Deservingness and emotions: Testing a structural model that relates discrete emotions to the perceived deservingness of positive or negative outcomes
A study is described that tested a model (Feather in Eur Rev Soc Psychol 17:38–73, 2006) relating emotions to the appraisal of outcome deservingness for self or other person. Outcome deservingness was assumed to depend on the evaluative structure of action/outcome relations (Feather in Values, achievement, and justice: studies in the psychology of deservingness. Kluwer/Plenum Publishers, New York, 1999b). The study tested predictions about relations between this structure and the emotions of pleasure, admiration, pride, resentment, anger, sadness, sympathy, guilt, regret, disappointment, and surprise. The study used a hypothetical scenario involving an applicant for a position in an organization where the applicant could either be other or self. Results that focused on planned comparisons and the action by focus interactions supported the analysis for both the positive outcome and the negative outcome conditions and they were consistent with the hypothesis that the appraisal of outcome deservingness would mediate at least in part the type of emotion that was reported when a positive or negative outcome followed a positive or negative action. Results were discussed in relation to the social psychology of justice and the emotions.
KeywordsDeservingness Action/outcome evaluations Discrete emotions Self and other perspective
In this article we describe a study that investigated the relation between deservingness and discrete emotions. The study addresses a relatively neglected area in social psychology, namely how people feel about positive or negative outcomes when these outcomes are perceived to be deserved or undeserved. Thus it focuses on how the appraisal of a central justice variable, deservingness, influences reported emotions about outcomes.
The study built on a structural model developed by Feather (1999a, b) and later expanded so as to encompass emotions such as pleasure, sadness, guilt, regret, sympathy, resentment, and pride (Feather 2006; Feather and McKee 2009). These emotions are assumed to bear some relation to the appraisal of deservingness and also to whether the outcomes that are appraised as deserved or undeserved are generated by the actions of self or by another person. First, however, we turn to a brief review of deservingness theory.
The model developed by Feather (1999a, b) involves a set of deservingness structures that are assumed to be subject to pressures toward consistency or balance (Heider 1958). A key part of each structure is the evaluative structure of the relation between the outcome and the action that led to the outcome. Thus, positive outcomes that are produced by positive actions and negative outcomes that are produced by negative actions would both be perceived to be deserved. In contrast, positive outcomes that are produced by negative actions and negative outcomes that are produced by positive actions would both be perceived to be undeserved.
To give some examples: A student would be judged to deserve a high grade (a positive outcome) if he or she put in a lot of preparation and effort prior to an exam (a positive action), but the high grade would be judged as undeserved if the student did not work hard and slackened off before the exam (a negative action). A company director would be judged to deserve dismissal (a negative outcome) following his or her dishonest behavior (a negative action), but the dismissal would be perceived as undeserved if he or she had behaved properly and according to the interests of the company (a positive action).
Deservingness thus emerges as a judgment that depends on consistency or balance between the evaluations of actions and their contingent outcomes. Consistent evaluations generate perceived deservingness; inconsistent evaluations generate judgments that outcomes are undeserved. It should be clear that deservingness is linked to a person’s behavior, to what a person does to produce either a positive or negative outcome. In this respect it differs from a person’s entitlement to an outcome where the emphasis is more on perceived rights and legal or quasi-legal prescriptions (Feather 1999b, 2008b).
Judgments of deservingness or undeservingness are restricted in Feather’s model to situations where there is some degree of personal causation rather than an external locus of causation for the outcome that occurs. This means that the person is perceived to be wholly or partially responsible for the action that he or she performs and the outcome that follows from the action. People are not deemed to deserve outcomes for which they are not responsible. Such outcomes would be attributed to external factors beyond the person’s control rather than to a person’s own actions. However, because perceived deservingness depends crucially on the evaluated relation between an outcome and the action that produced it, deservingness and perceived responsibility do not always run in parallel. A person can be responsible for an outcome that is perceived to be deserved but also responsible for an outcome that is perceived as undeserved.
