to the constructed image of a moral person—which is consistent with the definition of moral identity used by Aquino and Reed and also used in this article.
MORAL IDENTITY AND OUT-GROUPS
1273
others? If so, what are its cognitive and behavioral consequences? Four studies examine these questions. In Study 1, the positive relation between the self-importance of moral identity and the tendency to self-report greater moral regard for out-groups is empirically established. Studies 2- 4 examine the implications of an expanded circle of moral regard in the context of the September 11th terrorist attacks. Study 2 examines whether the selfimportance of moral identity is related to judgments of relief efforts intended to aid out-groups that are either closely or not closely associated with a salient intergroup conflict. Study 3 builds on Study 2 by testing whether a highly self-important moral identity predicts whether people differentially provide actual monetary assistance to equally deserving in-group and out-group members. Finally, Study 4 further explores the circle of moral regard hypothesis by examining the effect of a highly self-important moral identity on negative responses toward innocent out-group members as well as moral judgments about how to appropriately respond to out-group members held directly responsible for ingroup transgressions.
between these constructs. However, there is another, more indirect approach that provides a further test of the circle of moral regard hypothesis. This approach applies a well-established social psychological theory about how the kinds of resources that people exchange with one another in everyday life depend upon the nature and quality of relationships with exchange partners.
Exchange of Particularistic Resources
Social resource theory (Foa & Foa, 1974) posits that humans exchange six basic resources in social relationships: money, information, goods, services, status, and love. Foa and Foa (1974) refer to money, goods, and information as universalistic resources. In contrast, love, status, and services are referred to as particularistic resources. Social resource theory predicts that the qualities of the exchange partner and the nature of the relationship are more (less) important to the decision to exchange a social resource when the exchange involves a particularistic (universalistic) resource. For example, most people are willing to exchange money, which is the most universalistic resource, with just about anybody. They are much more selective when deciding with whom to exchange love or status. Indeed, studies have shown that particularistic resources are exchanged more frequently with people who are psychologically close rather than distant (e.g., Blieszner, 1993). On the basis of these findings, it is hypothesized that if moral identity is associated with an expansive circle of moral regard, then as the self-importance of this identity increases, so too should the person's perceived moral obligation to exchange highly particularistic resources with a stranger. The reason being that for a person with a highly self-important moral identity, a "stranger" is more likely to be included within the circle of moral regard. On the basis of these arguments, Study 1 tested two hypotheses. The first is that in terms of the self-importance of moral identity, high as compared with low moral identifiers would report having a more expansive circle of moral regard. The second hypothesis is that high as compared with low moral identifiers would be more willing to exchange particularistic resources with strangers.
Study 1: Moral Identity and the Circle of Moral Regard
It has been suggested that a person whose moral identity has high self-importance should show greater concern for a larger segment of humanity than someone whose moral identity is less important (Hart et al., 1998; Younis & Yates, 1999). This argument parallels the philosopher Peter Singer's (1981) assertion that moral progress can emerge from a fixed moral sense if people are willing to expand the mental dotted line that embraces the entities deemed worthy of moral consideration. According to Younis and Yates (1999), studies of European adults known to have protected and sheltered Jews during the Nazi occupation of World War II (e.g., Monroe & Epperson, 1994; Oliner & Oliner, 1988) provide anecdotal exemplars of this cognitive expansion of moral regard. These studies found that one characteristic that distinguished rescuers from nonrescuers is that the former tended to perceive Jews as fellow human beings who merited the same privileges and community rights. In contrast, nonrescuers viewed Jews as outsiders, as "they" rather than non-Jewish "we." According to Monroe and Epperson (1994), the ethical actions of rescuers flowed spontaneously from deep-seated dispositions revealed in the self's relationship to others.
Method Samples and Procedures
Data from two independent samples drawn from two different universities located in the northeastern United States were used to test the study hypotheses. The first sample (Sample 1) consisted of 137 undergraduates enrolled in an introductory marketing class. This sample was used to test the hypothesized relationship between the self-importance of moral identity and perceived moral obligations toward out-groups. Students received course credit for their participation. Seventy-one of the students were women. Fifty percent of the students identified themselves as White, 35% as Asian, 5% as Hispanic, 3% as African American, 2% as Middle Eastern, and 5% as other. Their average age was 19.4 years (SD 0.89). To fulfill a requirement for the marketing class, participants completed several penciland-paper questionnaires throughout the semester. As part of one of those sessions, participants arrived at the behavioral laboratory where they anonymously completed a measure of the self-importance of moral identity and several questions intended to assess their perceived obligation to show concern for various groups. After data collection was complete, participants were fully debriefed, thanked for their participation, and dismissed.
Whose Needs Merit Moral Consideration?
The self- other relationship is fundamental to a person's moral identity (Younis & Yates, 1999). As one rescuer interviewed by Oliner and Oliner (1988) put it: "I insist on saying that it was absolutely natural to have done this [rescuing]. You don't have to glorify yourself— considering that we are all children of God and that it is impossible to distinguish between one human and another" (p. 228).Younis and Yates (1999) interpreted this rescuer's rationale as evidence that a psychological consequence of a selfimportant moral identity is that it promotes a belief that all people share a common humanity. If this interpretation is valid, then there should be a positive relationship between the self-importance of moral identity and the extent to which people feel obligated to show moral concern for the welfare of out-groups. One way to evaluate this assertion directly is to show an empirical relationship
1274
REED AND AQUINO Following an approach that Turner, Foa, and Foa (1971) used to test social resource theory, participants were then presented with four descriptive statements of social exchanges involving particularistic resources. In two separate scenarios, participants were told that the resources were to be exchanged with both a "family member you feel closest to" and a "stranger." Two of the exchanges involved love ("Telling them I love you," "Telling them you feel affection for them") and two involved status ("Telling them you are proud of them," "Telling them you respect and admire them"). Participants were then asked to indicate how obligated they were to exchange these resources with a family member and a stranger. Responses were on a 5-point Likert scale that ranged from 1 (not at all obligated) to 5 (very much obligated). In each of the two scenarios, the four items describing the exchanges were combined into a single scale representing particularistic resources. The Cronbach's alpha reliabilities for this scale were .87 and .70 in the family member and stranger scenarios, respectively.
The second sample (Sample 2) consisted of 55 master's of business administration students enrolled in two master's-level organizational behavior classes. Sample 2 was used to test the hypothesis that, compared with low moral identifiers, high moral identifiers would be more willing to exchange particularistic resources with strangers. The students completed the moral identity self-importance measure prior to a class lecture on the role of personality in organizations. One month later they completed a questionnaire assessing the extent to which they felt obligated to exchange various kinds of particularistic resources with both a friend and a stranger. The confidentiality of students' responses was maintained by assigning code numbers to each survey. After completing both questionnaires, the results were reported to students in the aggregate and were used as the basis for a class discussion on social exchange theory. Forty-eight students provided usable data on both questionnaires. Thirty-nine respondents were men and 89% were White. Their average age was 29.1 years (SD 6.0).
Measures
Moral identity. Aquino and Reed's (2002) 10-item Self-Importance of Moral Identity Scale was used to measure this construct (see Appendix). The theoretical basis for the instrument is that by tapping certain trait stimuli, a broader associative network of related traits is invoked. Past research has shown that stimulus cues such as words can heighten the salience of a particular social identity (Forehand, Deshpande, & Reed, ´ 2002). Using an inductive procedure, Aquino and Reed identified a set of traits that many people may consider to be characteristic of a moral person (Lapsley & Lasky, 2001; Walker & Pitts, 1998). These traits were then used as identity-invoking stimuli in the measure, recognizing that the traits do not comprise an exhaustive set which all persons in all places would consider as being central to their unique moral identities. Respondents in Samples 1 and 2 answered the 10 questions on a 5-point Likert scale that ranged from 1 (strongly disagree) to 5 (strongly agree). Items for the Internalization and Symbolization subscales were averaged. The Cronbach's alpha reliability for both subscales was .83 in Sample 1, and the means were 4.5 (SD 0.5) and 3.1 (SD .7) for Internalization and Symbolization, respectively. In Sample 2, reliabilities were s .85 and .71 with means of 4.3 (SD 0.5) and 3.0 (SD 0.6) for the Internalization and Symbolization subscales, respectively. Moral regard for out-groups. The expansiveness of a person's circle of moral regard was measured by asking Sample 1 participants to report the extent to which they believed they had "a moral or ethical obligation to show concern for the welfare and interests" of four different out-groups: "People from another country," "Strangers," "People who practice a different religion than you," and "People of different ethnicities than you." Respondents answered the items on a 5-point Likert scale that ranged from 1 (absolutely no obligation) to 5 (very strong obligation). These items were averaged to form a scale ( .88) where a high score indicated a stronger moral obligation to show concern for the welfare of the four outgroups. Willingness to exchange particularistic resources. Sample 2 participants were asked to indicate the extent to which they felt obligated to provide to both a friend and a stranger, love and status—two highly particularistic resources identified by Foa and Foa (1980). To elicit these responses, participants were presented with the following instructions: All of us have many different relationships in our lives. In these relationships, we give and receive certain things from one another that may be described as "social resources." These resources may be material, like a diamond ring, or symbolic, like a smile. Listed below are several statements describing different types of social resources. Using the scale below, please indicate how obligated you are to provide each in the context of the relationship described, assuming that an appropriate opportunity presents itself for you to do so .
Results Concern for Out-Groups
The hypothesis that high moral identifiers would report a more expansive circle of moral regard was tested by regressing perceived moral obligation toward out-groups on the two dimensions of the self-importance of moral identity measure. Gender was used as a control variable because past research has shown that men tend to show greater intergroup bias than women (Sidanius, Levin, Liu, & Pratto, 2000; for reviews, see Pratto, Sidanius, & Stallworth, 1993; cf. Gaertner & Insko, 2000). Participants' race was also controlled because it is possible that compared with members of the socially dominant group (e.g., Whites), people who belong to socially subordinate racial groups (e.g., African Americans, Hispanics, Asians) within American society might identify more closely with out-groups. This might lead the latter to report stronger moral obligations to be concerned about out-group welfare, irrespective of the self-importance of their moral identity. Both control variables were dummy coded (0 male, 1 female and 0 non-White, 1 White). Results of this analysis showed one significant effect. The Internalization dimension was positively related to perceived moral obligations toward out-groups ( .30, p .01). The direction of the effect is consistent with the prediction that the higher a person's self-importance of moral identity, the more likely he or she was to report a stronger moral obligation to show concern for a variety of out-groups. This result is interpreted as supporting the hypothesis that high moral identifiers have a more expansive circle of moral regard than low moral identifiers. However, this was only true for the internalization dimension of the self-importance of moral identity measure. Having established a direct empirical relationship between the self-importance of a person's moral identity and concern for outgroups, Sample 2 data were analyzed to provide a further test of the circle of moral regard hypothesis using the predictions of social resource theory.
Exchange of Particularistic Resources
The second test of whether high moral identifiers have a more expansive circle of moral regard than low moral identifiers was performed by first dividing participants into two groups based on their scores on the two Self-Importance of Moral Identity sub-
MORAL IDENTITY AND OUT-GROUPS
1275
scales. Group assignment was determined by a median split on the Internalization, high (n 24) versus low (n 25), and Symbolization, high (n 25) versus low (n 24), subscales. Next, a 2 2 2 (Recipient Internalization Symbolization) analysis of variance with repeated measures on recipient (friend vs. stranger) was performed. There were no significant differences in the dependent variables as a function of gender, so this variable was not included in the analysis. Race was also not included because Sample 2 was composed almost exclusively of Whites, so there was insufficient variation on this characteristic to make meaningful comparisons. Results showed a significant main effect of recipient such that the participants reported a stronger moral obligation to exchange particularistic resources with a friend (M 4.4) than a stranger (M 1.7), F(1, 45) 293.86, p .01. This result supports social resource theory (Foa & Foa, 1980). A significant main effect of internalization was also found, F(1, 45) 4.49, p .05. The effect indicated that participants who scored higher on the Internalization subscale felt more obligated to exchange particularistic resources with both friends and strangers (M 3.2) than people who scored lower on this subscale (M 2.9). However, both of these main effects are qualified by a significant Recipient Internalization interaction, F(1, 45) 5.79, p .01. The pattern of means for this interaction supports the prediction that high moral identifiers are more willing to exchange particularistic resources with strangers than low moral identifiers. Specifically, high (M 4.4) and low (M 4.4) scorers on Internalization did not differ in their perceived obligations to exchange particularistic resources with a friend, t(47) .35, ns. However, they differed significantly in their felt obligation to exchange these resources with a stranger, t(47) 3.26, p .01, with high moral identifiers (M 2.1) feeling more obligated than low moral identifiers (M 1.4). No other main or interaction effects were significant.
feels strong bonds of attachment, trust, or affiliation. Data from Sample 2 suggest that as a person's moral identity gains selfimportance, he or she feels a stronger moral obligation to exchange such resources with strangers. This finding is interpreted as providing further evidence that having a self-important moral identity might be associated with a cognitive expansion of the boundaries of in-group membership. The data for Sample 2 were collected at two different time points and used adult nonstudent as compared with student participants. However, Sample 2 also has limitations. One is that the operationalization of out-group used in Sample 2 (i.e., the stranger scenario) is a somewhat abstract and far removed notion of the psychological concept of an out-group. Second, it was argued earlier that the expanded circle of moral regard would persist for high moral identifiers even when strong situational pressures (such as intergroup conflict) establish more distinct psychological boundaries between in-groups and out-groups. The Sample 2 data included no such situational factors. Finally, the dependent measure of perceived obligation does not necessarily mean that people would in fact behave differently toward out-groups as a function of the self-importance of moral identity. The limitations of both Samples 1 and 2 of Study 1 suggest that a stronger test of the circle of moral regard hypothesis would require an examination of a person's reactions toward real outgroups closely linked to salient intergroup conflict. It should also include a measure of actual behavior. These were the goals of the next two studies. In Studies 2 and 3, people's attitudes, intentions, and behavior toward organizations that render aid to specific out-groups were examined as a function of the self-importance of moral identity.
Study 2: Judgments of Relief Efforts That Aid Out-Groups
The findings from Study 1 offer preliminary evidence supporting the proposed link between the self-importance of moral identity and the expanding circle of moral regard. Study 2 presents a further test of this prediction by examining how people's moral identity self-importance influences perceptions of charitable organizations that aid specific out-groups. If a moral self is indeed associated with a more expansive circle of moral regard, a highly self-important moral identity should be associated with more favorable perceptions and judgments of relief efforts meant to aid out-groups (e.g., relief effort for Middle East refugees). More specifically, compared with low moral identifiers, high moral identifiers should evaluate these efforts as more worthwhile. Furthermore, in the aftermath of September 11th, it is hypothesized that this difference is influenced by whether the out-group receiving assistance is loosely associated with the event (e.g., Afghani refugees) as compared to a more neutral (e.g., Turkish refugees) out-group prototype. If a highly self-important moral identity mitigates out-group negativity by expanding the circle of moral regard and broadening the web of sympathy toward others, then high as compared with low moral identifiers should evaluate relief efforts that aid out-groups more favorably in general. However, this relation is expected to be stronger for high as compared with low moral identifiers even if the out-group (e.g., Afghani refugees)
Discussion and Limitations of Study 1
Of the two dimensions of the self-importance measure of moral identity proposed by Aquino and Reed (2002), the data from Sample 1 showed that internalization was a significant predictor of the extent to which people reported a moral obligation to show concern for the needs and welfare of out-groups (e.g., people from other countries, people of different religions, people of different ethnicities). A limitation of Sample 1 was that the independent and dependent measures were collected at the same time so the relationship between the self-importance of moral identity measure and perceived obligations toward out-groups might have been inflated because of common method bias. Moreover, the crosssectional nature of the research design used for Sample 1 may subject the results to the alternative explanation of self-generated validity (cf. Feldman & Lynch, 1988). The results using Sample 2 addressed these limitations while providing a further test of the proposed relationship between the self-importance of moral identity and the circle of moral regard, but using a different theoretical approach. Social resource theory (Foa & Foa, 1980) posits that particularistic resources are perceived as more personal and therefore are more likely to be exchanged only with those for whom the person
1276
REED AND AQUINO Thousands of pregnant women are among the Kurdish (Afghani) civilians who have fled their homes in recent days and are massed along the country's borders. The lack of shelter, food and medical care, and unsanitary conditions pose a serious risk to these women and their infant children. Even before the current crisis, poor health conditions and malnutrition made pregnancy and childbirth exceptionally dangerous for Kurdish (Afghani) women. UNPF is pre-positioning emergency relief supplies in the countries bordering Turkey (Afghanistan) in order to provide displaced Kurdish (Afghani) women with lifesaving reproductive health care services. The Fund is asking donors for $4.5 million to counter the health risks facing refugees and the internally displaced. PRESS RELEASE: UNITED NATIONS, New York, 28 September 2001—Responding to the grave health emergency now facing Kurdish (Afghani) women, the United Nations Population Fund (UNPF) is mounting its largest-ever humanitarian operation. The Fund is asking international donors for $4.5 million to support the effort. The charitable efforts were dummy coded (0 Turkish, 1 Afghani) for the purposes of analysis. Self-importance of moral identity. Aquino and Reed's (2002) 10-item Self-Importance of Moral Identity Scale was administered during the consumer questionnaire exercise. Respondents answered the items on a 7point Likert scale that ranged from 1 (strongly disagree) to 7 (strongly agree). The test was administered by computer. Items measuring the .87) and Symbolization ( .81) dimensions were Internalization ( averaged to form two subscales. Control variables. Gender and race were again entered as control variables and both were dummy coded (0 Male, 1 Female and 0 Non-White, 1 White).
is loosely associated with an intergroup conflict episode (e.g., the attacks on the World Trade Center and Pentagon).
Method Participants
Seventy-nine undergraduate students at a university located in the northeastern United States participated in the experiment. Average age was 19.6 years (SD 1.15). Forty-one were men and 38 were women. Thirty-seven identified themselves as White, 4 were African American, 27 were Asian, 6 were Hispanic, and 5 identified themselves as other. Seventy-five participants provided usable data on all study variables.
Procedure
The study took place 5 months after the September 11th terrorist attacks. Participants came to a behavioral laboratory and completed seven ostensibly unrelated research tasks for class credit. The first task was to complete a consumer questionnaire exercise. After completing three separate "filler tasks" that took approximately 30 min, participants completed an online research survey that measured perceptions of charitable organizations. Participants were shown a picture of an online brochure of a particular humanitarian relief effort to be described later. They were asked to read the information. They were then asked their perceptions of the organization and the relief effort as well as their intentions to donate. The confidentiality of students' responses was strictly maintained by assigning code numbers to all instruments. Participants then completed two additional filler tasks, were debriefed, and dismissed.
Independent Variables
Out-group target of relief effort. Participants were presented with information about one of two different efforts by the United Nations to aid out-groups. This information was presented using mock-up online brochures. Participants were randomly assigned to either the salient conflict out-group (Afghan) or the neutral out-group (Turkey) condition. All of the brochure information was identical except for the target group who would benefit from the relief effort. A picture of a Middle-Eastern woman and her child was shown along with the headline: Humanitarian Crisis in Turkey (Afghanistan). The information on the online brochure read as follows: The United Nations Population Fund (UNPF) is mounting its largestever humanitarian operation to save the lives of the most vulnerable victims of the current exodus and health crisis facing Turkey (Afghanistan): women and girls. Hospitals and clinics are in ruins after 23 years of war. Tens of thousands of civilians are on the move fearing famine and death.
Dependent Variables
Five items were written to assess judgments of the perceived worthiness of the United Nations' efforts to aid the two out-groups (ranked from 1 [strongly disagree] to 7 [strongly agree]). These items were as follows: "The above cause is not as important as other causes here in the U.S." (reverse scored); "Our government should provide money in this effort"; "If given the opportunity, I would donate to this effort"; "I don't feel very obligated to this particular cause" (reverse scored); and "I would be willing to volunteer my time to help this organization." A principal-components analysis showed that these items loaded on a single factor accounting for 57% of the variance. Consequently, they were averaged to form a scale ( .81).
