Abstract
Moral elevation has been shown to increase helping behavior. However, this might be due to a threatened moral self-image because people engage in a social comparison with a moral exemplar and conclude that their own moral integrity is inferior. Alternatively, feelings of elevation might provide a motivational impetus to act on one’s moral values. We provided participants with an opportunity to engage in self-affirmation, which was followed by an induction of moral elevation or a neutral control mood. Compared to the neutral mood, participants experiencing moral elevation showed higher levels of helping behavior following self-affirmation. This effect was especially pronounced in participants experiencing moral elevation who reminded themselves of previous prosocial behavior; they showed more helping than participants experiencing moral elevation who had not engaged in self-affirmation. Thus, rather than posing a threat to moral self-worth, feelings of elevation can provide the motivational trigger to act on affirmed moral values.
It is a rare person who does not think of himself or herself as morally upstanding and full of integrity. Even mass murderers, pedophiles, and other criminals seem to be able to convince themselves that they have positive moral qualities. Recent research has explored how people regulate their moral self-worth over time in the face of obvious moral transgressions (Zhong, Liljenquist, & Cain, 2009). Converging evidence suggests that on some implicit level people appear to keep track of their own moral behavior, as if they accumulate “moral credits” when doing something good, credits which they later “cash in” when doing something immoral. Thus, when people are reasonably sure of their moral integrity, they have little motivation to do further good, a phenomenon termed “moral licensing” (Mazar & Zhong, 2010; Merritt, Effron, & Monin, 2010; Monin & Miller, 2001; Sachdeva, Iliev, & Medin, 2009; Zhong et al., 2009). In the first clear demonstration of this effect, Monin and Miller (2001) gave some participants an opportunity to affirm their credentials as nonprejudiced people, for example, by disagreeing with sexist statements. In subsequent tasks, those participants exhibited more prejudicial opinions than participants who did not have the earlier opportunity of showing themselves to be unbiased. Similar effects have been demonstrated involving political attitudes, such that participants who in one task indicated a voting preference for Barak Obama indicated a positive prejudice toward Whites in a subsequent task (Effron, Cameron, & Monin, 2009).
Further research has shown that moral licensing effects occur when people merely think of concepts relating to a positive self-image (Sachdeva et al., 2009). Participants who copied positive trait words and used them in a story about themselves indicated they would donate less to a charity and cooperate less in an environmental task than participants who wrote a story involving neutral words. Presumably, thinking about their positive qualities affirmed participants’ moral integrity, and as a consequence, reduced any desire to engage in prosocial behavior. These findings suggest that making salient positive or negative traits can be enough to result in moral licensing effects; actually engaging in a moral action is not necessary. Although Sachdeva and colleagues (2009) did not explicitly interpret their findings in this way, their work can be viewed as applying the general principle of self-affirmation when facing threat (McQueen & Klein, 2006; Steele, 1988) to the moral domain. Indeed, a central consideration of self-affirmation theory is that people want to protect the integrity of their concept of self (Sherman & Cohen, 2006).
Such findings regarding moral licensing paint a rather pessimistic picture of human nature: On many occasions people may only act morally because it allows them to restore their own self-worth after it had been challenged. In other words, prosocial behaviors such as altruism might often be merely selfish, because once one feels like a morally superior person, there is no need to prove this fact any further. Sachdeva and colleagues (2009) end their article with the question “How does one avert the danger of moral licensing?” (p. 528) We propose a possible answer, one that involves eliciting elevation, a positive moral emotion.
Elevation describes the subjective experience when observing others perform acts of moral excellence such as kindness or heroism, namely behaviors that benefit others while incurring a cost to the person performing them (Haidt, 2003). When experiencing elevation, people describe feeling inspired and uplifted, and, importantly, having the motivation to become a better person (Haidt, 2003). People specifically report an urge of wanting to do good (Algoe & Haidt, 2009; Landis et al., 2009), and direct evidence for the link between elevation and helping behavior was provided by Schnall, Roper, and Fessler (2010). Participants watched an elevating clip from the “Oprah Winfrey Show” in which a music teacher mentored a young man from an impoverished background and played a major part in turning him away from a potential life of crime. Compared to participants who watched a neutral control clip or one eliciting general positive affect, participants who watched the elevation clip demonstrated a higher level of prosocial behavior, as indicated by their willingness to volunteer for further unpaid studies or to help an experimenter by completing a boring task. Similarly, Freeman, Aquino, and McFerran (2009) showed that elevation, induced by watching clips and reading stories about extraordinary acts of forgiveness, increased donations to a charity for the advancement of education of Black students among White individuals high in social dominance orientation, which is normally linked to racist attitudes. Further, people for whom moral identity is central to the self-concept report higher feelings of moral elevation when faced with moral excellence than people for whom moral identity is less important, and these feelings, in turn, predict the desire to engage in proscocial behavior (Aquino, McFerran, & Laven, 2011). All these findings suggest that moral elevation can have highly beneficial consequences because it inspires people to do good things.