Deservingness and discrete emotions
The structural model of deservingness was recently extended so as to enable predictions about emotional reactions to deserved and undeserved positive or negative outcomes (Feather 2006). Consistent with the theory just presented, a key component in this extension is the evaluative structure of action/outcome relations for self or other which is assumed to determine how deservingness is appraised. The extension also allows for different sets of emotions depending on whether outcomes for self or for other person are the focus.
Discrete emotions relating to perceived deservingness/undeservingness and evaluated action/outcome relations for other and self
Focus of emotion
Schadenfreude (or Pleasure)
This form of analysis also allows for the occurrence of emotions that relate to the action or outcome alone. For example, Table 1 shows that self is predicted to feel some pleasure about a positive outcome, independent of the appraisal of deservingness (i.e., pleasure may occur for both deserved and undeserved self-related positive outcomes). Similarly some feelings of guilt may occur about a self-related negative action independent of whether an outcome is appraised as deserved or undeserved. The analysis also allows for blends of emotions such as pleasure with guilt, and resentment with disappointment. However, the main focus is on the labeling of discrete emotions in terms of the form taken by the evaluated action/outcome sequence which is assumed to be the key determinant of perceived deservingness.
There have been other classifications of discrete emotions presented in the literature. The task of mapping different emotions is complex and different dimensions of classification have been proposed. Some examples follow. Shaver et al. (1987) conducted a hierarchical cluster analysis of emotion names and ended up with clusters that they labeled love, joy, surprise, anger, sadness, and fear. Ortony et al. (1988) proposed a cognitive structure of emotions that considered emotions in relation to whether the focus was on the consequences of an event, the actions of the agent producing the event, or the object involved in the event. Their account also considered self and other differences in the emotions that were expected to occur in each case. Roseman (1984, 2001) distinguished between positive and negative emotions that were classified in terms of the appraisal of unexpectedness, situational state, motivational state, probability, agency, control potential, and problem type. The classification encompassed the emotions of hope, joy, relief, love, and pride on the positive side, and fear, sadness, frustration, disgust, dislike, anger, contempt, regret, guilt, and shame on the negative side. In the classification surprise was related to unexpected outcomes.
Within the context of social comparison theory, Smith (2000) related discrete emotions to upward and downward comparisons, whether the emotion produced desirable or undesirable consequences for self or other person, the focus of attention of the emotion (self, other, or dual focus), and whether the emotion involved assimilative or contrastive processes. His discussion encompassed a broad range of emotions that included envy, resentment, pride, sympathy, and schadenfreude, among others.
Building on his influential attribution approach to motivation and emotion, Weiner (2006) classified positive and negative moral emotions in relation to causal link (ability, effort) and emotional target (self, other). When the causal link was ability the self-related emotion was shame and the other-related emotions were envy, scorn or contempt, and sympathy. When the causal link was effort, the self-related emotions were guilt and regret and the other-related emotions were admiration, anger, gratitude, indignation, jealousy, and schadenfreude. In this classification shame, scorn, and envy were considered to be “immoral” because the cause (e.g., ability or lack of ability) is beyond volitional control, and most of the remaining moral emotions were negative and seen as directed toward others and linked to controllable effort.
Other researchers have also made the distinction between self-related emotions and other-related emotions. For example, Rozin et al. (1999) investigated a triad of emotions (contempt, anger, and disgust) that were assumed to reflect a concern with the social order and the violation of moral codes. They also identified a second set of emotions (shame, embarrassment, and guilt) that were assumed to be self-conscious emotions concerned with the way a person assesses his or her moral worth (see also Tangney and Dearing 2002; Tangney and Fischer 1995; Tracy et al. 2007).