Results
Descriptive statistics and correlations among the study variables are shown in Table 1. Hierarchical regression analyses were used
Table 1 Study 2: Descriptive Statistics and Correlations Among Study Variables
Study variable 1. 2. 3. 4. 5. Worthiness of relief effort Female White Afghan relief effort Internalization 75. ** p .01. *** p M 5.4 0.5 0.5 0.5 5.2 .001. SD 1.3 0.5 0.5 0.5 0.6 1 — .18 .25* .21 .18 2 — .12 .18 .37*** 3 4 5 6
— .09 .06
— .21
—6.
Symbolization 2.9
0.7
.01
Note. N * p .05.
MORAL IDENTITY AND OUT-GROUPS
1277
to test the study hypotheses. The control and independent variables were entered in Step 1. The interactions between the two outgroups (Turkish vs. Afghani) served by the charitable organization and participants' scores on the Internalization and Symbolization subscales were entered in Step 2. In order to minimize multicollinearity between the interaction terms and their components (Aiken & West, 1991), all of the variables were centered prior to testing hypotheses. The results of the analysis are displayed in Table 2. The significance of individual regression weights was appraised to test the hypothesis that high moral identifiers judge the worthiness of the two relief efforts more favorably than low moral identifiers. The variance explained by the interaction terms was examined to test the hypothesis of differential perceptions of out-groups that are associated with a salient intergroup conflict (Afghani refugees and the September 11th attacks). As the hypothesis for a direct effect of self-importance of moral identity was directional, a one-tailed test was used to test this prediction. The results showed that two variables—race and internalization—were directly related to judgments of the worthiness of the relief effort. The effect of race indicated that White respondents judged the relief efforts less favorably than non-Whites ( .63, p .05). There was also a marginally significant effect of the target of the relief effort, with participants evaluating the Afghan effort more favorably than the Kurdish effort ( .58, p .10). Using a one-tailed test, we found that the effect of internalization supported the prediction that the self-importance of moral identity is positively related to the perceived worthiness of relief efforts directed toward out-groups ( .50, p .05). This main effect, however, is qualified by a significant interaction. The set of interactions explained a significant amount of incremental variance in the dependent variables (R2 .08, p .05). Inspection of the individual regression weights showed that the Internalization Out-group target interaction term was significant (B 1.35, p .05). We analyzed the form of the interaction by regressing judgments of cause worthiness on internalization within groups of participants who evaluated either the Afghan or Turkish
relief effort. The simple slopes in each of these groups showed that the relationship between internalization and judgments of worthiness was considerably stronger when participants evaluated the Afghani (B .58, p .01) as compared with the Turkish (B .05, ns) relief effort. This finding was consistent with the prediction that in terms of self-importance, low moral identifiers would be much less supportive of relief efforts that benefit an out-group closely associated with the events of September 11th.
Discussion and Limitations
The results of Study 2 lend partial support for the argument that the self-importance of moral identity is associated with an expanded circle of moral regard by means of differential perceptions of real out-groups. The direction of the effect indicates that the higher the self-importance of a person's moral identity, as indicated by the Internalization dimension, the more favorable one's evaluation of a relief effort directed toward two distinct outgroups, controlling for the effects of race and gender. Moreover, self-importance of moral identity showed a stronger effect on this evaluation when the out-group (Afghani refugees) was more likely to be loosely associated with a group —the Taliban regime in Afghanistan— directly implicated in the attacks on the World Trade Center and Pentagon. One interpretation of these findings is that in the face of a powerful contextual influence, low moral identifiers were more likely to alter their judgments of the worthiness of relief efforts meant to aid different out-groups based on the degree to which the out-group invoked an association to the perpetrators of the attacks.4 By inference, this effect can be explained by low moral identifiers having a more restricted and high moral identifiers a more inclusive circle of moral regard. This argument is further upheld by the direct relationship between the self-importance of moral identity and judgments of relief efforts to aid out-groups in general. A key element of the circle of moral regard hypothesis is that a highly self-important moral identity should be associated with an expansion of the psychological boundaries that differentiate ingroups and out-groups as a direct result of a person's conception of a moral self. Although the results of Study 2 are supportive of this argument, they do not provide a direct test because Study 2 involves two out-groups and not an in-group versus an out-group. Given the similarities between the two out-group stimuli (both Middle Eastern out-groups) one may argue that the circle of moral regard hypothesis was not fully tested in Study 2. A stronger test is to see how people respond when given a choice between helping in-group members directly affected by intergroup conflict or members of an out-group associated with that conflict. Finally, if the
4 We theorized that as a person's moral identity increases in selfimportance, so too does the expansiveness of his or her circle of moral regard. This should be reflected in more favorable reactions to those that are even loosely associated with an out-group prototype linked to a salient intergroup conflict. This conceptual argument is not negated by the fact that many Americans and participants in our sample view Afghani women and children as being victims rather than perpetrators. Indeed, this is a key part of the theoretical premise because we hypothesize that responsiveness to the victimization of this out-group will increase as a person's moral † p identity assumes greater self-importance.
Table 2 Study 2: Hierarchical Regression of Moral Identity and Target on Worthiness of Relief Effort
Unstandardized coefficients Model and independent variable Step 1 Female White Afghan relief effort (A) Internalization (I) Symbolization (S) Step 2 AI AS Note. N 75. F(7, 67) .08*, adjusted R2 .16. .10. * p .05. B 0.133 0.629 0.581 0.503 0.139 1.354 0.423 2.95. For Step 1, R2 SE 0.349 0.299 0.314 0.266 0.225 0.515 0.418 t 0.38 2.10* 1.85† 1.89* 0.62 2.62* 1.01
.16*. For Step 2, R2
1278
REED AND AQUINO
circle of moral regard hypothesis provides a sufficiently powerful theoretical explanation for moral action, then the self-importance of moral identity should be not only associated with judgment and cognition but also associated with actual moral behavior, defined as showing social responsiveness to the needs of others (cf. Aquino & Reed, 2002, Study 6). Addressing these concerns was the purpose of Study 3.
Study 3: Providing Assistance to Out-Groups Versus In-Groups
Study 3 confronted participants with two compelling and deserving beneficiary groups: one more likely to be viewed as an ingroup, the other more likely to be viewed as an out-group. The two beneficiary groups were the New York Police and Fire Widows and Children's Benefit Fund (in-group) and the United Nations Children's Fund (UNICEF) Emergency Effort for Afghan Children and Families (out-group). The in-group represented by the New York widows and children is presumed to be emotionally closer to the participants than the distant and foreign Afghan women and children. As in Study 2, it was also assumed that Afghan women and children were more likely to be categorized as members of a highly salient out-group (i.e., Middle Easterners) loosely associated with a recent hostile event directed toward the in-group (i.e., fellow American citizens). What the beneficiary groups held in common, however, was that they comprised primarily people who were unwitting victims of intergroup conflict events beyond their control. To test the influence of the self-importance of moral identity and the expanded circle of moral regard on actual behavior, we required participants in Study 3 to allocate a fixed sum of money to one or both of the beneficiary groups. Following the theoretical arguments presented earlier, it was expected that the selfimportance of a person's moral identity would increase the willingness to allocate money to benefit the out-group. However, this study also examined the effects of a potentially competing "American" identity that might reinforce the notion of an American ingroup in the post-September 11th context. Current models of attitude suggest that people can simultaneously hold "dual attitudes" (Wilson, Lindsey, & Schooler, 2000) that have opposite evaluative implications toward the same attitude object (cf. Reed, Wooten, & Bolton, 2002). Hence, a person might have not only a highly self-important moral identity that motivates him or her to behave one way but also an equally self-important corporate, ethnic, or national identity may be invoked that motivates him or her to behave in another way. In the context of the September 11th attacks, an identity that is likely to conflict directly with a highly self-important moral identity as a predictor of the willingness to aid out- versus in-groups (as operationalized in Study 3) is the strength of a person's American identity. For example, after the September 11th attacks, many Americans demonstrated a surge in patriotism and an affirmation of their self- and national identities by hoisting American flags over homes and businesses, posting bumper stickers with statements exhorting national unity (e.g., United We Stand), or wearing t-shirts expressing the desire to retaliate against those held responsible for attacks (e.g., Osama bin Laden: Wanted Dead or Alive). As noted earlier, these processes of self- and group affirmation
more firmly establish the idea of "us," which, in turn, should constrain and rigidify the boundaries between the in-group (e.g., Americans) and out-group (e.g., Muslims, Middle-Easterners). Therefore, although a highly self-important moral identity is hypothesized to increase the likelihood that people would donate money to the effort to Aid Afghan Children and Families, American identity was expected to reinforce the in-group mentality and to increase the likelihood of donating to the American New York Police and Fire Widows and Children's Benefit Fund.
Method Sample and Procedures
Participants were students taking two separate organizational behavior courses at an urban university in the southern United States. Fifty-eight were undergraduate students and 38 were graduate students taking a master's of business administration course. Fifty-two percent of the students were women. Fifty-five percent of the students were White, 25% were African American, 9% were Asian, 6% were Hispanic, 1% was Native American, and 4% identified their ethnicity as other. The average age was 26.2 years (SD 5.9). Data were collected in three phases 5 weeks apart. The first two phases assessed the independent variables described later. The materials were distributed in class, and students were told it was voluntary to complete them but that the results would be discussed later in the semester. The third phase of the study was the behavioral component. It was conducted during a regular classroom session using the following procedures. Students were given $3 to distribute between the two charities 12 weeks after September 11, 2001. They were then given a 9 12 inch envelope with the following instructions printed on it: This is a study of donation behavior. Three single dollar bills and two white envelopes are enclosed in this large manila envelope. Each of the two envelopes represents one charity, and the charities are described on an enclosed piece of paper. Distribute the three dollars between the two charities by placing the amount you choose in the corresponding envelope. You may not keep the cash. Once you have made your choice and placed the cash in the envelope(s), seal both white envelopes, place them in the large envelope, and seal the large envelope. A one-page description of each of the two charities was included in the large envelope. Descriptions of both charities were drawn from their respective websites. The UNICEF page described the plight of children and families in war-torn Afghanistan and described the type of humanitarian aid (e.g., food and tents and fuel to cope with the impending winter) that would be delivered to them by UNICEF. The graphic included in the description was from the Website, and it depicted a mother comforting her child, both enveloped in a single wrap. The other charity included the following description: "The New York Police and Fire Widows and Children's Benefit Fund provides funds to families at the time of a death to assist with immediate expenses. In addition, the benefit fund distributes an annual check to each of the surviving families." As in the Website, the shields of New York police and firefighters were graphic images shown at the top of the description. Three U.S. dollar bills were physically given to participants to make the cash contributions more tangible and to force a nonequal contribution to the two choices. The participants were told that the money was "really going to the designated charities." (The money was subsequently pooled and actually donated to the charities.) Students were asked not to discuss their decision until after the task was completed and the envelopes were sealed. No other oral instructions were given.
MORAL IDENTITY AND OUT-GROUPS Of the 96 students who completed the self-importance of moral identity questionnaire at Time 1, 76 completed all subsequent phases of the study and provided usable data on all study variables, described in more detail later.
1279
Measures
Self-importance of moral identity. Aquino and Reed's (2002) instrument measured the self-importance of this construct. Respondents answered the items on a 5-point Likert scale rated from 1 (strongly disagree) to 5 (strongly agree). As in Studies 1 and 2, the internalization ( .90) and symbolization ( .66) items were averaged to form two subscales. American identity. Four items interspersed within a general political and environmental opinion questionnaire measured this construct. The items were as follow: "It's great to be an American," "I am extremely proud of my affiliation with the United States of America," "Being a member of this country makes me feel like I share a common goal with others," and "Being an American is not an important part of who I am" (reverse scored). These items were answered on a 5-point Likert scale rated from 1 (strongly disagree) to 5 (strongly agree) and were averaged to form a scale ( .83). Control variables. As in previous studies, gender and ethnicity were used as control variables. Both variables were dummy coded (0 male, 1 female and 0 non-White, 1 White). As the sample consisted of undergraduate and graduate students, it was desirable to control for age because it is possible that the formation of self-importance of moral identity may be influenced by socialization and life experience (Hart et al., 1998). Finally, because the experiment pits dual identities against each other, it is important to consider how people may cognitively resolve the "struggle within." In the domain of moral psychology, one of the most influential constructs for explaining how people resolve moral dilemmas like the one posed in Study 3 is moral reasoning (e.g., Kohlberg, 1971; Rest, 1979). An assumption of the cognitive developmental approach to moral decision making is that the sophistication of one's moral reasoning is ultimately linked to moral behavior (Kohlberg, 1971). To see whether the motivation to reason through moral problems may influence people's resource allocation decisions to donate money to either an in- or out-group, independent of the self-importance of moral identity, the sophistication of moral reasoning was controlled for by our administering a short (three dilemma) version of Rest's (1979) Defining Issues Test (Center for the Study of Ethical Development, 1993). The P-score ( .62) was used to measure the extent to which a person's level of moral reasoning approaches the principled level that cognitive developmental theorists consider to be the highest level of moral reasoning. The moral reasoning measure was collected at Phase 2 of the study. All other nonbehavioral variables were collected during Phase 1.
Donation. Two dependent variables reflected donation behavior: the absolute amount of money donated to the out-group (Afghan women and children) and whether the person allocated a larger share of the $3 to the out-group (i.e., $2 or more). This second variable was treated as binary (0 smaller share to out-group, 1 larger share to out-group) and indicates the degree to which participants were willing to assist the outgroup at cost to the in-group.
Results
The descriptive statistics and correlations among the study variables are shown in Table 3. Multiple regression was performed to test the effects of the two identities (the self-importance of moral identity and the strength of American identity) on the absolute amount of money donated to the out-group. The results of this analysis are shown in Table 4. The results supported the prediction that the self-importance of moral identity predicts the donation of money to the out-group as evidenced by the positive relation between internalization and the dependent variable ( .60, p .01). The results also supported the prediction that the strength of American identity is negatively related to this behavior ( .44, p .01). The data suggest that the self-importance of moral identity and the strength of American identity exerted a strong effect on behavior but in opposite directions. This result is consistent with the earlier argument that these two identities are associated with different views of the self's relationship to others, with a highly self-important moral identity extending the breadth of this relation (to the out-group), and with a strong American identity restricting it (to the in-group). No other effects were significant. The results of a binary logistic regression on the willingness to favor the out-group are also shown in Table 4. The results of this analysis replicate those obtained from the multiple regression, but what is interesting here is the extent to which the independent variables predicted people's decision to favor the out-group at the expense of the in-group. As seen in Table 4, people who scored high as compared to low on the internalization dimension of the self-importance of moral identity were far more likely to allocate a larger proportion of the $3 to the out-group (Afghani women and children). In contrast, people who scored high on the measure of American identity were less likely to do so. More specifically, for every unit increase (decrease) in internalization (strength of Amer-
Table 3 Study 3: Means, Standard Deviations, and Correlations Among Study Variables
Study variable 1. 2. 3. 4. 5. 6. 7. UNICEF Age White Female Moral reasoning Internalization Symbolization M 1.16 26.67 0.55 0.50 31.01 4.48 3.23 SD 1.01 6.20 0.50 0.50 16.80 0.70 0.63 1 — .05 .18 .11 .02 .32* .14 2 — .19 .17 .06 .24* .004 3 4 5 6 7 8
— .16 .32* .12 .13
— .07 .36* .43***
— .19 .02
— .38*
—8.
American identity
Note. UNICEF United Nations Children's Fund. * p .05. *** p .001.
1280
REED AND AQUINO
Table 4 Study 3: Results of Multiple Regression Analysis and Logistic Regression on Allocation of Monetary Donations to UNICEF
Dependent variable Absolute $ to UNICEFa Independent variable Age White Female Moral reasoning Internalization Symbolization American identity B 0.02 0.28 0.12 0.00 0.60 0.13 0.44 SE 0.02 0.23 0.24 0.01 0.17 0.20 0.11 t 1.25 1.23 0.49 0.29 3.49*** 0.67 3.92*** B 0.03 0.73 0.23 0.01 1.38 0.14 0.66 Preference for UNICEFb SE 0.05 0.58 0.61 0.02 0.66 0.52 0.31 .17; Nagelkerke R2 Wald 0.46 1.61 0.14 0.25 4.36* 0.07 4.49* .24.
2
Note. UNICEF United Nations Children's Fund. a Adjusted R2 .22; F(7, 68) 3.95*. b Model 2 14.22*; Cox and Snell's R2 * p .05. *** p .001.
ican identity), the odds of giving a greater portion of money to the out-group increased (decreased) by 298% (48%). Again, these results are consistent with the argument that the self-importance of moral identity can overcome the tendency to favor the in-group, even during times of intergroup conflict. Indeed, they show that high moral identifiers were actually more likely to favor the outgroup, at least with regard to providing direct monetary assis- tance in this experimental context.
Study 4: Moral Identity and Revenge and Forgiveness Toward Out-Groups
As the United States waged its military campaign in Afghanistan to capture or kill those responsible for masterminding the terrorist attacks, one of the issues often debated in the public arena was how many innocent civilian casualties (presumably Afghani) the American people should be willing to accept. There have been varying public reactions. For example, not all Americans were equally disturbed by incidents like the one in which many innocent Afghani wedding attendees were killed when U.S. military planes construed celebratory rifle fire as an antiaircraft attack (Graham, 2002). Following the theoretical arguments in this article, it is proposed that a self-important moral identity can partly explain differences in the willingness to accept such civilian casualties. Specifically, because people whose moral identity has high selfimportance are more likely to view a wider set of out-groups as deserving moral concern, high moral identifiers should be less willing to accept the possibility of large numbers of civilian casualties following a military counterresponse to the terrorist attacks. In contrast, because low moral identifiers are likely to restrict their circle of moral regard, they may be more willing to accept considerable numbers of civilian deaths that are closely associated with a hostile out-group. On the basis of this reasoning, we hypothesized that a highly self-important moral identity would be negatively related to the willingness to accept civilian deaths as a direct result of a military attack against Afghanistan. The above hypothesis deals with people's reactions toward outgroup members who may be viewed as necessary, albeit undeserving, victims of military retaliation. This begs the question of whether the self-importance of moral identity might influence people's reactions toward out-group members held directly responsible for the September 11th attacks. If the effect of a highly self-important moral identity on the circle of moral regard is sufficiently robust and powerful, it is plausible that even out-group members who are viewed as having caused harm may be considered deserving of some moral concern. Examining the relationship between the self-importance of moral identity and two key retal-
Discussion and Limitations
Studies 2 and 3 focused on how the self-importance of moral identity relates to reactions to out-groups in the aftermath of intergroup conflict. These positive attitudes, intentions, and behaviors toward the out-group appear to be a consequence of an expanded circle of moral regard that is associated with having a highly self-important moral identity. However, it is clear that one of the strongest reactions to the attacks on the World Trade Center and Pentagon is to retaliate and punish the perpetrators and those who assist or harbor them. This was ostensibly the reason for the United States government sending troops into Afghanistan to hunt down the terrorists and for the continued United States campaign against Al-Qaeda and its leaders to this day. In the wake of these retaliatory responses toward out-groups, a question that was not addressed by the three previous studies is whether the circle of moral regard hypothesis applies not only to positive responses (i.e., perceived worthiness of assistance efforts, providing actual aid) but also to negative responses toward out-groups. Logically, an expanded circle of moral regard should be negatively associated with the willingness to inflict harm and injury on out-groups, even when these out-groups may be perceived as responsible for harming oneself or one's in-group. Study 4 tested this hypothesis by examining whether the self-importance of moral identity predicts (a) the willingness to accept civilian deaths among out-group members in retaliation for the attack against the United States and (b) the perceived morality of two hypothetical response options that vary in their level of punitiveness, which might be taken by the United States against the perpetrators.