However, in the light of the findings reviewed above concerning moral licensing (Mazar & Zhong, 2010; Merritt et al., 2010; Monin & Miller, 2001; Sachdeva et al., 2009; Zhong et al., 2009), it may be that the increased prosocial behavior observed after experiencing elevation is the result of feeling a threat to one’s moral self-worth. Monin (2007) proposes that observing exceptional moral behavior may trigger an upward social comparison, resulting in the conclusion that the perceiver’s own morality may be lacking. One response to this threat is to engage in a behavior that restores one’s self-worth, such as helping somebody in need. Thus, when people are given the opportunity to reaffirm their own moral value when feeling elevated, they might subsequently be less likely to help. Alternatively, when reminded of their own positive traits and moral convictions, feeling elevated might provide people with the motivation to act on their moral beliefs, thus empowering them to act in line with their values. Indeed, Crocker, Niiya, and Mischowski (2008) showed that self-affirmation produces other-directed emotions of self-transcendence, such as love and caring. They found that when writing an essay about their most treasured value, participants reported higher levels of love and empathy than when writing about a low-importance value. Because of its inherent motivational component to inspire people to do good, as an other-directed, in fact, “other-praising” emotion (Algoe & Haidt, 2009), elevation might provide a special opportunity to turn one’s core values into action.
The Current Research
This experiment tested the potential consequences of experiencing elevation following a reminder of positive aspects of one’s self-concept. Participants engaged in self-affirmation or a neutral control task and then were induced to experience either moral elevation or a neutral mood. Subsequently an opportunity to engage in helping behavior was presented that was ostensibly unconnected to the experiment itself. Two outcomes were considered possible: Following the logic of moral licensing (Sachdeva et al., 2009), participants experiencing moral elevation should only engage in increased helping behavior if they did not previously affirm their own positive qualities but not when they did. Alternatively, following the logic of self-affirmation as a reminder of core values (Crocker, Niiya, & Mischowski, 2008), relative to being in a neutral mood, when experiencing elevation with its inspirational effect to become a better person, participants should show increased helping behavior after having reminded themselves of their own positive qualities. If the second possibility was supported, the effect should be especially pronounced for participants who during the self-affirmation task considered their previous moral (i.e., prosocial) behavior rather than other positive qualities of a more selfish nature.
Method
Participants
Eighty female 1 participants recruited from the University of Plymouth participated as part of a psychology course requirement or for payment of £6. Data from three participants were excluded because of suspicion about the study purpose and from one additional participant because she failed to show up for an earlier time slot and consequently may have felt under some obligation to make it up to the experimenter.
Materials
Film clips for emotion induction
In the experimental condition, participants watched the 7-minute Oprah Winfrey clip previously used to induce elevation (Silvers & Haidt, 2008; Schnall, Roper, & Fessler, 2010). In the control condition, participants watched a 7-minute segment of “The Open Ocean” nature documentary by David Attenborough (1984), as previously used by Schnall et al. (2010).
Self-affirmation manipulation
Modeled after earlier studies (Cohen, Aronson, & Steele, 2000; Jordan & Monin, 2008), instructions for the “self-affirmation” condition were “Please write about a recent experience in which you demonstrated one of your most personally valued qualities and felt good about yourself. Examples of ‘personally valued qualities’ might include such things as artistic skills, sense of humor, spontaneity, athletic ability, musical talent, creativity or business skills.” Instructions for the “no self-affirmation” condition were “Please describe your typical journey from home to the university. Try to mentally walk through your route from door to door and be as detailed as you can in your description. Examples of what might be included are landmarks, street names, and length of time taken on each part of the route.”