These researchers have not explicitly included a justice-related dimension such as deservingness in their mapping of emotions, although some do refer to variables such as deservingness and legitimacy in their discussions (e.g., Ortony et al. 1988; Roseman 1984; Weiner 2006). The issue of how emotions relate to judgments of desert has been considered in philosophy, at least from Aristotle’s time (for recent discussion see Kristjánsson 2006; Portmann 2000). The classification presented in Table 1 attempts to redress that neglect in psychology and it has some overlap within the classifications just described in its recognition that emotions will differ in their positive or negative valence depending on whether outcomes are positive or negative, whether actions are positive or negative, and whether the emotions relate to self or other person. However, the analysis makes an important new contribution by adding deservingness to the mix, taking account of how outcomes relate to the actions that produced them, thereby generating perceived deservingness or perceived undeservingness.
Within the matrix of emotions in Table 1 evaluations of actions and their outcomes are crucially important. At a distal level needs and values would affect these evaluations (Feather 1999b). So would the prevailing social norms and social comparisons in a situation. These would also influence whether the actions that are performed and the outcomes that are generated are deemed to be appropriate or inappropriate, good or bad. Thus, judgments about the goodness or badness of actions and outcomes are complexly determined, with needs, values, and social norms as important determinants, involving both the person making the judgments and the wider social context.
The mapping of discrete emotions in Table 1 drew on theoretical accounts of emotions in the psychological literature and on some relevant empirical findings from our own and others’ research. For example, perspective differences (self vs. other) have been considered in some of the classifications mentioned above. So have positive versus negative emotions. The mapping was also based on everyday observations about how people label their emotions and how these emotions relate separately to actions and outcomes and to their interaction within a framework of needs and values, social norms and social comparisons.
The mapping of discrete values in Table 1 provides a much more differentiated and integrated picture when compared with other approaches that have included justice-related variables in their analysis of emotions. These other approaches have focused either on general affective reactions (e.g., distress or anger in equity theory; Adams 1965; Walster et al. 1978); or on discrete emotions that may occur in response to injustice, disadvantage or justice conditions (e.g., Krehbiel and Cropanzano 2000; Mikula et al. 1998; Montada 1994; Montada and Schneider 1989; Roseman et al. 1996; Weiss et al. 1999). Other researchers have targeted specific emotions such as moral outrage (e.g., Darley and Pittman 2003; Montada and Schneider 1989) or schadenfreude and sympathy (e.g., Brigham et al. 1997; Leach and Spears 2008; Smith et al. 1996; van Dijk et al. 2005).
The theoretical placement of discrete emotions in Table 1 is consistent with these approaches in accepting that the cognitive appraisal of events and outcomes is important (in this case the appraisal of outcome deservingness). But it goes further by providing a formal structure that is assumed to underlie this appraisal, a structure that takes account of evaluated actions and their evaluated contingent outcomes for self and for other person. Unlike the various attempts at classification of discrete emotion that we described previously and the limited literature on justice and discrete emotions just described, the model in Table 1 brings judgments of deservingness to center stage and does so in a formal and testable way, providing a new way of conceptualizing links between perceived deservingness/undeservingness and discrete emotions.
Research from our own research program on “tall poppies: or high status persons is consistent with this form of analysis (e.g., Feather 1994, 1996b). So is research on how people react to penalties for offenses (Feather 1996a, 1998). Perceived deservingness entered as a key variable in all of these studies and was associated with reported pleasure, sympathy, resentment, anger, and schadenfreude, depending on the particular study (e.g., Feather and Sherman 2002).
Most relevant is a recent study that tested predictions about emotions in Table 1 as they related to self-generated outcomes. Feather and McKee (2009) asked participants to recall an important positive or negative outcome in their lives that occurred in the last 12 months. The outcome had to be one for which they were partly or wholly responsible and one that they perceived as either a deserved outcome or an undeserved outcome. Participants rated a set of emotions in regard to how they felt at the time.
The results showed strong effects of outcome, with positive emotions such as pleasure, admiration, and pride all higher when the outcome was positive, and negative emotions such as sadness, sympathy, resentment, anger, guilt, regret, and disappointment all higher when the outcome was negative. However, most of these outcome effects were qualified by interaction effects involving outcome and deservingness, and these interaction effects were generally consistent with the location of the discrete emotions in Table 1 that related to self. For example, pride was especially strong for a deserved positive outcome; anger and resentment were especially high for an undeserved negative outcome. There were also effects that could be related to the valence of the action that produced the outcome (e.g., guilt and regret when the action was negative).