MORAL IDENTITY AND OUT-GROUPS
1281
iatory reactions toward the September 11th terrorist attackers tests this hypothesis.
Revenge and Forgiveness Toward Harm Doers
A natural response to having been harmed by another person or group is to seek revenge against the harm doer. Revenge has been justified on moral grounds because it balances the calculus of justice by restoring equity in the relationship between victim and offender (e.g., Hogan & Emler, 1981; Walster, Walster, & Berscheid, 1973). In the case of the September 11th attacks, thousands of lives were lost, the economic damages to the American economy were substantial, and the American "way of life" was perceived as having been threatened by a hostile out-group. Consequently, massive military retaliation directed against the perpetrators of the attack was seen by many Americans as a practical and moral imperative. Yet again, there were some Americans who believed that less punitive means should be pursued in response to the terrorist attacks. Study 4 focused on two possible responses to harm doing— revenge and forgiveness—that lie at extreme ends along a continuum of punitiveness. Revenge has been defined as an action taken in response to a perceived harm or wrongdoing by another party. Revenge is intended to inflict harm, damage, injury, discomfort, or punishment on the party judged responsible for causing the harm (Aquino, Tripp, & Bies, 2001; Stuckless & Goranson, 1992). Forgiveness has been defined as a deliberate decision by the victim of a harmful action to relinquish anger, resentment, and the desire to punish the party held responsible for inflicting harm (Aquino et al., 2001; Enright & The Human Development Study Group, 1991; North, 1987). By forgiving the offender, the victim essentially cancels the offender's "debt" and surrenders the "right" to exact punishment or seek revenge (Enright & the Human Development Study Group, 1991). This surrendering of justice by restoring equity is what forgiveness demands, although this does not mean that the offender is not held responsible for what he or she did or does not suffer some legal consequence (e.g., imprisonment; Enright & The Human Development Study Group, 1991). What forgiveness does mean is that the victim overcomes anger, hatred, and the desire to personally injure the harm doer. It is for this reason that forgiveness is typically viewed as the antithesis of revenge. It is argued that because these two responses to wrongdoing represent opposite poles along a dimension of punitiveness, examining people's beliefs about the moral appropriateness of these two actions provides a further test of the circle of moral regard hypothesis. If the self-importance of moral identity expands this circle, and if we assume that a defining characteristic of showing moral regard is the desire to minimize the suffering and harm of another human being, then it is reasonable to propose that high as compared with low moral identifiers should view a less punitive response to the perpetrators of the terrorist attacks on the World Trade Center and Pentagon as more morally acceptable. The reason for this is that the less punitive the response, the less the suffering of persons included within the circle of moral regard. Consistent with this argument, Westman and Lewandowski (1992) found that a stronger belief in the unity of humanity was negatively related to approval of war as a strategy.
Following this logic, Study 4 presented participants with two response options for dealing with the perpetrators of the World Trade Center and Pentagon attacks to see whether the selfimportance of moral identity influences judgments of the morality of (a) using any means necessary to kill those responsible or (b) forgiving those responsible by replacing negative emotions like hatred and anger with positive emotions like compassion and love. It was predicted that although a highly self-important moral identity would be negatively related to the perceived morality of Option a, it would be positively related to the perceived morality of Option b. The reason being that people whose moral identity has high self-importance and whose circle of moral regard is thereby more expansive, would be least likely to endorse the option which inflicts the most harm (death) on others. Conversely, high moral identifiers would be considerably more likely than low moral identifiers to endorse the morality of forgiving the perpetrators (Option b) based on the assumption that even the perpetrators are more likely to be included within the circle of moral regard.
Method Participants
One hundred six undergraduate students from a university located in the northeastern United States participated in the exercise for class credit. Average age was 20.6 years (SD 1.40). Forty-eight were men, 68 were women. Of those who reported race, 55 were White, 8 were African American, 29 were Asian, 3 were Hispanic, and 3 self-identified as other. Eighty-five participants provided usable data on all study variables.
Procedure
The exercise consisted of two time periods separated by 2 months. At Time 1, participants completed a consumer opinions questionnaire. At Time 2, participants completed a computer exercise intended to measure reactions toward historical events. They were told that the purpose of the Time 2 study was to gain a comprehensive understanding of how people respond to and remember major events. Pictures were taken from various online media regarding the World Trade Center attacks. Some of these pictures were pictures of devastation (e.g., the burning towers), others were pictures of sympathy (e.g., the outpouring of international support from other countries). The pictures were shown on a computer screen in a random order. Participants were then asked questions described later. The confidentiality of students' responses was strictly maintained by assigning code numbers at both time periods. After data collection was completed, participants were fully debriefed and dismissed.
Independent Variables
As in the previous studies, the 10-item Self-Importance of Moral Identity measure was administered at Time 1 along with several other measures. Using a 7-point scale ranging from 1 (low) to 7 (high), we calculated average scores on both the Internalization dimension (M 5.31, SD 0.53, .85) and the Symbolization dimension (M 2.96, SD 0.71, .78), and entered them into the analysis as an index of activation potential of a moral self.
Control Variables
Gender (0 male, 1 female), race (0 non-White, 1 White) and American identity, M 3.42 (SD 1.07); ( .91), were used as control variables.
1282 Dependent Variables
REED AND AQUINO
Participants responded to measures after viewing the pictures from September 11th. Participants were told to assume that a military strike against the persons held responsible was imminent and would result in civilian deaths among persons who had nothing to do with the attack. They were then asked to type in the number of (innocent) civilian deaths that would be morally acceptable in such a retaliatory attack. The text box for this response variable had exactly nine digits, so participants could have inputted a theoretical range of from 0 to 999999999. Second, using a 5-point Likert scale (1 extremely immoral, 5 extremely moral), participants assessed the moral appropriateness of the two responses described above: (a) use any means necessary to kill those responsible for these acts, (b) forgive those responsible meaning negative emotions like hatred and anger should be replaced with positive emotions like compassion and love.
means necessary) and forgiveness (showing compassion) as repeated measures and submitted them to a multivariate regression analysis that included gender, race, internalization, symbolization, and American identity as predictor variables. Gender and race had no effect and, because of the small size of the sample for Study 4, were deleted from the model to preserve degrees of freedom. Results showed a main effect of response, F(1, 69) 18.69, p .01, such that participants were more likely to perceive killing the perpetrators of the terrorist attacks as more moral (M 2.73) than forgiveness and compassion (M 2.42). However, this main effect was qualified by two significant interactions. There was a significant Response (revenge vs. forgiveness) Internalization interaction, F(1, 69) 3.70, p .05, and a significant Response Symbolization interaction, F(1, 69) 6.42, p .05. No other effects were significant. To examine the nature of these withinsubject effects, the zero-order correlations between internalization, symbolization, revenge (killing by any means necessary), and forgiveness (showing compassion) were examined. Consistent with predictions, internalization was negatively correlated with perceptions of morality of killing the perpetrators (r .37, p .01) and positively correlated with perceptions of morality of forgiving the perpetrators (r .35, p .01). Similarly, symbolization was negatively correlated with perceptions of the morality of seeking revenge by killing the perpetrators (r .30, p .01), although it showed no relationship to the perceived morality of forgiving the terrorist attackers ( r .15, ns).
Results
Correlations among all independent and dependent variables are shown in Table 5.
Number of Acceptable Civilian Deaths
A multiple regression analysis was performed to predict the number of acceptable civilian deaths as a function of the selfimportance of moral identity. The model included gender, race, internalization, symbolization, and American identity as independent and control variables. Results showed that the internalization dimension, F(1, 73) 3.87, p .05, 37, p .05, and the symbolization dimension, F(1, 73) 6.12, p .05, 25, p .05, of moral identity were both associated with reported acceptable collateral damage. Those who had a relatively stronger self-importance of moral identity as indexed by the two dimensions reported a lower number of acceptable civilian casualties. No other effects were significant.
General Discussion
Social psychologists have studied several approaches in the hope of reducing intergroup hostility and bias (cf. Hewstone et al., 2002). A review of these approaches provides insight into the cognitive and motivational processes that might interfere with what seems to be humanly universal, that is, the often automatic (cf. Greenwald, McGhee, & Schwartz, 1998; Wilson et al., 2000) tendency for people to favor in-groups and to negatively respond to out-groups. One effective approach to reducing negative responses to out-groups is to encourage people to bring their prejudices to mind and to then invoke self-directed guilt derived from the discrepancy between personal values and actual behavior (Devine, Plant, & Blair, 2001). Another approach seeks to suppress out-group hostility by increasing the salience of positive values (e.g., tolerance) and making people rationalize or account for their
Moral Appropriateness of Revenge and Forgiveness
It was argued that because less punitive responses are more likely to minimize the suffering of those persons who are included within the circle of moral regard, then high as compared with low moral identifiers would view a less punitive response to the perpetrators of the terrorist attacks on the World Trade Center and Pentagon as more morally acceptable. To test this hypothesis, we treated the two hypothetical responses of revenge (killing by any
Table 5 Study 4: Means, Standard Deviations, and Correlations Among Independent and Dependent Variables
Variable 1. 2. 3. 4. 5. 6. 7. 8. †p Gender Race Internalization Symbolization American identity Acceptable civilian deaths Kill Forgiveness .10. *p .05. *** p M SD 1 — .04 .10 .05 .04 .10 .05 .01 2 — .06 .12 .04 .10 .05 .00 3 4 5 6 7 8
5.31 2.96 3.42 1,826 2.73 2.42 .001.
0.53 0.71 1.07 11,073 1.33 1.41
— .45*** .08 .34*** .37*** .35***
— .08 .18† .30* .15
— .06 .14 .23†
— .18† .18†
— .63***
—
MORAL IDENTITY AND OUT-GROUPS
1283
negative reactions (Dobbs & Crano, 2001). Researchers have also examined approaches whereby out-group hostility is reduced by using dissonance principles. These approaches assume that promoting nonbiased behavior (Leippe & Eisenstadt, 1994) and increasing the quality and quantity of intergroup contact (Pettigrew, 1998) will produce attitudinal realignment as an outcome of dissonance reduction. What the intervention strategies described earlier share is that they are primarily cognitive and consequentialist, meaning their effectiveness depends on adequate levels of awareness, effort, and practice over time (Hewstone et al., 2002). Moreover, many of the aforementioned interventions are inhibitory in the sense that they attempt to suppress implicit and existing out-group hostility by getting people to undo or relearn them in some way. A complementary explanation for why some people may be less susceptible to in-group favoritism and out-group hostility, however, is that their moral identity has high self-importance. As a result, they mentally redefine the boundaries of their in-group. The purpose of this article was to test this theoretical premise referred to as the circle of moral regard hypothesis: The self-importance of a person's moral identity, conceptualized as the degree to which one's self-conception is privately and publicly organized around moral trait associations (Aquino & Reed, 2002), directly influences the self- other relation. Several psychological outcomes associated with an expanded circle of moral regard were investigated, and the findings of four studies provide preliminary but convergent evidence supporting the theorized link between a highly selfimportant moral identity and the expansion of the circle of moral regard. Evidence from prior research suggests that out-group negativity is often reduced by fostering perspective taking (cf. Levy, Freitas, & Salovey, 2002) or improving attitudes toward former out-group members by "recategorizing" them from out- to in-group (Hewstone et al., 2002). It appears that the self-importance of moral identity might represent a self-regulatory mechanism that affects how a person cognitively constructs the psychological boundaries that define "us" and "them." Data from Study 1 suggest that the extent to which a person possesses a strong internalized moral identity will be reflected in stronger self-reported moral obligation to help various conceptions of out-groups and a desire to exchange highly personalized resources with them. This self-reported moral regard extended to out-groups was also reflected in secondary judgments regarding relief efforts. Data from Study 2 showed that individuals who had a strong internalized moral identity were more likely to favorably perceive the worthiness of relief efforts to assist out-groups, even when those out-groups were loosely associated with a highly salient intergroup conflict. This is an especially interesting finding because prior research has shown that during times of intergroup conflict, broad-based stereotypes of out-group members often trigger hostile expectations of interactions wrought with conflict (e.g., Stephan et al., 1998). Yet, even in the aftermath of one of the most disturbing domestic intergroup conflicts in 60 years, such negative reactions did not unfavorably affect judgments toward out-group relief efforts when the internalized selfimportance of moral identification was high. A more rigorous test of the circle of moral regard hypothesis was examined in Study 3. Results suggested that a strong internalized moral identity was associated with actual donation behav-
ior that reflected preference for the out-group. When forced to choose between giving more monetary assistance to either a deserving in-group (New York Police and Fire Widows and Children's Benefit Fund) or a deserving out-group (Afghani women and children), participants with a high internalized moral identity chose to extend assistance to the more socially distant out-group. These results provide the most direct evidence for the circle of moral regard hypothesis because they involve real behavior that emerged even when controlling for a powerful internal force (i.e., the opposing American identity) that seems likely to motivate the contraction rather than the expansion of the circle of moral regard.5 Finally, it appears that the psychological outcomes of an expansive circle of moral regard apply not only to positive actions toward the out-group but also to negative reactions toward out-groups that are loosely associated with a salient intergroup conflict and even to those held directly responsible for causing harm to the in-group. Study 4 showed that people who had both a high internalized moral identity and a high symbolized moral identity were less likely to accept harming innocent out-group members as a result of military retaliation (cf. Westman & Lewandowski, 1992). Ironically, those who have a particularly strong moral identity can be more sensitive to the aggressor's intentions in their perceptions of hostility (cf. Berkowitz, Mueller, Schnell, & Padberg, 1986), and yet they may also be more inclined to temper their desire for retaliation in the wake of hostility. This interpretation is based on the results from Study 4 showing that low as compared with high moral identifiers were less likely to judge killing the perpetrators of intergroup conflict as an appropriate moral response and more likely to judge forgiveness as morally appropriate. Although the data from the four studies reported here provide preliminary evidence for the link between the self-importance of moral identity and the psychological outcomes associated with an expanded circle of moral regard, several limitations of the studies should be noted. With the exception of two findings, the internalization dimension of moral identity was a stronger predictor of the criterion variables. One possible reason for this result is that even though the people in the samples presented here had relatively strong internalized senses of their moral identity, there may be other factors that impede their motivation, ability, or opportunity to express that moral identity publicly. Alternatively, it could be that people symbolically express their moral identities in ways other than what was measured with the items used to tap the
5 It should be noted that these results do not imply that if a person has a strong American identity, then they are immoral. Two points should be made: First, an examination of the correlation between American identity, internalization, and symbolization shows that they were slightly positively correlated in Study 3 (see Table 3). Second, Aquino and Reed (2002) have already made the point that individuals who score relatively lower on the 10-item measure of the self-importance of moral identity are not necessarily immoral individuals. For example, examination of the means for internalization and symbolization across all of the studies reported here suggest that people self-report relatively high levels (above the midpoint of the scale) on the two subscales. Therefore, a more appropriate interpretation at least in these data and in the data from Aquino and Reed is that people who score relatively higher on the subscales simply have a self-concept that tends to be relatively more morally schematic (cf. Bem, 1981).
1284
REED AND AQUINO moral actions that differ in self investment. Manuscript submitted for publication. Aquino, K., Tripp, T. M., & Bies, R. J. (2001). How employees respond to personal offense: The effects of blame attribution, victim status, and offender status on revenge and reconciliation in the workplace. Journal of Applied Psychology, 86, 52-59. Bem, S. L. (1981). Gender schema theory: A cognitive account of sex typing. Psychological Review, 88, 354 -364. Berkowitz, M. W., Mueller, C. W., Schnell, S. V., & Padberg, M. T. (1986). Moral reasoning and judgments of aggression. Journal of Personality and Social Psychology, 51, 885- 891. Blasi, A. (1984). Moral identity: Its role in moral functioning. In W. Kurtines & J. Gewirtz, (Eds.), Morality, moral behavior and moral development (pp. 128 -139). New York: Wiley. Blieszner, R. (1993). Resource exchange in the social networks of elderly women. In U. G. Foa, J. Converse Jr., K. Turnblom, & E. B. Foa (Eds.), Resource theory: Explorations and applications (pp. 67-79). San Diego, CA: Academic Press. Brewer, M. B., & Gardner, W. (1996). Who is this "We"? Levels of collective identity and self representations. Journal of Personality and Social Psychology, 71, 83-93. Brewer, M. B., & Kramer, R. M. (1985). The psychology of intergroup attitudes and behavior. Annual Review of Psychology, 36, 219 -243. Brown, J. D., & Gallagher, F. M. (1992). Coming to terms with failure: Private self-enhancement and public self-effacement. Journal of Experimental Social Psychology, 28, 3-22. Center for the Study of Ethical Development. (1993). Guide for the Defining Issues Test. Minneapolis: University of Minnesota. Damon, W. (1984). Self-understanding and moral development from childhood to adolescence. In W. Kurtines & J. Gewirtz (Eds.), Morality, moral behavior and moral development (pp. 109 -127). New York: Wiley. Dart, B. (2001, December 5). Kindness honored in wake of attacks: Groups helped Muslims, Arabs. The Atlanta Journal-Constitution. Retrieved December 5, 2001, from http://www.accessatlanta.com/ajc/ Deaux, K. (2002) Negotiating identity and community after September 11. Retrieved March 5, 2003, from the Graduate Center at the City University of New York: Social Science Research Council: http://www .ssrc.org/sept11/essays/deaux.htm Devine, P. G. (1989). Stereotypes and prejudice: Their controlled and automatic components. Journal of Personality and Social Psychology, 56, 5-18. Devine, P. G., Plant, E. A., & Blair, I. V. (2001). Classic and contemporary analyses of racial prejudice. In R. Brown & S. Gaertner (Eds.), Blackwell handbook of social psychology: Intergroup processes (pp. 198 -217). Malden, MA: Blackwell. Dobbs, M., & Crano, W. D. (2001). Outgroup accountability in the minimal group paradigm: Implications for aversive discrimination and social identity theory. Personality and Social Psychology Bulletin, 12, 355365. Dovidio, J. F., & Gaertner, S. L. (1991). Changes in the expression and assessment of racial prejudice. In H. J. Knopke, R. J. Norrell, & R. W. Rogers (Eds.), Opening doors: Perspectives on race relations in contemporary America (pp. 119 -148). Tuscaloosa: University of Alabama Press. Dovidio, J. F., & Gaertner, S. L. (1993). Stereotypes and evaluative intergroup bias. In D. M. Mackie & D. L. Hamilton (Eds.), Affect, cognition, and stereotyping (pp. 167-193). San Diego, CA: Academic Press. Eisenberg, N. (2000). Emotion, regulation, and moral development. Annual Review of Psychology, 51, 665- 697. Enright, R. D., & The Human Development Study Group. (1991). The moral development of forgiveness. In W. Kurtines & J. Gewritz (Eds.),
symbolization dimension. Future research should examine conditions under whether the private (internalization) or public (symbolization) aspects of the self-importance of a person's moral identity will be stronger predictors of particular outcomes. One plausible hypothesis is that the symbolization dimension might be more predictive of overt, public behaviors. In contrast, internalization might be less affected by the presence of an audience. Another limitation of the studies in this article is that the processes that directly link the self-importance of moral identity to psychological and cognitive functioning were not explored. It should be noted, though, that this study is preliminary, and it uses one of the first validated measures of the self-importance of moral identity. Moreover, the goal of this article was to broadly investigate the relationship between the self importance of moral identity and the psychological outcomes of an expanded circle of moral regard. Future work should examine the micropsychological processes that account for this relationship. Finally, none of the data in this article permits strong causal inferences. In all of the studies, the self-importance of moral identity was related to other outcomes through correlational rather than experimental analysis. Although many other variables were controlled, the possibility remains that one or more unmeasured variables may systematically covary with the dimensions of moral identity's self-importance and the outcomes measured in the four studies. To make stronger causal inferences, future research might attempt to manipulate the selfimportance of a person's moral identity and systematically test its relationship to moral outcomes.