Procedure
Participants were tested individually by a female experimenter and randomly assigned to one of four conditions, crossing elevation versus control emotion with self-affirmation versus no self-affirmation. As a cover study, participants were told that the research tested “memory for subjective versus objective experiences” and that they would complete one or more writing tasks and watch a film clip. After being seated at a computer cubicle they performed the first writing task for 5 min, which in fact manipulated self-affirmation. Then participants watched either the elevation or control emotion clip. 2 The remaining procedure followed Bartlett and DeSteno (2006) and Schnall et al. (2010, Experiment 2) and involved the next part of the experiment, a computer task, being rigged to fail, and the participant was told she was free to leave while still receiving full course credit or payment. Thus, the study that participants expected to last for 1 hr ostensibly ended after just about 15 minutes. Before the participant got ready to leave the room, however, the experimenter asked whether she would be willing to help her by completing a long and somewhat boring math questionnaire. If she agreed she was told to complete as much of the questionnaire as she wished, and that any amount of help would be greatly appreciated. Completion time of the questionnaire was secretly timed by the experimenter. Participants were then questioned regarding suspicions about the true nature of the experiment and were debriefed.
Results
Two sets of analyses were conducted. First we tested the basic notion that self-affirmation qualifies the effect of moral elevation. Additional analyses subsequently compared the self-affirmation conditions regarding the content of affirmed qualities that participants spontaneously produced, distinguishing between prosocial and selfish positive qualities.
A two-way between-subject analysis of variance (ANOVA) was conducted with minutes spent on the questionnaire as dependent variable and emotion (elevation vs. control) and self-affirmation (self affirmation vs. control) as independent variables. Replicating earlier finding (Schnall et al., 2010), there was a main effect of emotion condition, such that participants in the elevation condition (M = 34.34, SD = 14.32) spent significantly more time helping the experimenter by completing the questionnaire than participants in the neutral control condition (M = 27.52, SD = 13.56), F(1, 72) = 4.21, p = .04, ηp 2 = .06 (see Figure 1 ). No main effect was obtained for self-affirmation, F(1, 72) = .02, p = .89, but critically, there was an interaction between Emotion and Self-Affirmation, F(1, 72) = 4.05, p = .05, ηp 2 = .05. Planned comparisons showed that participants in the elevation condition who had completed the self-affirmation helped for longer (M = 37.14, SD = 14.54) than participants who had self-affirmed but were in the control condition (M = 24.30, SD = 12.77), F(1, 72) = 8.25, p = .005, d = .94. We further compared the elevation condition regarding whether there was a difference in self-affirmation. Although the means suggested that elevation following self-affirmation (M = 37.14, SD = 14.54) resulted in more helping than elevation without prior self-affirmation (M = 31.22, SD = 13.81), this planned comparison did not reach statistical significance, F(1, 72) = 1.75, p = .18, d = .42. However, this could be due to the self-affirmation instructions having led to a variety of qualities being recalled, with some relating to positive moral values, whereas others relating to other skills or abilities without a prosocial theme. Thus, we proceeded to conduct follow-up analyses that took into account the competence domain that participants had written about during the self-affirmation task.
Minutes spent completing boring questionnaire to help experimenter.
Self-affirmation content
Participants’ affirmation narratives were classified by two independent coders (who were blind to emotion condition) into whether they involved helping behavior, defined as another person benefiting from the actions of the participant (e.g., helping a friend who was ill), or were essentially selfish in nature, defined as descriptions of skills or behaviors that did not promote another person’s well-being (e.g., completing a jungle trek). The analysis above was broken down further by self-affirmation content, and planned comparisons were conducted (see Figure 2 for means). Participants who wrote about helping-related values spent significantly more time on the helping task if they were in the elevation emotion condition than if were in the control emotion condition, F(1, 70) = 9.06, p = .004, d = 1.48. Helping was not as high for participants who affirmed selfish skills and talents, reducing the effect between the elevation condition and the control emotion condition to being marginally significant, F(1, 70) = 3.01, p = .09, d = .72. Further, participants in the elevation condition who engaged in prosocial self-affirmation differed from those who did not engage in self-affirmation, F(1, 70) = 4.13, p = .05, d = −.84. Thus, the empowering effect of elevation following self-affirmation was especially pronounced when participants had reminded themselves of previous prosocial behaviors. In contrast, self-affirmation content did not matter for participants in the control emotion condition, whose relatively low level of helping did not differ for prosocial compared to selfish affirmations, F(1, 70) = .09, p = .76, d = .15.