What is lacking from these previous studies in our research program is the inclusion of a self versus other comparison in relation to the predicted emotions in Table 1. The present study was designed to complete the picture by providing this comparison.
The present research
The study to be described investigated the emotions listed in Table 1 for self and other and extended the research to a hypothetical organizational setting where a person in a scenario was either successful or unsuccessful in a job application. The main aim of the study was to investigate perspective or focus effects relating to whether the emotions referred to self or to other person, thus providing an important extension of the previous research on deservingness and emotions. We were not concerned with outcome effects in this study. The main focus was on perspective effects holding outcome constant.
We also included surprise as an additional emotion in our inquiry. Roseman (2001) related surprise to unexpected events in his classification of emotions. It is plausible to assume that outcomes that are perceived to be undeserved may also be perceived as less expected and thus they may generate more surprise when compared with outcomes that are perceived to be deserved. These feelings of surprise would occur for both positive and negative outcomes. Hence we now include surprise as an emotion in the undeserved columns of Table 1, predicting that this emotion will be more likely to occur when outcomes are perceived to be undeserved.
Finally, we assessed both perceived responsibility for the outcome and the valence of the actions taken by the person applying for the job in order to check on our manipulations. As noted previously, the deservingness model (Feather 1999b, 2006) assumes that a person has some degree of responsibility for the action and its outcome (i.e., personal causation), though this might vary depending on experimental condition. The deservingness model also includes positive or negative actions that may lead to either positive or negative outcomes. We wanted to check that our action manipulations (high vs. low effort) led to differences in positive versus negative action valence.
- 1.When an outcome is positive:
Pleasure will be weakest when the other person’s positive outcome is undeserved;
Pride will be strongest when self’s positive outcome is deserved;
Admiration will be strongest when other’s positive outcome is deserved;
Resentment and anger will be strongest when other’s positive outcome is undeserved;
Guilt and regret will be strongest when self’s positive outcome is undeserved.
- 2.When an outcome is negative:
Pleasure will be strongest when the other person’s negative outcome is deserved;
Sadness will be weakest when other’s negative outcome is deserved;
Sympathy will be strongest when other’s negative outcome is undeserved;
Resentment and anger will be strongest when self’s negative outcome is undeserved;
Guilt and regret will be strongest when self’s negative outcome is deserved.
Surprise will be stronger when an outcome is undeserved independent of outcome.
Participants were 165 undergraduates (52 men, 111 women, 2 of unspecified gender) who were recruited from an introductory psychology course as partial fulfilment of course requirements. Participants were randomly assigned to the different experimental conditions. Mean age of the total sample was 23.54 years (SD = 7.62).
Design and procedure
Participants responded to questionnaires containing a hypothetical scenario that described a person seeking employment in an IT organization. The scenario could take different forms depending on whether the stimulus person in the scenario put in a lot of effort in the application and interview process or only a small amount of effort, and either was appointed to the position or failed to get the job. The dependent variables were ratings of how participants felt about the outcome for a wide range of emotions and their ratings of deservingness for the successful or unsuccessful outcome.
Participants were first told that they would be asked to read a scenario and to imagine that the scenario related to an actual event experienced either by a work colleague called Chris (other condition) or by self (self-related condition). Immediately before the scenario, they were asked to “try to picture the events in the scenario exactly as described.”
Other person scenarios
The scenario described a person (Chris) who had helped an accounting firm develop software applications and information technology. In order to enhance plausibility and empathy, male participants were asked to imagine Chris as a male work colleague and female participants were asked to imagine Chris as a female work colleague.
In the positive action condition Chris was described as a diligent worker for the firm, putting in a lot of effort and care in the tasks required for the job. In the negative action condition Chris was described as not always the most diligent of workers, sometimes avoiding the amount of effort and care needed to do the tasks required in the job.