Conclusions
Emily Dickinson once observed that human beings select their own societies and then "shut the door" to the claims and needs of those who fall outside the circle of their moral regard. The natural propensity to distinguish "us" from "them" has been at the forefront of work on intergroup conflict. It has long been recognized that "some [people] may consider their morality to be central to their self-identities, whereas others may consider it to be peripheral" (Damon, 1984, p. 110). However, very little empirical research has systematically examined the relationship between the self-importance of moral identity and its cognitive and behavioral implications for the construction of psychological group boundaries. Consequently, this article makes two important contributions to the literature. First, it refines and further extends a definition of moral identity grounded in theories of social identity and the selfconcept. Second, it provides evidence for the relationship between the self-importance of moral identity and various reac- tions toward out-groups. The data suggest that the concept of a moral self may collapse the mental barriers separating in-groups and out-groups, thus highlighting the possibility that in some "simple, perfectly impenetrable" way, we are all brothers.
References
Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and interpreting interactions. Thousand Oaks, CA: Sage. Aquino, K. F., & Reed, A., II. (2002). The self-importance of moral identity. Journal of Personality and Social Psychology, 83, 1423-1440. Aquino, K. F., & Reed, A., II. (2003). To thine own selves be true: The effects of moral identity and organizational status on the preference for
MORAL IDENTITY AND OUT-GROUPS Handbook of moral behavior and development (Vol. 1, pp. 123-152). Hillsdale, NJ: Erlbaum. Erikson, E. H. (1964). Insight and responsibility. New York: Norton. Esses, V. M., Haddock, G., & Zanna, M. P. (1993). Values, stereotypes, and emotions as determinants of intergroup attitudes. In D. M. Mackie & D. L. Hamilton (Eds.), Affect, cognition, and stereotyping: Interactive processes in group perception (pp. 167-193). San Diego, CA: Academic Press. Fein, S., & Spencer S. J. (1997). Prejudice as self-image maintenance: Affirming the self through derogating others. Journal of Personality and Social Psychology, 73, 31- 44. Feldman, J., & Lynch, J. G. (1988). Self-generated validity and other effects of measurement on belief, attitude, intention and behavior. Journal of Applied Psychology, 73, 421- 435. Foa, E. B., & Foa, U. G. (1974). Societal structures of the mind. Springfield, IL: Charles C Thomas. Foa, E. B., & Foa, U. G. (1980). Resource theory: Interpersonal behavior in exchange. In K. J. Gergen, M. S. Greenberg, & R. H. Willis (Eds.), Social exchange: Advances in theory and research (pp. 77-101). New York: Plenum Press. Forehand, M., Deshpande, R., & Reed, A., II. (2002). Identity salience and ´ the influence of differential activation of the social self-schema on advertising response. Journal of Applied Psychology, 87, 1086 -1099. Gaertner, S. L., & Insko, C. A. (2000). Intergroup discrimination in the minimal group paradigm: Categorization, reciprocation, or fear? Journal of Personality and Social Psychology, 79, 77-94. Gibbons, F. X., & Gerrard, M. (1991). Downward social comparison and coping with threat. In J. M. Suls & T. A. Wills (Eds.), Social comparison: Theory and research (pp. 317-345). Hillsdale, NJ: Erlbaum. Graham, B. (2002, July 9). Inquiry set in Afghans' deaths in air-strike: Pentagon still disputes number of casualties. Washington Post, p. A17. Greenwald, A. G., & Farnham, S. D. (2000). Using the Implicit Association Test to measure self-esteem and self-concept. Journal of Personality and Social Psychology, 79, 1022-1038. Greenwald, A. G., McGhee, D. E., & Schwartz, J. L. K. (1998). Measuring individual differences in implicit cognition: The Implicit Association Test. Journal of Personality and Social Psychology, 74, 1464 -1480. Hamilton, D. L., & Trolier, T. K. (1986). Stereotypes and stereotyping: An overview of the cognitive approach. In J. F. Dovidio & S. L. Gaertner (Eds.), Prejudice, discrimination, and racism (pp. 127-163). Orlando, FL: Academic Press. Hart, D., Atkins, R., & Ford, D. (1998). Urban America as a context for the development of moral identity in adolescence. Journal of Social Issues, 54, 513-530. Hart, D., & Fegley, S. (1995). Prosocial behavior and caring in adolescence: Relations to moral judgment and self-understanding. Child Development, 66, 1346 -1359. Hewstone, M., Rubin, M., & Willis, H. (2002). Intergroup bias. Annual Review of Psychology, 53, 575- 604. Hogan, R., & Emler, N. P. (1981). Retributive justice. In M. J. Lerner & S. C. Lerner (Eds.), The justice motive in social behavior (pp. 125-143). New York: Plenum. Hume, D. (1751/1996). An inquiry concerning the principles of morals. LaSalle, IL: Open Court. Kihlstrom, J. F., & Klein, S. B. (1994). The self as a knowledge structure. In R. S. Wyer Jr. & K. Thomas (Eds.), Handbook of social cognition: Vol. 1. Basic processes: Vol. 2. Applications (2nd ed.; pp. 153-208). Hillsdale, NJ: Erlbaum. Kohlberg, L. (1971). From is to ought: How to commit the naturalistic fallacy and get away with it in the study of moral development. In T.
1285
Mischel (Ed.), Cognitive development and epistemology (pp. 164 -165). New York: Academic Press. Lapsley, D. K., & Lasky, B. (2001). Prototypic moral character. Identity, 1, 345-363. Laughlin, H. P. (1970). The ego and its defenses. New York: AppletonCentury-Crofts. Leippe, M. R., & Eisenstadt, D. (1994). Generalization of dissonance reduction: Decreasing prejudice through induced compliance. Journal of Personality and Social Psychology, 67, 395- 413. Levy, S. R., Freitas, A. L., & Salovey, P. (2002). Construing action abstractly and blurring social distinctions: Implications for perceiving homogeneity among, but also empathizing with and helping, others. Journal of Personality and Social Psychology, 83, 1224-1238. Markus, H. (1977). Self-schemata and processing information about the self. Journal of Personality and Social Psychology, 35, 63-78. Mikulincer, M., & Shaver, P. R. (2001). Attachment theory and intergroup bias: Evidence that priming the secure base schema attenuates negative reactions to out-groups. Journal of Personality and Social Psychology, 81, 97-115. Monroe, K. R., & Epperson, C. (1994). "But what else could I do?" Choice, identity, and cognitive perceptual theory. Political Psychology, 15, 201226. North, J. (1987). Wrongdoing and forgiveness. Philosophy, 6, 499 -508. Oliner, S. P., & Oliner, P. M. (1988). The altruistic personality: Rescuers of the Jews in Nazi-occupied Poland. New York: Free Press. Pettigrew, T. F. (1998). Intergroup contact theory. Annual Review of Psychology, 49, 65- 85. Pratto, F., Sidanius, J., & Stallworth, L. M. (1993). Sexual selection and the sexual and ethnic basis of social hierarchy. In L. Ellis (Ed.), Social stratification and socioeconomic inequality: A comparative biosocial analysis (pp. 111-137). New York: Praeger. Quillian, L. (1995). Prejudice as a response to perceived group threat: Population composition and anti-immigrant and racial prejudice in Europe. American Sociological Review, 60, 586 - 611. Reed, A., II. (2002). Social identity as a useful perspective for self-conceptbased consumer research. Psychology and Marketing, 19, 235-266. Reed, A., II, Aquino, K. F., & Ray, S. (2003). Violent media imagery and the September 11th terrorist attacks: Moral identity moderates the relationship between empathy and moral disengagement. Manuscript submitted for publication. Reed, A., II, Wooten, D. B., & Bolton L. (2002). The temporary construction of consumer attitudes. Journal of Consumer Psychology, 12, 375388. Rest, J. (1979). Development in judging moral issues. Minneapolis: University of Minnesota Press. Schlenker, B. R. (Ed.). (1985). The self in social life. New York: McGrawHill. Sidanius, J., Levin, S., Liu, J., & Pratto, F. (2000). Social dominance orientation, anti-egalitarianism and the political psychology of gender: An extension and cross-cultural replication. European Journal of Social Psychology, 30, 41- 67. Singer, P. (1981). The expanding circle: Ethics and sociobiology. New York: Farrar, Straus, & Giroux. Steele, C. M. (1988). The psychology of self-affirmation: Sustaining the integrity of the self. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 21, pp. 261-302). New York: Academic Press. Steele, C. M., & Liu, T. J. (1983). Dissonance processes as selfaffirmation. Journal of Personality and Social Psychology, 45, 5-19. Steele, C. M., Spencer, S. J., & Lynch, M. (1993). Self-image resilience and dissonance: The role of affirmation resources. Journal of Personality and Social Psychology, 64, 885- 896. Stephan, W. G., & Stephan, C. (1993). Cognition and affect in stereo-
1286
REED AND AQUINO Turner, J. L., Foa, E. B., & Foa, U. G. (1971). Interpersonal reinforcers: Classification, interrelationships, and some differential properties. Journal of Personality and Social Psychology, 19, 168 -180. Turner, J. C., Hogg, M. A., Oakes, P. J., Reicher, S. D., & Wetherell, M. S. (1987). Rediscovering the social group: A self-categorization theory. Oxford, England: Blackwell. Walker, L. J., & Pitts, R. C. (1998). Naturalistic conceptions of moral maturity. Developmental Psychology, 34, 403- 419. Walster, E., Walster, G., & Berscheid, E. (1978). Equity: Theory and research. Boston: Allyn & Bacon. Westman, A. S., & Lewandowski, L. M. (1992). How empathy, egocentrism, Kohlberg's moral development, and Erikson's psychosocial development are related to attitudes toward war. Psychological Reports, 69(Pt. 2), 1123-1127. Wilson, T. D., Lindsey, S., & Schooler, T. T. (2000). A model of dual attitudes. Psychological Review, 107, 101-126. Younis, J., & Yates, M. (1999). Youth service and moral-civic identity: A case of everyday morality. Educational Psychology Review, 11, 361376.
typing: Parallel interactive networks. In D. M. Mackie & D. L. Hamilton (Eds.), Affect, cognition, and stereotyping: Interactive processes in group perception (pp. 111-136). Orlando, FL: Academic Press. Stephan, W. G., Ybarra, O., Martinez, C. M., Schwarzwald, J., & TurKaspa, M. (1998). Prejudice toward immigrants to Spain and Israel: An integrated threat theory analysis. Journal of Cross-Cultural Psychology, 29, 559 -576. Stryker, S. (1980). Symbolic interactionism: A social structural version. Menlo Park, CA: Benjamin/Cummings. Stuckless, N., & Goranson, R. (1992). The Vengeance Scale: Development of a measure of attitudes toward revenge. Journal of Social Behavior and Personality, 7, 25- 42. Tajfel, H. (1959). Quantitative judgment in social perception. British Journal of Psychology, 50, 16 -29. Tajfel, H., & Turner, J. C. (1979). An integrative theory of intergroup conflict. In W. G. Austin & S. Worchel (Eds.), The social psychology of intergroup relations (pp. 33- 47). Monterey, CA: Brooks-Cole. Tetlock, P. E. (1983). Accountability and the perseverance of first impressions. Social Psychology Quarterly, 46, 285-292.
Appendix The Self-Importance of Moral Identity Measure
Listed below are some characteristics that might describe a person: Caring, Compassionate, Fair, Friendly, Generous, Helpful, Hardworking, Honest, and Kind The person with these characteristics could be you or it could be someone else. For a moment, visualize in your mind the kind of person who has these characteristics. Imagine how that person would think, feel, and act. When you have a clear image of what this person would be like, answer the following questions. I I S I S S I S S I 1. It would make me feel good to be a person who has these characteristics. 2. Being someone who has these characteristics is an important part of who I am. 3. I often wear clothes that identify me as having these characteristics. 4. I would be ashamed to be a person who had these characteristics. (R) 5. The types of things I do in my spare time (e.g., hobbies) clearly identify me as having these characteristics. 6. The kinds of books and magazines that I read identify me as having these characteristics. 7. Having these characteristics is not really important to me. (R) 8. The fact that I have these characteristics is communicated to others by my membership in certain organizations. 9. I am actively involved in activities that communicate to others that I have these characteristics. 10. I strongly desire to have these characteristics. I Internalization; S Symbolization; R Reverse coded.
doc_640697534.docx
Case Study on Moral Identity and the Expanding Circle of Moral Regard toward Out-Groups
An economic expansion is an increase in the level of economic activity, and of the goods and services available. It is a period of economic growth as measured by a rise in real GDP.[1] The explanation of such fluctuations in aggregate economic activity is one of the primary concerns of macroeconomics.
Case Study on Moral Identity and the Expanding Circle of Moral Regard toward Out-Groups
This article examines moral identity and reactions to out-groups during intergroup conflict. Four studies suggest that a highly self-important moral identity is associated with an expansive circle of moral regard toward out-group members (Study 1) and more favorable attitudes toward relief efforts to aid out-group members (Study 2). Study 3 examines moral identity and national identity influences on the provision of financial assistance to out-groups. Study 4 investigates the relationship between moral identity and (a) the willingness to harm innocent out-group members not involved in the conflict and (b) moral judgments of revenge and forgiveness toward out-group members directly responsible for transgressions against the ingroup. Results are discussed in terms of self-regulatory mechanisms that mitigate in-group favoritism and out-group hostility.
A well-established finding in social psychology is the in-group favoritism and out-group hostility that occurs during intergroup conflict. For many Americans, the September 11th attacks on the World Trade Center and Pentagon had profound psychological effects on in-group national identity. Not surprisingly, these events triggered a variety of hostile reactions toward out-group members perceived as being closely, or even loosely, associated with the
Americus Reed II, Marketing Department, The Wharton School, University of Pennsylvania; Karl F. Aquino, Management Department, College of Business, University of Delaware. Both authors contributed equally to this research. Order of authorship was reverse alphabetical. This study was partially supported by a College of Business and Economics Research Summer Grant awarded to Karl F. Aquino. For either administrative or data collection assistance or for comments on a draft of the article, we thank Alan Ercan Aquino, Mark Forehand, Gavan Fitzsimmons, Nancy Kim, Karen Reed, Christophe Van den Bulte, Patti Williams, and participants of the four-school (Columbia University; New York University; The Wharton School, University of Pennsylvania; and Yale University) Marketing Colloquium held at New York University in May of 2002. Young Lee is especially thanked for assistance with computer programming for Studies 2 and 4. Steve Grover is especially thanked for data collection assistance for Study 3. Correspondence concerning this article should be addressed to Americus Reed II, The Wharton School, University of Pennsylvania, 3730 Walnut Street, 700 Jon M. Huntsman Hall, Philadelphia, Pennsylvania 191046371. E-mail: [email protected]
terrorist attacks. For example, reports of anti-Muslim sentiment expressed by some Americans were evidenced by increased racial and ethnic violence, heightened suspicion and mistrust of immigrants from Middle-East countries, and public support for targeted surveillance of people whose physical characteristics fit the terrorist prototype (cf. American Arab Anti Discrimination Committee: http://www.adc.org/education/advice.htm) One explanation for out-group hostility in the aftermath of September 11th is self-protective motives at either the individual (selfidentity) or group (social identity) level. That is, out-group denigration often upholds the self- or national image in the face of an external threat. Although the motivation to enhance one's self or one's group by negatively evaluating others is unquestionably powerful, some people are less susceptible to this motivation. For example, in contrast to the hostility that some Americans showed toward Muslims and Arab Americans after September 11th, others pleaded for racial tolerance and openly condemned acts of discrimination directed against fellow citizens and even noncitizens (cf. Coalition for World Peace: http://www.actionla.org/peace/#point). This article explores a possible social psychological explanation for the attitudes and behaviors of people who respond less negatively toward out-groups even in the midst of inter- group conflict. This explanation derives from the psychological effects of a self-regulating construct referred to as moral iden- tity that connects the individual (personal identity) to others (social identity) through the evaluative implications of a set of moral associations that define the moral self (Aquino & Reed, 2002). We hypothesize that moral identity can influence out- group hostility because its self-importance alters the psycho- logical boundaries that define in-group membership. When the activation potential of moral identity is high —that is, when moral identity assumes high self-importance—the self/other relation should be characterized by a more expansive concep- tion of the in-group toward which a person feels obligated to
1270
MORAL IDENTITY AND OUT-GROUPS
1271
exhibit moral regard.1 We refer to this premise as the "circle of moral regard" hypothesis: Rather than confining one's in-group to a narrow set of others (e.g., family, kin, fellow citizens), a person whose moral identity has high self-importance should include a larger set of social groups. In the extreme, this psychological boundary might include all of humanity. The goal of this article is to broadly examine the circle of moral regard hypothesis. We begin by briefly reviewing the literature on self-protective mechanisms based on personal (self-affirmation) and social (self-esteem through group status) identities. Next, some key conditions that may generate in-group favoritism and out-group hostility are described. Third, the concept of moral identity is defined and its relationship to out-group delineations and in-group favoritism is highlighted. Finally, the results of four studies that have investigated the connection between moral identity and perceptions and behavior toward out-groups are reported and discussed.
Triggers of Out-Group Hostility
Out-group hostility is often triggered when out-group members pose a realistic threat to in-group physical and psychological well-being (e.g., Quillian, 1995) or symbolically threaten the stability of in-group values and beliefs (Esses, Haddock, & Zanna, 1993) or when encounters with out-group members cause anxiety as a result of condemnation and rejection by those out-group members (e.g., Stephan & Stephan, 1993). Moreover, if particular circumstances (e.g., a vivid or dramatic event) make intergroup differences highly salient, then broad-based stereotypes of outgroup members can evoke unfavorable expectations of conflictladen interactions (e.g., Stephan, Ybarra, Martinez, Schwarzwald, & Tur-Kaspa, 1998). The attacks on the World Trade Center and Pentagon created many of the conditions that give rise to hostility toward outgroups. As a result, for some Americans, the image of Arab Americans and Muslims now begins and ends with a terrorist group prototype. This image lays the psychological foundation for acts of prejudice and discrimination directed toward the target out-group (Deaux, 2002). In the wake of strong situational forces invoked in the aftermath of September 11th, an increase in hostility directed toward out-group members fitting the terrorist prototype would seem inevitable. Yet, as mentioned earlier, many Americans have openly condemned acts of discrimination and some have gone out of their way to provide comfort and protection for their Arab and Muslim neighbors. For example, when one suburban mosque was vandalized after the September 11th terrorist attacks, a group of concerned American citizens volunteered at considerable personal risk to accompany Islamic women to the grocery store, to medical appointments, or on other errands after discovering that these Muslim women were afraid to go out in public (Dart, 2001). We theorize that the attitudes and behaviors of people who assisted these out-group members may be partially explained by a person's conception of a moral self (Aquino &Reed, 2002) that defines the self's relation to others and establishes the psychological boundaries between in-groups (us) and out- groups (them).