Minutes spent completing boring questionnaire to help experimenter, grouped by self-affirmation content.
Discussion
The results of this experiment suggest that experiencing the emotion of elevation can propel people to put into action the core moral values after they had become salient to them: Participants who first reminded themselves of a positive quality that they possessed and then watched an elevating clip engaged in more helping behavior than participants who self-affirmed and then watched a clip with relatively neutral content. Further, when specifically taking into account the type of qualities that participants self-affirmed, we found that having affirmed previous prosocial behavior resulted in the highest level of helping behavior in combination with the feeling of moral elevation. Although it would have been equally plausible that helping behavior is reduced under conditions of self-affirmation and moral elevation, similar to the reported effects of moral licensing (Zhong et al., 2009), this was not the case.
Before we draw conclusions of our findings, several possible confounding factors need to be considered. First, although it would have been preferable to include a mood measure of elevation and other feelings as a manipulation check, the nature of the design made it difficult to assess mood at an appropriate time. In our previous work (Schnall et al., 2010), we did administer a manipulation check immediately after the mood induction, but this was less than optimal given that the manipulation check preceded the assessment of the dependent variable, namely participants’ helping behavior. Nevertheless, this earlier work did provide strong evidence that the elevation film clip led to significantly higher reports of feeling uplifted, moved, having a warm feeling in the chest, and wanting to become a better person than that of the neutral film clip, and thus we assume that the intended feelings were also successfully elicited in the current experiment. Second, one might object that because we only compared elevation to a control condition, we cannot rule out that general positive affect played a role. This was also an issue we addressed in detail in our earlier research (Schnall et al., 2010): The elevation and neutral conditions did not differ regarding reported happiness, and more importantly, whereas participants’ elevation-relevant feelings predicted subsequent helping, their happiness did not. Finally, a concern might be that the helping behavior might have been due to modeling or imitation rather than being the result of feelings of elevation. This possibility can be ruled out for two reasons: The helping behavior exhibited by participants differed qualitatively from the behavior displayed in the elevation-eliciting film clip, and further, effects on helping were qualified by previous self-affirmation.
Our findings point to two components of engaging in prosocial behavior in the context of one’s moral self-concept and moral elevation. First, rather than constituting a threat, the self-affirmation provided participants with the knowledge that they possess a particular positive quality (e.g., having helped others in the past) and thus made accessible evidence of their own ability to do good. Second, combined with this salience of past good deeds, momentary feelings of elevation provided the motivational impetus to act on this knowledge. Thus, our findings provide evidence for a moral empowerment effect of elevation. Indeed, Crocker et al. (2008) noted that self-affirmation is less of an affirmation of the self than it is an affirmation of values of the self.
We believe that the domain in which self-affirmation occurs is critical. Although self-affirmation theories (e.g., Steele, 1988) generally assume that the blow of a threat in one domain is lessened by the reminder of one’s positive self-worth in another, unrelated domain, in the context of elevation, a reminder of one’s values in the moral domain might serve as relevant evidence that one is capable of certain good actions (Sherman & Cohen, 2006). Indeed, the effect of elevation following self-affirmation was most notable in participants whose prior affirmation involved describing ways in which they had helped others, rather than self-centered skills or talents. These might have been participants for whom being a good, moral person was an especially central and salient part of self-concept, because they activated previous helping behaviors despite the study instructions nudging them to think of personal qualities such as artistic skills, athletic ability, or musical talent; in fact, none of the examples listed in the instructions were of any prosocial or social nature. Those participants who generated examples of previous moral behavior might find moral integrity as more central to their own self-concept than other participants; indeed, individual differences have been established regarding the extent to which people value being a moral person (Aquino & Reed, 2002; Aquino, McFerran, & Laven, 2011). In our study, some participants spontaneously activated prosocial values when asked to engage in self-affirmation, but future studies will need to determine whether people could be induced to do so, thus systematically encouraging desirable helping behaviors.
Overall, our findings suggest that being focally aware of one’s moral values while experiencing elevation can have the powerful effect of translating a person’s moral beliefs into action, thus shifting attention from one’s own moral self to those in need who stand to gain from it.
Footnotes
Notes
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and publication of this article.
The author(s) disclosed receipt of the following financial support for the research and authorship of this article: Grants RES-061-25-0119 and RES-000-22-4453 from the ESRC to Simone Schnall.