Self-Protection Through Self-Affirmation of Personal and Group Identity
Several studies have shown that self-image threat can lead to negative evaluations of out-groups (Brown & Gallagher, 1992; Gibbons & Gerrard, 1991). One reason this occurs is because information that threatens perceivers' sense of self-worth evokes the need to restore a positive self-image. This goal can be achieved by negatively evaluating out-groups (cf. Steele, 1988; Steele &Liu, 1983; Steele, Spencer, &Lynch, 1993). Unless other internal motivations are activated, such as a goal of accountability (Tetlock, 1983), social desirability and/or egalitarian motives (Dovidio & Gaertner, 1991), or self-image bolstered with a self-affirmation procedure (Fein &Spencer, 1997; Study 1), people may find the denigration of out-groups to be an effective way of maintaining "an image of self-integrity" (Steele et al., 1993, p. 885). Social identity theory (Tajfel, 1959; Tajfel & Turner, 1979), an influential theoretical explanation for in-group favoritism and outgroup denigration at the group level, makes a similar assertion. Like the self-affirming processes at the individual level, positive evaluations of in-groups and unfavorable responses toward outgroups are presumed to serve self-protective and self-enhancing functions. One way people maintain or increase their self-esteem is through group membership (Turner, Hogg, Oakes, Reicher, &Wetherell, 1987). Consequently, it is beneficial for people to ascribe favorable characteristics to the groups to which they belong (Hewstone, Rubin & Willis, 2002). Once the concept of "us" is established, however, self-esteem can also be enhanced by negatively evaluating groups to which one does not belong. For example, a person may selectively search for intergroup differences that favor their group and dismiss information that favors the out-group (e.g., Brewer & Kramer, 1985; Devine, 1989; Mikulincer & Shaver, 2001). This can lead them to view out-group members as inferiors. Self-serving causal explanations for ingroup outcomes (e.g., Hamilton & Trolier, 1986) and the use of linguistic labels to differentiate in-groups and out-groups also increase the salience of intergroup differences (e.g., Dovidio & Gaertner, 1993), further encouraging in-group favoritism. Several conditions seem to reliably invoke these kinds of in-group/outgroup effects. The more prominent are reviewed briefly as follows.
The Role of a Moral Self-Conception
It has been argued that a defining characteristic of a person with a legitimate moral identity is that he or she extends feelings of sympathy and affiliation toward a larger segment of humanity than someone whose moral identity is less important (Hart, Atkins & Ford, 1998; Younis & Yates, 1999). If this assertion is valid, then compared with those whose moral identity has low selfimportance, people whose moral identity has high self-importance should be relatively less likely to exhibit in-group favoritism and out-group hostility even during times of intergroup conflict. One of the problems with testing this prediction is that there have been few attempts to provide a definition of moral identity that corresponds to a psychometrically sound measure of its self-importance.
1 Following the definition of morality proposed by moral philosophers (e.g., Hume, 1751/1996) and psychologists (e.g., Eisenberg, 2000; Rest, 1979), moral regard is defined in this study as showing concern for the needs and welfare of others.
1272
REED AND AQUINO
Recent work by Aquino and Reed (2002) addressed this issue by offering both a theoretical framework and a methodological approach for testing the relationship between moral identity and ingroup/out-group effects.
gence of a spontaneous self-concept organized around moral themes (cf. Hart & Fegley, 1995), self-reported volunteering, and the actual donation of food to the needy.
Moral Identity: A Theoretical Framework
Aquino and Reed (2002) proposed a trait-based conceptualization of moral identity and presented empirical evidence supporting the construct and predictive validity of a measure of its selfimportance. They defined moral identity as "a self-conception organized around a set of moral traits." On the basis of social cognitive theories of the self (cf. Kihlstrom & Klein, 1994), Aquino and Reed (2002) argued that moral identity is linked to specific moral traits (cf. Lapsley & Lasky, 2001; Walker & Pitts, 1998) and can, therefore, be invoked by tapping a subset of traits that are associatively linked to other traits that form part of the person's unique moral self-conception. They further proposed that because moral identity is amenable to a distinct mental image of what a moral person is likely to think, feel, and do (Kihlstrom & Klein, 1994), it can be associated with a social referent that could be a real membership group (e.g., fellow Peace Corp volunteers), an abstracted ideal (e.g., God), a known individual (e.g., a big brother in a mentoring program), an unknown individual (e.g., Mother Teresa), or any social construction (cf. Reed, 2002). Consequently, moral identity is one of the many possible social identities that comprise a person's social self-schema (cf. Markus, 1977). Because the traits comprising this identity have moral content, however, its self-importance should be a stronger predictor of moral affect, cognitions, and behaviors than other identities (e.g., occupation, gender) that are less likely to be imbued with moral meaning (cf. Aquino & Reed, 2002, 2003; Reed, 2002; Reed, Aquino, & Ray, 2003). Building on their trait-based conception, Aquino and Reed (2002) developed an instrument that taps two aspects of identity as defined by Erikson (1964): identity as being rooted in the very core of one's being and as being true to one's self in action. Procedurally, Aquino and Reed's (2002) instrument uses trait stimuli to make a moral identity schema salient and then asks respondents several questions to measure the self-importance of their unique moral identities along the two dimensions (see Appendix).2 In six studies using diverse samples, Aquino and Reed (2002) showed that the items tap the two dimensions of identity that are consistent with Erikson's (1964) conception of an identity as being rooted in the very core of one's being and as being true to oneself in action. Aquino and Reed (2002) labeled these dimensions internalization and symbolization. Internalization reflects the degree to which a set of moral traits is central to the self-concept, whereas Symbolization reflects the degree to which these traits are expressed publicly through the person's actions in the world.3 This latter dimension corresponds to Laughlin's (1970) definition of symbolization as the process "through which an external object becomes the disguised outward representation for another internal and hidden object, idea, person, or complex" (p. 414). Aquino and Reed demonstrated that the instrument they developed to measure these dimensions of moral identity had a stable factor structure; was internally consistent; possessed nomological validity; and predicted a variety of morally relevant outcomes, including the emer-
Present Research
In this research we examine whether the self-importance of moral identity as defined by the two dimensions of internalization and symbolization (Aquino & Reed, 2002) is associated with an expansive circle of moral regard. Does a highly self-important moral identity diminish the psychological boundary between ingroups and out-groups and extend the web of sympathy toward
2 Following Blasi (1984), Aquino and Reed (2002) made several assumptions about moral identity that are also assumed in this article. First, they acknowledge that the specific content of people's moral identities is likely to vary. So whereas one person may see being compassionate as critical to his or her moral identity, another person might emphasize being just and fair. Even though there may be some nonoverlapping moral traits that compose each person's unique moral identity, a common set of traits appears to be central to many individuals' moral self-definitions (Aquino &Reed, 2002; Lapsley & Lasky, 2001; Walker & Pitts, 1998). Second, Aquino and Reed assume that being a moral person may be, but need not be, a part of a person's overall self-definition. This assumption is consistent with the notion of the self as comprising multiple identities (Brewer & Gardner, 1996) that are hierarchically ordered (Stryker, 1980). Consequently, the self-importance of moral identity may change as a function of place and time (Hart et al., 1998). Last, Aquino and Reed (2002) do not claim that having a highly self-important moral identity makes a normative or prescriptive statement about a person's moral character or beliefs. This last point is important because there are many different views of morality and what the ideal moral character should be. Aquino and Reed's conception of moral identity does not presume to reflect or endorse any particular moral view or position about what moral self-representation a person ought to possess (cf. Aquino & Reed, 2003). 3 One way to view Aquino and Reed's (2002) instrument, or any instrument that asks people to rate themselves relative to any dimension of the self, is from an impression management perspective (Schlenker, 1985). In other words, one could argue that the moral identity instrument (see Appendix) captures merely the importance of maintaining and conveying a socially constructed image of a moral person. Such motivation could conceivably be purely self-presentational. However, the conceptual definition of moral identity offered by Aquino and Reed and the underlying dimensions linked to Erikson's notion of identity assume that the operationalization of moral identity using Aquino and Reed's instrument captures (at least along the internalization dimension) the importance of how critical that social image is a part of a person's objective sense of who they are (i.e., their actual self). Although the instrument can be thought of as tapping merely into the projected image of a social construction, Aquino and Reed have presented empirical evidence that the instrument taps into a deeper sense of self-relevance. For example, in Study 2 of their article, they show a strong correlation between the internalization dimension and an Implicit Association Test (Greenwald et al., 1998; Greenwald & Farnham, 2000) modified to include the moral traits used as salience induction stimuli in their instrument. To the extent that this implicit measure reflects aspects of self that are not amenable merely to self-presentational projected image concerns, then there is at least some preliminary evidence that the measure really captures aspects beyond the social desirability to adhere to the constructed image of a moral person—which is consistent with the definition of moral identity used by Aquino and Reed and also used in this article.
MORAL IDENTITY AND OUT-GROUPS
1273
others? If so, what are its cognitive and behavioral consequences? Four studies examine these questions. In Study 1, the positive relation between the self-importance of moral identity and the tendency to self-report greater moral regard for out-groups is empirically established. Studies 2- 4 examine the implications of an expanded circle of moral regard in the context of the September 11th terrorist attacks. Study 2 examines whether the selfimportance of moral identity is related to judgments of relief efforts intended to aid out-groups that are either closely or not closely associated with a salient intergroup conflict. Study 3 builds on Study 2 by testing whether a highly self-important moral identity predicts whether people differentially provide actual monetary assistance to equally deserving in-group and out-group members. Finally, Study 4 further explores the circle of moral regard hypothesis by examining the effect of a highly self-important moral identity on negative responses toward innocent out-group members as well as moral judgments about how to appropriately respond to out-group members held directly responsible for ingroup transgressions.
between these constructs. However, there is another, more indirect approach that provides a further test of the circle of moral regard hypothesis. This approach applies a well-established social psychological theory about how the kinds of resources that people exchange with one another in everyday life depend upon the nature and quality of relationships with exchange partners.
Exchange of Particularistic Resources
Social resource theory (Foa & Foa, 1974) posits that humans exchange six basic resources in social relationships: money, information, goods, services, status, and love. Foa and Foa (1974) refer to money, goods, and information as universalistic resources. In contrast, love, status, and services are referred to as particularistic resources. Social resource theory predicts that the qualities of the exchange partner and the nature of the relationship are more (less) important to the decision to exchange a social resource when the exchange involves a particularistic (universalistic) resource. For example, most people are willing to exchange money, which is the most universalistic resource, with just about anybody. They are much more selective when deciding with whom to exchange love or status. Indeed, studies have shown that particularistic resources are exchanged more frequently with people who are psychologically close rather than distant (e.g., Blieszner, 1993). On the basis of these findings, it is hypothesized that if moral identity is associated with an expansive circle of moral regard, then as the self-importance of this identity increases, so too should the person's perceived moral obligation to exchange highly particularistic resources with a stranger. The reason being that for a person with a highly self-important moral identity, a "stranger" is more likely to be included within the circle of moral regard. On the basis of these arguments, Study 1 tested two hypotheses. The first is that in terms of the self-importance of moral identity, high as compared with low moral identifiers would report having a more expansive circle of moral regard. The second hypothesis is that high as compared with low moral identifiers would be more willing to exchange particularistic resources with strangers.
Study 1: Moral Identity and the Circle of Moral Regard
It has been suggested that a person whose moral identity has high self-importance should show greater concern for a larger segment of humanity than someone whose moral identity is less important (Hart et al., 1998; Younis & Yates, 1999). This argument parallels the philosopher Peter Singer's (1981) assertion that moral progress can emerge from a fixed moral sense if people are willing to expand the mental dotted line that embraces the entities deemed worthy of moral consideration. According to Younis and Yates (1999), studies of European adults known to have protected and sheltered Jews during the Nazi occupation of World War II (e.g., Monroe & Epperson, 1994; Oliner & Oliner, 1988) provide anecdotal exemplars of this cognitive expansion of moral regard. These studies found that one characteristic that distinguished rescuers from nonrescuers is that the former tended to perceive Jews as fellow human beings who merited the same privileges and community rights. In contrast, nonrescuers viewed Jews as outsiders, as "they" rather than non-Jewish "we." According to Monroe and Epperson (1994), the ethical actions of rescuers flowed spontaneously from deep-seated dispositions revealed in the self's relationship to others.
Method Samples and Procedures
Data from two independent samples drawn from two different universities located in the northeastern United States were used to test the study hypotheses. The first sample (Sample 1) consisted of 137 undergraduates enrolled in an introductory marketing class. This sample was used to test the hypothesized relationship between the self-importance of moral identity and perceived moral obligations toward out-groups. Students received course credit for their participation. Seventy-one of the students were women. Fifty percent of the students identified themselves as White, 35% as Asian, 5% as Hispanic, 3% as African American, 2% as Middle Eastern, and 5% as other. Their average age was 19.4 years (SD 0.89). To fulfill a requirement for the marketing class, participants completed several penciland-paper questionnaires throughout the semester. As part of one of those sessions, participants arrived at the behavioral laboratory where they anonymously completed a measure of the self-importance of moral identity and several questions intended to assess their perceived obligation to show concern for various groups. After data collection was complete, participants were fully debriefed, thanked for their participation, and dismissed.
Whose Needs Merit Moral Consideration?
The self- other relationship is fundamental to a person's moral identity (Younis & Yates, 1999). As one rescuer interviewed by Oliner and Oliner (1988) put it: "I insist on saying that it was absolutely natural to have done this [rescuing]. You don't have to glorify yourself— considering that we are all children of God and that it is impossible to distinguish between one human and another" (p. 228).Younis and Yates (1999) interpreted this rescuer's rationale as evidence that a psychological consequence of a selfimportant moral identity is that it promotes a belief that all people share a common humanity. If this interpretation is valid, then there should be a positive relationship between the self-importance of moral identity and the extent to which people feel obligated to show moral concern for the welfare of out-groups. One way to evaluate this assertion directly is to show an empirical relationship
1274
REED AND AQUINO Following an approach that Turner, Foa, and Foa (1971) used to test social resource theory, participants were then presented with four descriptive statements of social exchanges involving particularistic resources. In two separate scenarios, participants were told that the resources were to be exchanged with both a "family member you feel closest to" and a "stranger." Two of the exchanges involved love ("Telling them I love you," "Telling them you feel affection for them") and two involved status ("Telling them you are proud of them," "Telling them you respect and admire them"). Participants were then asked to indicate how obligated they were to exchange these resources with a family member and a stranger. Responses were on a 5-point Likert scale that ranged from 1 (not at all obligated) to 5 (very much obligated). In each of the two scenarios, the four items describing the exchanges were combined into a single scale representing particularistic resources. The Cronbach's alpha reliabilities for this scale were .87 and .70 in the family member and stranger scenarios, respectively.
The second sample (Sample 2) consisted of 55 master's of business administration students enrolled in two master's-level organizational behavior classes. Sample 2 was used to test the hypothesis that, compared with low moral identifiers, high moral identifiers would be more willing to exchange particularistic resources with strangers. The students completed the moral identity self-importance measure prior to a class lecture on the role of personality in organizations. One month later they completed a questionnaire assessing the extent to which they felt obligated to exchange various kinds of particularistic resources with both a friend and a stranger. The confidentiality of students' responses was maintained by assigning code numbers to each survey. After completing both questionnaires, the results were reported to students in the aggregate and were used as the basis for a class discussion on social exchange theory. Forty-eight students provided usable data on both questionnaires. Thirty-nine respondents were men and 89% were White. Their average age was 29.1 years (SD 6.0).
Measures
Moral identity. Aquino and Reed's (2002) 10-item Self-Importance of Moral Identity Scale was used to measure this construct (see Appendix). The theoretical basis for the instrument is that by tapping certain trait stimuli, a broader associative network of related traits is invoked. Past research has shown that stimulus cues such as words can heighten the salience of a particular social identity (Forehand, Deshpande, & Reed, ´ 2002). Using an inductive procedure, Aquino and Reed identified a set of traits that many people may consider to be characteristic of a moral person (Lapsley & Lasky, 2001; Walker & Pitts, 1998). These traits were then used as identity-invoking stimuli in the measure, recognizing that the traits do not comprise an exhaustive set which all persons in all places would consider as being central to their unique moral identities. Respondents in Samples 1 and 2 answered the 10 questions on a 5-point Likert scale that ranged from 1 (strongly disagree) to 5 (strongly agree). Items for the Internalization and Symbolization subscales were averaged. The Cronbach's alpha reliability for both subscales was .83 in Sample 1, and the means were 4.5 (SD 0.5) and 3.1 (SD .7) for Internalization and Symbolization, respectively. In Sample 2, reliabilities were s .85 and .71 with means of 4.3 (SD 0.5) and 3.0 (SD 0.6) for the Internalization and Symbolization subscales, respectively. Moral regard for out-groups. The expansiveness of a person's circle of moral regard was measured by asking Sample 1 participants to report the extent to which they believed they had "a moral or ethical obligation to show concern for the welfare and interests" of four different out-groups: "People from another country," "Strangers," "People who practice a different religion than you," and "People of different ethnicities than you." Respondents answered the items on a 5-point Likert scale that ranged from 1 (absolutely no obligation) to 5 (very strong obligation). These items were averaged to form a scale ( .88) where a high score indicated a stronger moral obligation to show concern for the welfare of the four outgroups. Willingness to exchange particularistic resources. Sample 2 participants were asked to indicate the extent to which they felt obligated to provide to both a friend and a stranger, love and status—two highly particularistic resources identified by Foa and Foa (1980). To elicit these responses, participants were presented with the following instructions: All of us have many different relationships in our lives. In these relationships, we give and receive certain things from one another that may be described as "social resources." These resources may be material, like a diamond ring, or symbolic, like a smile. Listed below are several statements describing different types of social resources. Using the scale below, please indicate how obligated you are to provide each in the context of the relationship described, assuming that an appropriate opportunity presents itself for you to do so .
Results Concern for Out-Groups
The hypothesis that high moral identifiers would report a more expansive circle of moral regard was tested by regressing perceived moral obligation toward out-groups on the two dimensions of the self-importance of moral identity measure. Gender was used as a control variable because past research has shown that men tend to show greater intergroup bias than women (Sidanius, Levin, Liu, & Pratto, 2000; for reviews, see Pratto, Sidanius, & Stallworth, 1993; cf. Gaertner & Insko, 2000). Participants' race was also controlled because it is possible that compared with members of the socially dominant group (e.g., Whites), people who belong to socially subordinate racial groups (e.g., African Americans, Hispanics, Asians) within American society might identify more closely with out-groups. This might lead the latter to report stronger moral obligations to be concerned about out-group welfare, irrespective of the self-importance of their moral identity. Both control variables were dummy coded (0 male, 1 female and 0 non-White, 1 White). Results of this analysis showed one significant effect. The Internalization dimension was positively related to perceived moral obligations toward out-groups ( .30, p .01). The direction of the effect is consistent with the prediction that the higher a person's self-importance of moral identity, the more likely he or she was to report a stronger moral obligation to show concern for a variety of out-groups. This result is interpreted as supporting the hypothesis that high moral identifiers have a more expansive circle of moral regard than low moral identifiers. However, this was only true for the internalization dimension of the self-importance of moral identity measure. Having established a direct empirical relationship between the self-importance of a person's moral identity and concern for outgroups, Sample 2 data were analyzed to provide a further test of the circle of moral regard hypothesis using the predictions of social resource theory.
Exchange of Particularistic Resources
The second test of whether high moral identifiers have a more expansive circle of moral regard than low moral identifiers was performed by first dividing participants into two groups based on their scores on the two Self-Importance of Moral Identity sub-
MORAL IDENTITY AND OUT-GROUPS
1275
scales. Group assignment was determined by a median split on the Internalization, high (n 24) versus low (n 25), and Symbolization, high (n 25) versus low (n 24), subscales. Next, a 2 2 2 (Recipient Internalization Symbolization) analysis of variance with repeated measures on recipient (friend vs. stranger) was performed. There were no significant differences in the dependent variables as a function of gender, so this variable was not included in the analysis. Race was also not included because Sample 2 was composed almost exclusively of Whites, so there was insufficient variation on this characteristic to make meaningful comparisons. Results showed a significant main effect of recipient such that the participants reported a stronger moral obligation to exchange particularistic resources with a friend (M 4.4) than a stranger (M 1.7), F(1, 45) 293.86, p .01. This result supports social resource theory (Foa & Foa, 1980). A significant main effect of internalization was also found, F(1, 45) 4.49, p .05. The effect indicated that participants who scored higher on the Internalization subscale felt more obligated to exchange particularistic resources with both friends and strangers (M 3.2) than people who scored lower on this subscale (M 2.9). However, both of these main effects are qualified by a significant Recipient Internalization interaction, F(1, 45) 5.79, p .01. The pattern of means for this interaction supports the prediction that high moral identifiers are more willing to exchange particularistic resources with strangers than low moral identifiers. Specifically, high (M 4.4) and low (M 4.4) scorers on Internalization did not differ in their perceived obligations to exchange particularistic resources with a friend, t(47) .35, ns. However, they differed significantly in their felt obligation to exchange these resources with a stranger, t(47) 3.26, p .01, with high moral identifiers (M 2.1) feeling more obligated than low moral identifiers (M 1.4). No other main or interaction effects were significant.
feels strong bonds of attachment, trust, or affiliation. Data from Sample 2 suggest that as a person's moral identity gains selfimportance, he or she feels a stronger moral obligation to exchange such resources with strangers. This finding is interpreted as providing further evidence that having a self-important moral identity might be associated with a cognitive expansion of the boundaries of in-group membership. The data for Sample 2 were collected at two different time points and used adult nonstudent as compared with student participants. However, Sample 2 also has limitations. One is that the operationalization of out-group used in Sample 2 (i.e., the stranger scenario) is a somewhat abstract and far removed notion of the psychological concept of an out-group. Second, it was argued earlier that the expanded circle of moral regard would persist for high moral identifiers even when strong situational pressures (such as intergroup conflict) establish more distinct psychological boundaries between in-groups and out-groups. The Sample 2 data included no such situational factors. Finally, the dependent measure of perceived obligation does not necessarily mean that people would in fact behave differently toward out-groups as a function of the self-importance of moral identity. The limitations of both Samples 1 and 2 of Study 1 suggest that a stronger test of the circle of moral regard hypothesis would require an examination of a person's reactions toward real outgroups closely linked to salient intergroup conflict. It should also include a measure of actual behavior. These were the goals of the next two studies. In Studies 2 and 3, people's attitudes, intentions, and behavior toward organizations that render aid to specific out-groups were examined as a function of the self-importance of moral identity.
Study 2: Judgments of Relief Efforts That Aid Out-Groups
The findings from Study 1 offer preliminary evidence supporting the proposed link between the self-importance of moral identity and the expanding circle of moral regard. Study 2 presents a further test of this prediction by examining how people's moral identity self-importance influences perceptions of charitable organizations that aid specific out-groups. If a moral self is indeed associated with a more expansive circle of moral regard, a highly self-important moral identity should be associated with more favorable perceptions and judgments of relief efforts meant to aid out-groups (e.g., relief effort for Middle East refugees). More specifically, compared with low moral identifiers, high moral identifiers should evaluate these efforts as more worthwhile. Furthermore, in the aftermath of September 11th, it is hypothesized that this difference is influenced by whether the out-group receiving assistance is loosely associated with the event (e.g., Afghani refugees) as compared to a more neutral (e.g., Turkish refugees) out-group prototype. If a highly self-important moral identity mitigates out-group negativity by expanding the circle of moral regard and broadening the web of sympathy toward others, then high as compared with low moral identifiers should evaluate relief efforts that aid out-groups more favorably in general. However, this relation is expected to be stronger for high as compared with low moral identifiers even if the out-group (e.g., Afghani refugees)
Discussion and Limitations of Study 1
Of the two dimensions of the self-importance measure of moral identity proposed by Aquino and Reed (2002), the data from Sample 1 showed that internalization was a significant predictor of the extent to which people reported a moral obligation to show concern for the needs and welfare of out-groups (e.g., people from other countries, people of different religions, people of different ethnicities). A limitation of Sample 1 was that the independent and dependent measures were collected at the same time so the relationship between the self-importance of moral identity measure and perceived obligations toward out-groups might have been inflated because of common method bias. Moreover, the crosssectional nature of the research design used for Sample 1 may subject the results to the alternative explanation of self-generated validity (cf. Feldman & Lynch, 1988). The results using Sample 2 addressed these limitations while providing a further test of the proposed relationship between the self-importance of moral identity and the circle of moral regard, but using a different theoretical approach. Social resource theory (Foa & Foa, 1980) posits that particularistic resources are perceived as more personal and therefore are more likely to be exchanged only with those for whom the person
1276
REED AND AQUINO Thousands of pregnant women are among the Kurdish (Afghani) civilians who have fled their homes in recent days and are massed along the country's borders. The lack of shelter, food and medical care, and unsanitary conditions pose a serious risk to these women and their infant children. Even before the current crisis, poor health conditions and malnutrition made pregnancy and childbirth exceptionally dangerous for Kurdish (Afghani) women. UNPF is pre-positioning emergency relief supplies in the countries bordering Turkey (Afghanistan) in order to provide displaced Kurdish (Afghani) women with lifesaving reproductive health care services. The Fund is asking donors for $4.5 million to counter the health risks facing refugees and the internally displaced. PRESS RELEASE: UNITED NATIONS, New York, 28 September 2001—Responding to the grave health emergency now facing Kurdish (Afghani) women, the United Nations Population Fund (UNPF) is mounting its largest-ever humanitarian operation. The Fund is asking international donors for $4.5 million to support the effort. The charitable efforts were dummy coded (0 Turkish, 1 Afghani) for the purposes of analysis. Self-importance of moral identity. Aquino and Reed's (2002) 10-item Self-Importance of Moral Identity Scale was administered during the consumer questionnaire exercise. Respondents answered the items on a 7point Likert scale that ranged from 1 (strongly disagree) to 7 (strongly agree). The test was administered by computer. Items measuring the .87) and Symbolization ( .81) dimensions were Internalization ( averaged to form two subscales. Control variables. Gender and race were again entered as control variables and both were dummy coded (0 Male, 1 Female and 0 Non-White, 1 White).
is loosely associated with an intergroup conflict episode (e.g., the attacks on the World Trade Center and Pentagon).
Method Participants
Seventy-nine undergraduate students at a university located in the northeastern United States participated in the experiment. Average age was 19.6 years (SD 1.15). Forty-one were men and 38 were women. Thirty-seven identified themselves as White, 4 were African American, 27 were Asian, 6 were Hispanic, and 5 identified themselves as other. Seventy-five participants provided usable data on all study variables.
Procedure
The study took place 5 months after the September 11th terrorist attacks. Participants came to a behavioral laboratory and completed seven ostensibly unrelated research tasks for class credit. The first task was to complete a consumer questionnaire exercise. After completing three separate "filler tasks" that took approximately 30 min, participants completed an online research survey that measured perceptions of charitable organizations. Participants were shown a picture of an online brochure of a particular humanitarian relief effort to be described later. They were asked to read the information. They were then asked their perceptions of the organization and the relief effort as well as their intentions to donate. The confidentiality of students' responses was strictly maintained by assigning code numbers to all instruments. Participants then completed two additional filler tasks, were debriefed, and dismissed.
Independent Variables
Out-group target of relief effort. Participants were presented with information about one of two different efforts by the United Nations to aid out-groups. This information was presented using mock-up online brochures. Participants were randomly assigned to either the salient conflict out-group (Afghan) or the neutral out-group (Turkey) condition. All of the brochure information was identical except for the target group who would benefit from the relief effort. A picture of a Middle-Eastern woman and her child was shown along with the headline: Humanitarian Crisis in Turkey (Afghanistan). The information on the online brochure read as follows: The United Nations Population Fund (UNPF) is mounting its largestever humanitarian operation to save the lives of the most vulnerable victims of the current exodus and health crisis facing Turkey (Afghanistan): women and girls. Hospitals and clinics are in ruins after 23 years of war. Tens of thousands of civilians are on the move fearing famine and death.
Dependent Variables
Five items were written to assess judgments of the perceived worthiness of the United Nations' efforts to aid the two out-groups (ranked from 1 [strongly disagree] to 7 [strongly agree]). These items were as follows: "The above cause is not as important as other causes here in the U.S." (reverse scored); "Our government should provide money in this effort"; "If given the opportunity, I would donate to this effort"; "I don't feel very obligated to this particular cause" (reverse scored); and "I would be willing to volunteer my time to help this organization." A principal-components analysis showed that these items loaded on a single factor accounting for 57% of the variance. Consequently, they were averaged to form a scale ( .81).
Results
Descriptive statistics and correlations among the study variables are shown in Table 1. Hierarchical regression analyses were used
Table 1 Study 2: Descriptive Statistics and Correlations Among Study Variables
Study variable 1. 2. 3. 4. 5. Worthiness of relief effort Female White Afghan relief effort Internalization 75. ** p .01. *** p M 5.4 0.5 0.5 0.5 5.2 .001. SD 1.3 0.5 0.5 0.5 0.6 1 — .18 .25* .21 .18 2 — .12 .18 .37*** 3 4 5 6
— .09 .06
— .21
—6.
Symbolization 2.9
0.7
.01
Note. N * p .05.
MORAL IDENTITY AND OUT-GROUPS
1277
to test the study hypotheses. The control and independent variables were entered in Step 1. The interactions between the two outgroups (Turkish vs. Afghani) served by the charitable organization and participants' scores on the Internalization and Symbolization subscales were entered in Step 2. In order to minimize multicollinearity between the interaction terms and their components (Aiken & West, 1991), all of the variables were centered prior to testing hypotheses. The results of the analysis are displayed in Table 2. The significance of individual regression weights was appraised to test the hypothesis that high moral identifiers judge the worthiness of the two relief efforts more favorably than low moral identifiers. The variance explained by the interaction terms was examined to test the hypothesis of differential perceptions of out-groups that are associated with a salient intergroup conflict (Afghani refugees and the September 11th attacks). As the hypothesis for a direct effect of self-importance of moral identity was directional, a one-tailed test was used to test this prediction. The results showed that two variables—race and internalization—were directly related to judgments of the worthiness of the relief effort. The effect of race indicated that White respondents judged the relief efforts less favorably than non-Whites ( .63, p .05). There was also a marginally significant effect of the target of the relief effort, with participants evaluating the Afghan effort more favorably than the Kurdish effort ( .58, p .10). Using a one-tailed test, we found that the effect of internalization supported the prediction that the self-importance of moral identity is positively related to the perceived worthiness of relief efforts directed toward out-groups ( .50, p .05). This main effect, however, is qualified by a significant interaction. The set of interactions explained a significant amount of incremental variance in the dependent variables (R2 .08, p .05). Inspection of the individual regression weights showed that the Internalization Out-group target interaction term was significant (B 1.35, p .05). We analyzed the form of the interaction by regressing judgments of cause worthiness on internalization within groups of participants who evaluated either the Afghan or Turkish
relief effort. The simple slopes in each of these groups showed that the relationship between internalization and judgments of worthiness was considerably stronger when participants evaluated the Afghani (B .58, p .01) as compared with the Turkish (B .05, ns) relief effort. This finding was consistent with the prediction that in terms of self-importance, low moral identifiers would be much less supportive of relief efforts that benefit an out-group closely associated with the events of September 11th.
Discussion and Limitations
The results of Study 2 lend partial support for the argument that the self-importance of moral identity is associated with an expanded circle of moral regard by means of differential perceptions of real out-groups. The direction of the effect indicates that the higher the self-importance of a person's moral identity, as indicated by the Internalization dimension, the more favorable one's evaluation of a relief effort directed toward two distinct outgroups, controlling for the effects of race and gender. Moreover, self-importance of moral identity showed a stronger effect on this evaluation when the out-group (Afghani refugees) was more likely to be loosely associated with a group —the Taliban regime in Afghanistan— directly implicated in the attacks on the World Trade Center and Pentagon. One interpretation of these findings is that in the face of a powerful contextual influence, low moral identifiers were more likely to alter their judgments of the worthiness of relief efforts meant to aid different out-groups based on the degree to which the out-group invoked an association to the perpetrators of the attacks.4 By inference, this effect can be explained by low moral identifiers having a more restricted and high moral identifiers a more inclusive circle of moral regard. This argument is further upheld by the direct relationship between the self-importance of moral identity and judgments of relief efforts to aid out-groups in general. A key element of the circle of moral regard hypothesis is that a highly self-important moral identity should be associated with an expansion of the psychological boundaries that differentiate ingroups and out-groups as a direct result of a person's conception of a moral self. Although the results of Study 2 are supportive of this argument, they do not provide a direct test because Study 2 involves two out-groups and not an in-group versus an out-group. Given the similarities between the two out-group stimuli (both Middle Eastern out-groups) one may argue that the circle of moral regard hypothesis was not fully tested in Study 2. A stronger test is to see how people respond when given a choice between helping in-group members directly affected by intergroup conflict or members of an out-group associated with that conflict. Finally, if the
4 We theorized that as a person's moral identity increases in selfimportance, so too does the expansiveness of his or her circle of moral regard. This should be reflected in more favorable reactions to those that are even loosely associated with an out-group prototype linked to a salient intergroup conflict. This conceptual argument is not negated by the fact that many Americans and participants in our sample view Afghani women and children as being victims rather than perpetrators. Indeed, this is a key part of the theoretical premise because we hypothesize that responsiveness to the victimization of this out-group will increase as a person's moral † p identity assumes greater self-importance.
Table 2 Study 2: Hierarchical Regression of Moral Identity and Target on Worthiness of Relief Effort
Unstandardized coefficients Model and independent variable Step 1 Female White Afghan relief effort (A) Internalization (I) Symbolization (S) Step 2 AI AS Note. N 75. F(7, 67) .08*, adjusted R2 .16. .10. * p .05. B 0.133 0.629 0.581 0.503 0.139 1.354 0.423 2.95. For Step 1, R2 SE 0.349 0.299 0.314 0.266 0.225 0.515 0.418 t 0.38 2.10* 1.85† 1.89* 0.62 2.62* 1.01
.16*. For Step 2, R2
1278
REED AND AQUINO
circle of moral regard hypothesis provides a sufficiently powerful theoretical explanation for moral action, then the self-importance of moral identity should be not only associated with judgment and cognition but also associated with actual moral behavior, defined as showing social responsiveness to the needs of others (cf. Aquino & Reed, 2002, Study 6). Addressing these concerns was the purpose of Study 3.
Study 3: Providing Assistance to Out-Groups Versus In-Groups
Study 3 confronted participants with two compelling and deserving beneficiary groups: one more likely to be viewed as an ingroup, the other more likely to be viewed as an out-group. The two beneficiary groups were the New York Police and Fire Widows and Children's Benefit Fund (in-group) and the United Nations Children's Fund (UNICEF) Emergency Effort for Afghan Children and Families (out-group). The in-group represented by the New York widows and children is presumed to be emotionally closer to the participants than the distant and foreign Afghan women and children. As in Study 2, it was also assumed that Afghan women and children were more likely to be categorized as members of a highly salient out-group (i.e., Middle Easterners) loosely associated with a recent hostile event directed toward the in-group (i.e., fellow American citizens). What the beneficiary groups held in common, however, was that they comprised primarily people who were unwitting victims of intergroup conflict events beyond their control. To test the influence of the self-importance of moral identity and the expanded circle of moral regard on actual behavior, we required participants in Study 3 to allocate a fixed sum of money to one or both of the beneficiary groups. Following the theoretical arguments presented earlier, it was expected that the selfimportance of a person's moral identity would increase the willingness to allocate money to benefit the out-group. However, this study also examined the effects of a potentially competing "American" identity that might reinforce the notion of an American ingroup in the post-September 11th context. Current models of attitude suggest that people can simultaneously hold "dual attitudes" (Wilson, Lindsey, & Schooler, 2000) that have opposite evaluative implications toward the same attitude object (cf. Reed, Wooten, & Bolton, 2002). Hence, a person might have not only a highly self-important moral identity that motivates him or her to behave one way but also an equally self-important corporate, ethnic, or national identity may be invoked that motivates him or her to behave in another way. In the context of the September 11th attacks, an identity that is likely to conflict directly with a highly self-important moral identity as a predictor of the willingness to aid out- versus in-groups (as operationalized in Study 3) is the strength of a person's American identity. For example, after the September 11th attacks, many Americans demonstrated a surge in patriotism and an affirmation of their self- and national identities by hoisting American flags over homes and businesses, posting bumper stickers with statements exhorting national unity (e.g., United We Stand), or wearing t-shirts expressing the desire to retaliate against those held responsible for attacks (e.g., Osama bin Laden: Wanted Dead or Alive). As noted earlier, these processes of self- and group affirmation
more firmly establish the idea of "us," which, in turn, should constrain and rigidify the boundaries between the in-group (e.g., Americans) and out-group (e.g., Muslims, Middle-Easterners). Therefore, although a highly self-important moral identity is hypothesized to increase the likelihood that people would donate money to the effort to Aid Afghan Children and Families, American identity was expected to reinforce the in-group mentality and to increase the likelihood of donating to the American New York Police and Fire Widows and Children's Benefit Fund.
Method Sample and Procedures
Participants were students taking two separate organizational behavior courses at an urban university in the southern United States. Fifty-eight were undergraduate students and 38 were graduate students taking a master's of business administration course. Fifty-two percent of the students were women. Fifty-five percent of the students were White, 25% were African American, 9% were Asian, 6% were Hispanic, 1% was Native American, and 4% identified their ethnicity as other. The average age was 26.2 years (SD 5.9). Data were collected in three phases 5 weeks apart. The first two phases assessed the independent variables described later. The materials were distributed in class, and students were told it was voluntary to complete them but that the results would be discussed later in the semester. The third phase of the study was the behavioral component. It was conducted during a regular classroom session using the following procedures. Students were given $3 to distribute between the two charities 12 weeks after September 11, 2001. They were then given a 9 12 inch envelope with the following instructions printed on it: This is a study of donation behavior. Three single dollar bills and two white envelopes are enclosed in this large manila envelope. Each of the two envelopes represents one charity, and the charities are described on an enclosed piece of paper. Distribute the three dollars between the two charities by placing the amount you choose in the corresponding envelope. You may not keep the cash. Once you have made your choice and placed the cash in the envelope(s), seal both white envelopes, place them in the large envelope, and seal the large envelope. A one-page description of each of the two charities was included in the large envelope. Descriptions of both charities were drawn from their respective websites. The UNICEF page described the plight of children and families in war-torn Afghanistan and described the type of humanitarian aid (e.g., food and tents and fuel to cope with the impending winter) that would be delivered to them by UNICEF. The graphic included in the description was from the Website, and it depicted a mother comforting her child, both enveloped in a single wrap. The other charity included the following description: "The New York Police and Fire Widows and Children's Benefit Fund provides funds to families at the time of a death to assist with immediate expenses. In addition, the benefit fund distributes an annual check to each of the surviving families." As in the Website, the shields of New York police and firefighters were graphic images shown at the top of the description. Three U.S. dollar bills were physically given to participants to make the cash contributions more tangible and to force a nonequal contribution to the two choices. The participants were told that the money was "really going to the designated charities." (The money was subsequently pooled and actually donated to the charities.) Students were asked not to discuss their decision until after the task was completed and the envelopes were sealed. No other oral instructions were given.
MORAL IDENTITY AND OUT-GROUPS Of the 96 students who completed the self-importance of moral identity questionnaire at Time 1, 76 completed all subsequent phases of the study and provided usable data on all study variables, described in more detail later.
1279
Measures
Self-importance of moral identity. Aquino and Reed's (2002) instrument measured the self-importance of this construct. Respondents answered the items on a 5-point Likert scale rated from 1 (strongly disagree) to 5 (strongly agree). As in Studies 1 and 2, the internalization ( .90) and symbolization ( .66) items were averaged to form two subscales. American identity. Four items interspersed within a general political and environmental opinion questionnaire measured this construct. The items were as follow: "It's great to be an American," "I am extremely proud of my affiliation with the United States of America," "Being a member of this country makes me feel like I share a common goal with others," and "Being an American is not an important part of who I am" (reverse scored). These items were answered on a 5-point Likert scale rated from 1 (strongly disagree) to 5 (strongly agree) and were averaged to form a scale ( .83). Control variables. As in previous studies, gender and ethnicity were used as control variables. Both variables were dummy coded (0 male, 1 female and 0 non-White, 1 White). As the sample consisted of undergraduate and graduate students, it was desirable to control for age because it is possible that the formation of self-importance of moral identity may be influenced by socialization and life experience (Hart et al., 1998). Finally, because the experiment pits dual identities against each other, it is important to consider how people may cognitively resolve the "struggle within." In the domain of moral psychology, one of the most influential constructs for explaining how people resolve moral dilemmas like the one posed in Study 3 is moral reasoning (e.g., Kohlberg, 1971; Rest, 1979). An assumption of the cognitive developmental approach to moral decision making is that the sophistication of one's moral reasoning is ultimately linked to moral behavior (Kohlberg, 1971). To see whether the motivation to reason through moral problems may influence people's resource allocation decisions to donate money to either an in- or out-group, independent of the self-importance of moral identity, the sophistication of moral reasoning was controlled for by our administering a short (three dilemma) version of Rest's (1979) Defining Issues Test (Center for the Study of Ethical Development, 1993). The P-score ( .62) was used to measure the extent to which a person's level of moral reasoning approaches the principled level that cognitive developmental theorists consider to be the highest level of moral reasoning. The moral reasoning measure was collected at Phase 2 of the study. All other nonbehavioral variables were collected during Phase 1.
Donation. Two dependent variables reflected donation behavior: the absolute amount of money donated to the out-group (Afghan women and children) and whether the person allocated a larger share of the $3 to the out-group (i.e., $2 or more). This second variable was treated as binary (0 smaller share to out-group, 1 larger share to out-group) and indicates the degree to which participants were willing to assist the outgroup at cost to the in-group.
Results
The descriptive statistics and correlations among the study variables are shown in Table 3. Multiple regression was performed to test the effects of the two identities (the self-importance of moral identity and the strength of American identity) on the absolute amount of money donated to the out-group. The results of this analysis are shown in Table 4. The results supported the prediction that the self-importance of moral identity predicts the donation of money to the out-group as evidenced by the positive relation between internalization and the dependent variable ( .60, p .01). The results also supported the prediction that the strength of American identity is negatively related to this behavior ( .44, p .01). The data suggest that the self-importance of moral identity and the strength of American identity exerted a strong effect on behavior but in opposite directions. This result is consistent with the earlier argument that these two identities are associated with different views of the self's relationship to others, with a highly self-important moral identity extending the breadth of this relation (to the out-group), and with a strong American identity restricting it (to the in-group). No other effects were significant. The results of a binary logistic regression on the willingness to favor the out-group are also shown in Table 4. The results of this analysis replicate those obtained from the multiple regression, but what is interesting here is the extent to which the independent variables predicted people's decision to favor the out-group at the expense of the in-group. As seen in Table 4, people who scored high as compared to low on the internalization dimension of the self-importance of moral identity were far more likely to allocate a larger proportion of the $3 to the out-group (Afghani women and children). In contrast, people who scored high on the measure of American identity were less likely to do so. More specifically, for every unit increase (decrease) in internalization (strength of Amer-
Table 3 Study 3: Means, Standard Deviations, and Correlations Among Study Variables
Study variable 1. 2. 3. 4. 5. 6. 7. UNICEF Age White Female Moral reasoning Internalization Symbolization M 1.16 26.67 0.55 0.50 31.01 4.48 3.23 SD 1.01 6.20 0.50 0.50 16.80 0.70 0.63 1 — .05 .18 .11 .02 .32* .14 2 — .19 .17 .06 .24* .004 3 4 5 6 7 8
— .16 .32* .12 .13
— .07 .36* .43***
— .19 .02
— .38*
—8.
American identity
Note. UNICEF United Nations Children's Fund. * p .05. *** p .001.
1280
REED AND AQUINO
Table 4 Study 3: Results of Multiple Regression Analysis and Logistic Regression on Allocation of Monetary Donations to UNICEF
Dependent variable Absolute $ to UNICEFa Independent variable Age White Female Moral reasoning Internalization Symbolization American identity B 0.02 0.28 0.12 0.00 0.60 0.13 0.44 SE 0.02 0.23 0.24 0.01 0.17 0.20 0.11 t 1.25 1.23 0.49 0.29 3.49*** 0.67 3.92*** B 0.03 0.73 0.23 0.01 1.38 0.14 0.66 Preference for UNICEFb SE 0.05 0.58 0.61 0.02 0.66 0.52 0.31 .17; Nagelkerke R2 Wald 0.46 1.61 0.14 0.25 4.36* 0.07 4.49* .24.
2
Note. UNICEF United Nations Children's Fund. a Adjusted R2 .22; F(7, 68) 3.95*. b Model 2 14.22*; Cox and Snell's R2 * p .05. *** p .001.
ican identity), the odds of giving a greater portion of money to the out-group increased (decreased) by 298% (48%). Again, these results are consistent with the argument that the self-importance of moral identity can overcome the tendency to favor the in-group, even during times of intergroup conflict. Indeed, they show that high moral identifiers were actually more likely to favor the outgroup, at least with regard to providing direct monetary assis- tance in this experimental context.
Study 4: Moral Identity and Revenge and Forgiveness Toward Out-Groups
As the United States waged its military campaign in Afghanistan to capture or kill those responsible for masterminding the terrorist attacks, one of the issues often debated in the public arena was how many innocent civilian casualties (presumably Afghani) the American people should be willing to accept. There have been varying public reactions. For example, not all Americans were equally disturbed by incidents like the one in which many innocent Afghani wedding attendees were killed when U.S. military planes construed celebratory rifle fire as an antiaircraft attack (Graham, 2002). Following the theoretical arguments in this article, it is proposed that a self-important moral identity can partly explain differences in the willingness to accept such civilian casualties. Specifically, because people whose moral identity has high selfimportance are more likely to view a wider set of out-groups as deserving moral concern, high moral identifiers should be less willing to accept the possibility of large numbers of civilian casualties following a military counterresponse to the terrorist attacks. In contrast, because low moral identifiers are likely to restrict their circle of moral regard, they may be more willing to accept considerable numbers of civilian deaths that are closely associated with a hostile out-group. On the basis of this reasoning, we hypothesized that a highly self-important moral identity would be negatively related to the willingness to accept civilian deaths as a direct result of a military attack against Afghanistan. The above hypothesis deals with people's reactions toward outgroup members who may be viewed as necessary, albeit undeserving, victims of military retaliation. This begs the question of whether the self-importance of moral identity might influence people's reactions toward out-group members held directly responsible for the September 11th attacks. If the effect of a highly self-important moral identity on the circle of moral regard is sufficiently robust and powerful, it is plausible that even out-group members who are viewed as having caused harm may be considered deserving of some moral concern. Examining the relationship between the self-importance of moral identity and two key retal-
Discussion and Limitations
Studies 2 and 3 focused on how the self-importance of moral identity relates to reactions to out-groups in the aftermath of intergroup conflict. These positive attitudes, intentions, and behaviors toward the out-group appear to be a consequence of an expanded circle of moral regard that is associated with having a highly self-important moral identity. However, it is clear that one of the strongest reactions to the attacks on the World Trade Center and Pentagon is to retaliate and punish the perpetrators and those who assist or harbor them. This was ostensibly the reason for the United States government sending troops into Afghanistan to hunt down the terrorists and for the continued United States campaign against Al-Qaeda and its leaders to this day. In the wake of these retaliatory responses toward out-groups, a question that was not addressed by the three previous studies is whether the circle of moral regard hypothesis applies not only to positive responses (i.e., perceived worthiness of assistance efforts, providing actual aid) but also to negative responses toward out-groups. Logically, an expanded circle of moral regard should be negatively associated with the willingness to inflict harm and injury on out-groups, even when these out-groups may be perceived as responsible for harming oneself or one's in-group. Study 4 tested this hypothesis by examining whether the self-importance of moral identity predicts (a) the willingness to accept civilian deaths among out-group members in retaliation for the attack against the United States and (b) the perceived morality of two hypothetical response options that vary in their level of punitiveness, which might be taken by the United States against the perpetrators.
MORAL IDENTITY AND OUT-GROUPS
1281
iatory reactions toward the September 11th terrorist attackers tests this hypothesis.
Revenge and Forgiveness Toward Harm Doers
A natural response to having been harmed by another person or group is to seek revenge against the harm doer. Revenge has been justified on moral grounds because it balances the calculus of justice by restoring equity in the relationship between victim and offender (e.g., Hogan & Emler, 1981; Walster, Walster, & Berscheid, 1973). In the case of the September 11th attacks, thousands of lives were lost, the economic damages to the American economy were substantial, and the American "way of life" was perceived as having been threatened by a hostile out-group. Consequently, massive military retaliation directed against the perpetrators of the attack was seen by many Americans as a practical and moral imperative. Yet again, there were some Americans who believed that less punitive means should be pursued in response to the terrorist attacks. Study 4 focused on two possible responses to harm doing— revenge and forgiveness—that lie at extreme ends along a continuum of punitiveness. Revenge has been defined as an action taken in response to a perceived harm or wrongdoing by another party. Revenge is intended to inflict harm, damage, injury, discomfort, or punishment on the party judged responsible for causing the harm (Aquino, Tripp, & Bies, 2001; Stuckless & Goranson, 1992). Forgiveness has been defined as a deliberate decision by the victim of a harmful action to relinquish anger, resentment, and the desire to punish the party held responsible for inflicting harm (Aquino et al., 2001; Enright & The Human Development Study Group, 1991; North, 1987). By forgiving the offender, the victim essentially cancels the offender's "debt" and surrenders the "right" to exact punishment or seek revenge (Enright & the Human Development Study Group, 1991). This surrendering of justice by restoring equity is what forgiveness demands, although this does not mean that the offender is not held responsible for what he or she did or does not suffer some legal consequence (e.g., imprisonment; Enright & The Human Development Study Group, 1991). What forgiveness does mean is that the victim overcomes anger, hatred, and the desire to personally injure the harm doer. It is for this reason that forgiveness is typically viewed as the antithesis of revenge. It is argued that because these two responses to wrongdoing represent opposite poles along a dimension of punitiveness, examining people's beliefs about the moral appropriateness of these two actions provides a further test of the circle of moral regard hypothesis. If the self-importance of moral identity expands this circle, and if we assume that a defining characteristic of showing moral regard is the desire to minimize the suffering and harm of another human being, then it is reasonable to propose that high as compared with low moral identifiers should view a less punitive response to the perpetrators of the terrorist attacks on the World Trade Center and Pentagon as more morally acceptable. The reason for this is that the less punitive the response, the less the suffering of persons included within the circle of moral regard. Consistent with this argument, Westman and Lewandowski (1992) found that a stronger belief in the unity of humanity was negatively related to approval of war as a strategy.
Following this logic, Study 4 presented participants with two response options for dealing with the perpetrators of the World Trade Center and Pentagon attacks to see whether the selfimportance of moral identity influences judgments of the morality of (a) using any means necessary to kill those responsible or (b) forgiving those responsible by replacing negative emotions like hatred and anger with positive emotions like compassion and love. It was predicted that although a highly self-important moral identity would be negatively related to the perceived morality of Option a, it would be positively related to the perceived morality of Option b. The reason being that people whose moral identity has high self-importance and whose circle of moral regard is thereby more expansive, would be least likely to endorse the option which inflicts the most harm (death) on others. Conversely, high moral identifiers would be considerably more likely than low moral identifiers to endorse the morality of forgiving the perpetrators (Option b) based on the assumption that even the perpetrators are more likely to be included within the circle of moral regard.
Method Participants
One hundred six undergraduate students from a university located in the northeastern United States participated in the exercise for class credit. Average age was 20.6 years (SD 1.40). Forty-eight were men, 68 were women. Of those who reported race, 55 were White, 8 were African American, 29 were Asian, 3 were Hispanic, and 3 self-identified as other. Eighty-five participants provided usable data on all study variables.
Procedure
The exercise consisted of two time periods separated by 2 months. At Time 1, participants completed a consumer opinions questionnaire. At Time 2, participants completed a computer exercise intended to measure reactions toward historical events. They were told that the purpose of the Time 2 study was to gain a comprehensive understanding of how people respond to and remember major events. Pictures were taken from various online media regarding the World Trade Center attacks. Some of these pictures were pictures of devastation (e.g., the burning towers), others were pictures of sympathy (e.g., the outpouring of international support from other countries). The pictures were shown on a computer screen in a random order. Participants were then asked questions described later. The confidentiality of students' responses was strictly maintained by assigning code numbers at both time periods. After data collection was completed, participants were fully debriefed and dismissed.
Independent Variables
As in the previous studies, the 10-item Self-Importance of Moral Identity measure was administered at Time 1 along with several other measures. Using a 7-point scale ranging from 1 (low) to 7 (high), we calculated average scores on both the Internalization dimension (M 5.31, SD 0.53, .85) and the Symbolization dimension (M 2.96, SD 0.71, .78), and entered them into the analysis as an index of activation potential of a moral self.
Control Variables
Gender (0 male, 1 female), race (0 non-White, 1 White) and American identity, M 3.42 (SD 1.07); ( .91), were used as control variables.
1282 Dependent Variables
REED AND AQUINO
Participants responded to measures after viewing the pictures from September 11th. Participants were told to assume that a military strike against the persons held responsible was imminent and would result in civilian deaths among persons who had nothing to do with the attack. They were then asked to type in the number of (innocent) civilian deaths that would be morally acceptable in such a retaliatory attack. The text box for this response variable had exactly nine digits, so participants could have inputted a theoretical range of from 0 to 999999999. Second, using a 5-point Likert scale (1 extremely immoral, 5 extremely moral), participants assessed the moral appropriateness of the two responses described above: (a) use any means necessary to kill those responsible for these acts, (b) forgive those responsible meaning negative emotions like hatred and anger should be replaced with positive emotions like compassion and love.
means necessary) and forgiveness (showing compassion) as repeated measures and submitted them to a multivariate regression analysis that included gender, race, internalization, symbolization, and American identity as predictor variables. Gender and race had no effect and, because of the small size of the sample for Study 4, were deleted from the model to preserve degrees of freedom. Results showed a main effect of response, F(1, 69) 18.69, p .01, such that participants were more likely to perceive killing the perpetrators of the terrorist attacks as more moral (M 2.73) than forgiveness and compassion (M 2.42). However, this main effect was qualified by two significant interactions. There was a significant Response (revenge vs. forgiveness) Internalization interaction, F(1, 69) 3.70, p .05, and a significant Response Symbolization interaction, F(1, 69) 6.42, p .05. No other effects were significant. To examine the nature of these withinsubject effects, the zero-order correlations between internalization, symbolization, revenge (killing by any means necessary), and forgiveness (showing compassion) were examined. Consistent with predictions, internalization was negatively correlated with perceptions of morality of killing the perpetrators (r .37, p .01) and positively correlated with perceptions of morality of forgiving the perpetrators (r .35, p .01). Similarly, symbolization was negatively correlated with perceptions of the morality of seeking revenge by killing the perpetrators (r .30, p .01), although it showed no relationship to the perceived morality of forgiving the terrorist attackers ( r .15, ns).
Results
Correlations among all independent and dependent variables are shown in Table 5.
Number of Acceptable Civilian Deaths
A multiple regression analysis was performed to predict the number of acceptable civilian deaths as a function of the selfimportance of moral identity. The model included gender, race, internalization, symbolization, and American identity as independent and control variables. Results showed that the internalization dimension, F(1, 73) 3.87, p .05, 37, p .05, and the symbolization dimension, F(1, 73) 6.12, p .05, 25, p .05, of moral identity were both associated with reported acceptable collateral damage. Those who had a relatively stronger self-importance of moral identity as indexed by the two dimensions reported a lower number of acceptable civilian casualties. No other effects were significant.
General Discussion
Social psychologists have studied several approaches in the hope of reducing intergroup hostility and bias (cf. Hewstone et al., 2002). A review of these approaches provides insight into the cognitive and motivational processes that might interfere with what seems to be humanly universal, that is, the often automatic (cf. Greenwald, McGhee, & Schwartz, 1998; Wilson et al., 2000) tendency for people to favor in-groups and to negatively respond to out-groups. One effective approach to reducing negative responses to out-groups is to encourage people to bring their prejudices to mind and to then invoke self-directed guilt derived from the discrepancy between personal values and actual behavior (Devine, Plant, & Blair, 2001). Another approach seeks to suppress out-group hostility by increasing the salience of positive values (e.g., tolerance) and making people rationalize or account for their
Moral Appropriateness of Revenge and Forgiveness
It was argued that because less punitive responses are more likely to minimize the suffering of those persons who are included within the circle of moral regard, then high as compared with low moral identifiers would view a less punitive response to the perpetrators of the terrorist attacks on the World Trade Center and Pentagon as more morally acceptable. To test this hypothesis, we treated the two hypothetical responses of revenge (killing by any
Table 5 Study 4: Means, Standard Deviations, and Correlations Among Independent and Dependent Variables
Variable 1. 2. 3. 4. 5. 6. 7. 8. †p Gender Race Internalization Symbolization American identity Acceptable civilian deaths Kill Forgiveness .10. *p .05. *** p M SD 1 — .04 .10 .05 .04 .10 .05 .01 2 — .06 .12 .04 .10 .05 .00 3 4 5 6 7 8
5.31 2.96 3.42 1,826 2.73 2.42 .001.
0.53 0.71 1.07 11,073 1.33 1.41
— .45*** .08 .34*** .37*** .35***
— .08 .18† .30* .15
— .06 .14 .23†
— .18† .18†
— .63***
—
MORAL IDENTITY AND OUT-GROUPS
1283
negative reactions (Dobbs & Crano, 2001). Researchers have also examined approaches whereby out-group hostility is reduced by using dissonance principles. These approaches assume that promoting nonbiased behavior (Leippe & Eisenstadt, 1994) and increasing the quality and quantity of intergroup contact (Pettigrew, 1998) will produce attitudinal realignment as an outcome of dissonance reduction. What the intervention strategies described earlier share is that they are primarily cognitive and consequentialist, meaning their effectiveness depends on adequate levels of awareness, effort, and practice over time (Hewstone et al., 2002). Moreover, many of the aforementioned interventions are inhibitory in the sense that they attempt to suppress implicit and existing out-group hostility by getting people to undo or relearn them in some way. A complementary explanation for why some people may be less susceptible to in-group favoritism and out-group hostility, however, is that their moral identity has high self-importance. As a result, they mentally redefine the boundaries of their in-group. The purpose of this article was to test this theoretical premise referred to as the circle of moral regard hypothesis: The self-importance of a person's moral identity, conceptualized as the degree to which one's self-conception is privately and publicly organized around moral trait associations (Aquino & Reed, 2002), directly influences the self- other relation. Several psychological outcomes associated with an expanded circle of moral regard were investigated, and the findings of four studies provide preliminary but convergent evidence supporting the theorized link between a highly selfimportant moral identity and the expansion of the circle of moral regard. Evidence from prior research suggests that out-group negativity is often reduced by fostering perspective taking (cf. Levy, Freitas, & Salovey, 2002) or improving attitudes toward former out-group members by "recategorizing" them from out- to in-group (Hewstone et al., 2002). It appears that the self-importance of moral identity might represent a self-regulatory mechanism that affects how a person cognitively constructs the psychological boundaries that define "us" and "them." Data from Study 1 suggest that the extent to which a person possesses a strong internalized moral identity will be reflected in stronger self-reported moral obligation to help various conceptions of out-groups and a desire to exchange highly personalized resources with them. This self-reported moral regard extended to out-groups was also reflected in secondary judgments regarding relief efforts. Data from Study 2 showed that individuals who had a strong internalized moral identity were more likely to favorably perceive the worthiness of relief efforts to assist out-groups, even when those out-groups were loosely associated with a highly salient intergroup conflict. This is an especially interesting finding because prior research has shown that during times of intergroup conflict, broad-based stereotypes of out-group members often trigger hostile expectations of interactions wrought with conflict (e.g., Stephan et al., 1998). Yet, even in the aftermath of one of the most disturbing domestic intergroup conflicts in 60 years, such negative reactions did not unfavorably affect judgments toward out-group relief efforts when the internalized selfimportance of moral identification was high. A more rigorous test of the circle of moral regard hypothesis was examined in Study 3. Results suggested that a strong internalized moral identity was associated with actual donation behav-
ior that reflected preference for the out-group. When forced to choose between giving more monetary assistance to either a deserving in-group (New York Police and Fire Widows and Children's Benefit Fund) or a deserving out-group (Afghani women and children), participants with a high internalized moral identity chose to extend assistance to the more socially distant out-group. These results provide the most direct evidence for the circle of moral regard hypothesis because they involve real behavior that emerged even when controlling for a powerful internal force (i.e., the opposing American identity) that seems likely to motivate the contraction rather than the expansion of the circle of moral regard.5 Finally, it appears that the psychological outcomes of an expansive circle of moral regard apply not only to positive actions toward the out-group but also to negative reactions toward out-groups that are loosely associated with a salient intergroup conflict and even to those held directly responsible for causing harm to the in-group. Study 4 showed that people who had both a high internalized moral identity and a high symbolized moral identity were less likely to accept harming innocent out-group members as a result of military retaliation (cf. Westman & Lewandowski, 1992). Ironically, those who have a particularly strong moral identity can be more sensitive to the aggressor's intentions in their perceptions of hostility (cf. Berkowitz, Mueller, Schnell, & Padberg, 1986), and yet they may also be more inclined to temper their desire for retaliation in the wake of hostility. This interpretation is based on the results from Study 4 showing that low as compared with high moral identifiers were less likely to judge killing the perpetrators of intergroup conflict as an appropriate moral response and more likely to judge forgiveness as morally appropriate. Although the data from the four studies reported here provide preliminary evidence for the link between the self-importance of moral identity and the psychological outcomes associated with an expanded circle of moral regard, several limitations of the studies should be noted. With the exception of two findings, the internalization dimension of moral identity was a stronger predictor of the criterion variables. One possible reason for this result is that even though the people in the samples presented here had relatively strong internalized senses of their moral identity, there may be other factors that impede their motivation, ability, or opportunity to express that moral identity publicly. Alternatively, it could be that people symbolically express their moral identities in ways other than what was measured with the items used to tap the
5 It should be noted that these results do not imply that if a person has a strong American identity, then they are immoral. Two points should be made: First, an examination of the correlation between American identity, internalization, and symbolization shows that they were slightly positively correlated in Study 3 (see Table 3). Second, Aquino and Reed (2002) have already made the point that individuals who score relatively lower on the 10-item measure of the self-importance of moral identity are not necessarily immoral individuals. For example, examination of the means for internalization and symbolization across all of the studies reported here suggest that people self-report relatively high levels (above the midpoint of the scale) on the two subscales. Therefore, a more appropriate interpretation at least in these data and in the data from Aquino and Reed is that people who score relatively higher on the subscales simply have a self-concept that tends to be relatively more morally schematic (cf. Bem, 1981).
1284
REED AND AQUINO moral actions that differ in self investment. Manuscript submitted for publication. Aquino, K., Tripp, T. M., & Bies, R. J. (2001). How employees respond to personal offense: The effects of blame attribution, victim status, and offender status on revenge and reconciliation in the workplace. Journal of Applied Psychology, 86, 52-59. Bem, S. L. (1981). Gender schema theory: A cognitive account of sex typing. Psychological Review, 88, 354 -364. Berkowitz, M. W., Mueller, C. W., Schnell, S. V., & Padberg, M. T. (1986). Moral reasoning and judgments of aggression. Journal of Personality and Social Psychology, 51, 885- 891. Blasi, A. (1984). Moral identity: Its role in moral functioning. In W. Kurtines & J. Gewirtz, (Eds.), Morality, moral behavior and moral development (pp. 128 -139). New York: Wiley. Blieszner, R. (1993). Resource exchange in the social networks of elderly women. In U. G. Foa, J. Converse Jr., K. Turnblom, & E. B. Foa (Eds.), Resource theory: Explorations and applications (pp. 67-79). San Diego, CA: Academic Press. Brewer, M. B., & Gardner, W. (1996). Who is this "We"? Levels of collective identity and self representations. Journal of Personality and Social Psychology, 71, 83-93. Brewer, M. B., & Kramer, R. M. (1985). The psychology of intergroup attitudes and behavior. Annual Review of Psychology, 36, 219 -243. Brown, J. D., & Gallagher, F. M. (1992). Coming to terms with failure: Private self-enhancement and public self-effacement. Journal of Experimental Social Psychology, 28, 3-22. Center for the Study of Ethical Development. (1993). Guide for the Defining Issues Test. Minneapolis: University of Minnesota. Damon, W. (1984). Self-understanding and moral development from childhood to adolescence. In W. Kurtines & J. Gewirtz (Eds.), Morality, moral behavior and moral development (pp. 109 -127). New York: Wiley. Dart, B. (2001, December 5). Kindness honored in wake of attacks: Groups helped Muslims, Arabs. The Atlanta Journal-Constitution. Retrieved December 5, 2001, from http://www.accessatlanta.com/ajc/ Deaux, K. (2002) Negotiating identity and community after September 11. Retrieved March 5, 2003, from the Graduate Center at the City University of New York: Social Science Research Council: http://www .ssrc.org/sept11/essays/deaux.htm Devine, P. G. (1989). Stereotypes and prejudice: Their controlled and automatic components. Journal of Personality and Social Psychology, 56, 5-18. Devine, P. G., Plant, E. A., & Blair, I. V. (2001). Classic and contemporary analyses of racial prejudice. In R. Brown & S. Gaertner (Eds.), Blackwell handbook of social psychology: Intergroup processes (pp. 198 -217). Malden, MA: Blackwell. Dobbs, M., & Crano, W. D. (2001). Outgroup accountability in the minimal group paradigm: Implications for aversive discrimination and social identity theory. Personality and Social Psychology Bulletin, 12, 355365. Dovidio, J. F., & Gaertner, S. L. (1991). Changes in the expression and assessment of racial prejudice. In H. J. Knopke, R. J. Norrell, & R. W. Rogers (Eds.), Opening doors: Perspectives on race relations in contemporary America (pp. 119 -148). Tuscaloosa: University of Alabama Press. Dovidio, J. F., & Gaertner, S. L. (1993). Stereotypes and evaluative intergroup bias. In D. M. Mackie & D. L. Hamilton (Eds.), Affect, cognition, and stereotyping (pp. 167-193). San Diego, CA: Academic Press. Eisenberg, N. (2000). Emotion, regulation, and moral development. Annual Review of Psychology, 51, 665- 697. Enright, R. D., & The Human Development Study Group. (1991). The moral development of forgiveness. In W. Kurtines & J. Gewritz (Eds.),
symbolization dimension. Future research should examine conditions under whether the private (internalization) or public (symbolization) aspects of the self-importance of a person's moral identity will be stronger predictors of particular outcomes. One plausible hypothesis is that the symbolization dimension might be more predictive of overt, public behaviors. In contrast, internalization might be less affected by the presence of an audience. Another limitation of the studies in this article is that the processes that directly link the self-importance of moral identity to psychological and cognitive functioning were not explored. It should be noted, though, that this study is preliminary, and it uses one of the first validated measures of the self-importance of moral identity. Moreover, the goal of this article was to broadly investigate the relationship between the self importance of moral identity and the psychological outcomes of an expanded circle of moral regard. Future work should examine the micropsychological processes that account for this relationship. Finally, none of the data in this article permits strong causal inferences. In all of the studies, the self-importance of moral identity was related to other outcomes through correlational rather than experimental analysis. Although many other variables were controlled, the possibility remains that one or more unmeasured variables may systematically covary with the dimensions of moral identity's self-importance and the outcomes measured in the four studies. To make stronger causal inferences, future research might attempt to manipulate the selfimportance of a person's moral identity and systematically test its relationship to moral outcomes.
Conclusions
Emily Dickinson once observed that human beings select their own societies and then "shut the door" to the claims and needs of those who fall outside the circle of their moral regard. The natural propensity to distinguish "us" from "them" has been at the forefront of work on intergroup conflict. It has long been recognized that "some [people] may consider their morality to be central to their self-identities, whereas others may consider it to be peripheral" (Damon, 1984, p. 110). However, very little empirical research has systematically examined the relationship between the self-importance of moral identity and its cognitive and behavioral implications for the construction of psychological group boundaries. Consequently, this article makes two important contributions to the literature. First, it refines and further extends a definition of moral identity grounded in theories of social identity and the selfconcept. Second, it provides evidence for the relationship between the self-importance of moral identity and various reac- tions toward out-groups. The data suggest that the concept of a moral self may collapse the mental barriers separating in-groups and out-groups, thus highlighting the possibility that in some "simple, perfectly impenetrable" way, we are all brothers.
References
Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and interpreting interactions. Thousand Oaks, CA: Sage. Aquino, K. F., & Reed, A., II. (2002). The self-importance of moral identity. Journal of Personality and Social Psychology, 83, 1423-1440. Aquino, K. F., & Reed, A., II. (2003). To thine own selves be true: The effects of moral identity and organizational status on the preference for
MORAL IDENTITY AND OUT-GROUPS Handbook of moral behavior and development (Vol. 1, pp. 123-152). Hillsdale, NJ: Erlbaum. Erikson, E. H. (1964). Insight and responsibility. New York: Norton. Esses, V. M., Haddock, G., & Zanna, M. P. (1993). Values, stereotypes, and emotions as determinants of intergroup attitudes. In D. M. Mackie & D. L. Hamilton (Eds.), Affect, cognition, and stereotyping: Interactive processes in group perception (pp. 167-193). San Diego, CA: Academic Press. Fein, S., & Spencer S. J. (1997). Prejudice as self-image maintenance: Affirming the self through derogating others. Journal of Personality and Social Psychology, 73, 31- 44. Feldman, J., & Lynch, J. G. (1988). Self-generated validity and other effects of measurement on belief, attitude, intention and behavior. Journal of Applied Psychology, 73, 421- 435. Foa, E. B., & Foa, U. G. (1974). Societal structures of the mind. Springfield, IL: Charles C Thomas. Foa, E. B., & Foa, U. G. (1980). Resource theory: Interpersonal behavior in exchange. In K. J. Gergen, M. S. Greenberg, & R. H. Willis (Eds.), Social exchange: Advances in theory and research (pp. 77-101). New York: Plenum Press. Forehand, M., Deshpande, R., & Reed, A., II. (2002). Identity salience and ´ the influence of differential activation of the social self-schema on advertising response. Journal of Applied Psychology, 87, 1086 -1099. Gaertner, S. L., & Insko, C. A. (2000). Intergroup discrimination in the minimal group paradigm: Categorization, reciprocation, or fear? Journal of Personality and Social Psychology, 79, 77-94. Gibbons, F. X., & Gerrard, M. (1991). Downward social comparison and coping with threat. In J. M. Suls & T. A. Wills (Eds.), Social comparison: Theory and research (pp. 317-345). Hillsdale, NJ: Erlbaum. Graham, B. (2002, July 9). Inquiry set in Afghans' deaths in air-strike: Pentagon still disputes number of casualties. Washington Post, p. A17. Greenwald, A. G., & Farnham, S. D. (2000). Using the Implicit Association Test to measure self-esteem and self-concept. Journal of Personality and Social Psychology, 79, 1022-1038. Greenwald, A. G., McGhee, D. E., & Schwartz, J. L. K. (1998). Measuring individual differences in implicit cognition: The Implicit Association Test. Journal of Personality and Social Psychology, 74, 1464 -1480. Hamilton, D. L., & Trolier, T. K. (1986). Stereotypes and stereotyping: An overview of the cognitive approach. In J. F. Dovidio & S. L. Gaertner (Eds.), Prejudice, discrimination, and racism (pp. 127-163). Orlando, FL: Academic Press. Hart, D., Atkins, R., & Ford, D. (1998). Urban America as a context for the development of moral identity in adolescence. Journal of Social Issues, 54, 513-530. Hart, D., & Fegley, S. (1995). Prosocial behavior and caring in adolescence: Relations to moral judgment and self-understanding. Child Development, 66, 1346 -1359. Hewstone, M., Rubin, M., & Willis, H. (2002). Intergroup bias. Annual Review of Psychology, 53, 575- 604. Hogan, R., & Emler, N. P. (1981). Retributive justice. In M. J. Lerner & S. C. Lerner (Eds.), The justice motive in social behavior (pp. 125-143). New York: Plenum. Hume, D. (1751/1996). An inquiry concerning the principles of morals. LaSalle, IL: Open Court. Kihlstrom, J. F., & Klein, S. B. (1994). The self as a knowledge structure. In R. S. Wyer Jr. & K. Thomas (Eds.), Handbook of social cognition: Vol. 1. Basic processes: Vol. 2. Applications (2nd ed.; pp. 153-208). Hillsdale, NJ: Erlbaum. Kohlberg, L. (1971). From is to ought: How to commit the naturalistic fallacy and get away with it in the study of moral development. In T.
1285
Mischel (Ed.), Cognitive development and epistemology (pp. 164 -165). New York: Academic Press. Lapsley, D. K., & Lasky, B. (2001). Prototypic moral character. Identity, 1, 345-363. Laughlin, H. P. (1970). The ego and its defenses. New York: AppletonCentury-Crofts. Leippe, M. R., & Eisenstadt, D. (1994). Generalization of dissonance reduction: Decreasing prejudice through induced compliance. Journal of Personality and Social Psychology, 67, 395- 413. Levy, S. R., Freitas, A. L., & Salovey, P. (2002). Construing action abstractly and blurring social distinctions: Implications for perceiving homogeneity among, but also empathizing with and helping, others. Journal of Personality and Social Psychology, 83, 1224-1238. Markus, H. (1977). Self-schemata and processing information about the self. Journal of Personality and Social Psychology, 35, 63-78. Mikulincer, M., & Shaver, P. R. (2001). Attachment theory and intergroup bias: Evidence that priming the secure base schema attenuates negative reactions to out-groups. Journal of Personality and Social Psychology, 81, 97-115. Monroe, K. R., & Epperson, C. (1994). "But what else could I do?" Choice, identity, and cognitive perceptual theory. Political Psychology, 15, 201226. North, J. (1987). Wrongdoing and forgiveness. Philosophy, 6, 499 -508. Oliner, S. P., & Oliner, P. M. (1988). The altruistic personality: Rescuers of the Jews in Nazi-occupied Poland. New York: Free Press. Pettigrew, T. F. (1998). Intergroup contact theory. Annual Review of Psychology, 49, 65- 85. Pratto, F., Sidanius, J., & Stallworth, L. M. (1993). Sexual selection and the sexual and ethnic basis of social hierarchy. In L. Ellis (Ed.), Social stratification and socioeconomic inequality: A comparative biosocial analysis (pp. 111-137). New York: Praeger. Quillian, L. (1995). Prejudice as a response to perceived group threat: Population composition and anti-immigrant and racial prejudice in Europe. American Sociological Review, 60, 586 - 611. Reed, A., II. (2002). Social identity as a useful perspective for self-conceptbased consumer research. Psychology and Marketing, 19, 235-266. Reed, A., II, Aquino, K. F., & Ray, S. (2003). Violent media imagery and the September 11th terrorist attacks: Moral identity moderates the relationship between empathy and moral disengagement. Manuscript submitted for publication. Reed, A., II, Wooten, D. B., & Bolton L. (2002). The temporary construction of consumer attitudes. Journal of Consumer Psychology, 12, 375388. Rest, J. (1979). Development in judging moral issues. Minneapolis: University of Minnesota Press. Schlenker, B. R. (Ed.). (1985). The self in social life. New York: McGrawHill. Sidanius, J., Levin, S., Liu, J., & Pratto, F. (2000). Social dominance orientation, anti-egalitarianism and the political psychology of gender: An extension and cross-cultural replication. European Journal of Social Psychology, 30, 41- 67. Singer, P. (1981). The expanding circle: Ethics and sociobiology. New York: Farrar, Straus, & Giroux. Steele, C. M. (1988). The psychology of self-affirmation: Sustaining the integrity of the self. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 21, pp. 261-302). New York: Academic Press. Steele, C. M., & Liu, T. J. (1983). Dissonance processes as selfaffirmation. Journal of Personality and Social Psychology, 45, 5-19. Steele, C. M., Spencer, S. J., & Lynch, M. (1993). Self-image resilience and dissonance: The role of affirmation resources. Journal of Personality and Social Psychology, 64, 885- 896. Stephan, W. G., & Stephan, C. (1993). Cognition and affect in stereo-
1286
REED AND AQUINO Turner, J. L., Foa, E. B., & Foa, U. G. (1971). Interpersonal reinforcers: Classification, interrelationships, and some differential properties. Journal of Personality and Social Psychology, 19, 168 -180. Turner, J. C., Hogg, M. A., Oakes, P. J., Reicher, S. D., & Wetherell, M. S. (1987). Rediscovering the social group: A self-categorization theory. Oxford, England: Blackwell. Walker, L. J., & Pitts, R. C. (1998). Naturalistic conceptions of moral maturity. Developmental Psychology, 34, 403- 419. Walster, E., Walster, G., & Berscheid, E. (1978). Equity: Theory and research. Boston: Allyn & Bacon. Westman, A. S., & Lewandowski, L. M. (1992). How empathy, egocentrism, Kohlberg's moral development, and Erikson's psychosocial development are related to attitudes toward war. Psychological Reports, 69(Pt. 2), 1123-1127. Wilson, T. D., Lindsey, S., & Schooler, T. T. (2000). A model of dual attitudes. Psychological Review, 107, 101-126. Younis, J., & Yates, M. (1999). Youth service and moral-civic identity: A case of everyday morality. Educational Psychology Review, 11, 361376.
typing: Parallel interactive networks. In D. M. Mackie & D. L. Hamilton (Eds.), Affect, cognition, and stereotyping: Interactive processes in group perception (pp. 111-136). Orlando, FL: Academic Press. Stephan, W. G., Ybarra, O., Martinez, C. M., Schwarzwald, J., & TurKaspa, M. (1998). Prejudice toward immigrants to Spain and Israel: An integrated threat theory analysis. Journal of Cross-Cultural Psychology, 29, 559 -576. Stryker, S. (1980). Symbolic interactionism: A social structural version. Menlo Park, CA: Benjamin/Cummings. Stuckless, N., & Goranson, R. (1992). The Vengeance Scale: Development of a measure of attitudes toward revenge. Journal of Social Behavior and Personality, 7, 25- 42. Tajfel, H. (1959). Quantitative judgment in social perception. British Journal of Psychology, 50, 16 -29. Tajfel, H., & Turner, J. C. (1979). An integrative theory of intergroup conflict. In W. G. Austin & S. Worchel (Eds.), The social psychology of intergroup relations (pp. 33- 47). Monterey, CA: Brooks-Cole. Tetlock, P. E. (1983). Accountability and the perseverance of first impressions. Social Psychology Quarterly, 46, 285-292.
Appendix The Self-Importance of Moral Identity Measure
Listed below are some characteristics that might describe a person: Caring, Compassionate, Fair, Friendly, Generous, Helpful, Hardworking, Honest, and Kind The person with these characteristics could be you or it could be someone else. For a moment, visualize in your mind the kind of person who has these characteristics. Imagine how that person would think, feel, and act. When you have a clear image of what this person would be like, answer the following questions. I I S I S S I S S I 1. It would make me feel good to be a person who has these characteristics. 2. Being someone who has these characteristics is an important part of who I am. 3. I often wear clothes that identify me as having these characteristics. 4. I would be ashamed to be a person who had these characteristics. (R) 5. The types of things I do in my spare time (e.g., hobbies) clearly identify me as having these characteristics. 6. The kinds of books and magazines that I read identify me as having these characteristics. 7. Having these characteristics is not really important to me. (R) 8. The fact that I have these characteristics is communicated to others by my membership in certain organizations. 9. I am actively involved in activities that communicate to others that I have these characteristics. 10. I strongly desire to have these characteristics. I Internalization; S Symbolization; R Reverse coded.
doc_640697534.docx