Each scenario then continued the narrative by stating that Chris was attracted to work more in the IT area now that the accounting firm’s direction had been well-established. Chris responded to a newspaper advertisement for employment in an IT firm with the task of developing up-to-the-minute software for a range of purposes.
In the positive action condition Chris was described as working hard on updating his/her CV and as putting in a lot of care in writing the application for the position. Chris gave an accurate portrayal of his/her abilities and indicated a willingness to work hard, sometimes even outside normal working hours, in order to complete a job. Chris made sure to get adequate time for sleep in the night before the interview so that he/she would be ready and alert for the interview. At the interview Chris listened carefully to the questions asked by the panel and answered them accurately and thoroughly, giving appropriate and relevant information about his/her abilities and knowledge. All of these pieces of information were designed to create the perception that Chris acted in a positive and appropriate way in attempting to get the job.
In the negative action condition Chris was described as spending only a small amount of time on making some effort to update his/her CV and as producing an application that was rather hastily written and that could have been better presented. Although Chris gave a reasonably accurate portrayal of his/her abilities, he/she said little about how hard he/she would work and gave no indication of a willingness to work outside normal working hours in order to complete a job. Chris stayed up late the night before the interview playing video games and did not get much sleep. At times he/she did not answer accurately or thoroughly enough to give appropriate and relevant information in the interview about his/her abilities and knowledge. All of these pieces of information were designed to create the perception that Chris acted in a more negative way and rather inappropriately in attempting to get the job.
Information was then provided about the outcome of the job application. In the positive outcome condition participants were told that five days after the interview Chris received a letter offering him/her the position. In the negative outcome condition Chris received a letter written to say that his/her application was unsuccessful.
Participants were first told that they would be asked to begin by reading a scenario that related to “an actual event experienced by you”. Immediately before the scenario participants were asked to “try to picture the events in the scenario and to imagine yourself in it exactly as described”.
The scenarios that were used in the other person condition were modified so that they now referred to self. Thus the information about actions and outcomes now referred to self, manipulated by asking participants to take the perspective of the job applicant (e.g. “You work for an accounting firm …”; “You did not listen as carefully to the questions asked by the panel as you might have”; “You received a letter offering you the position”; etc.).
After reading their scenario participants rated a set of 30 emotions about how they would feel about the success or failure of the other person (Chris) or of self, depending on the condition. Participants were told that they could refer back to their scenario as often as they liked.
Participants used 1–7 rating scales with end labels to indicate how strongly they felt. Examples of items for the other person scenario with their end labels are “How happy would you feel about Chris’s failure?” (I would not feel happy at all, I would feel very happy); “How resentful would you feel about Chris’s success?” (I would not feel resentful at all, I would feel very resentful) and “How sorry would you feel for Chris?” (I would not feel sorry at all, I would feel very sorry). Examples of items for the self-related scenario are “How guilty would you feel about your success? (I would not feel guilty at all, I would feel very guilty) and “How surprised would you feel about your failure? (I would not feel at all surprised, I would feel very surprised).
The emotions to which participants responded in all conditions were as follows in the order in which they were presented: happy, disappointed, uneasy, contemptuous, satisfied, resentful, feeling of injustice, surprised, envious, guilty, sympathetic, angry, pleased, remorseful, sad, feeling of admiration, jealous, frustrated, joyful, disgusted, respect, upset, fulfilled, pity, depressed, ashamed, annoyed, proud, indignant, and sorry. Participants were informed that “Some of these feelings…could be stronger than others” and that “Some may be more usual and appropriate than others.”
From these emotions we selected those that were assumed to reflect the emotions listed in the cells of Table 1. The selection was made on the basis of face validity but we also used supporting evidence from previous studies (e.g., Feather 2008a; Feather and Nairn 2005; Feather and Sherman 2002). Single items were used for some emotions; composite scales were used for other emotions. In the case of the composite scales we computed the mean rating as the score for each participant.
The composite scales with the emotions that made them up and their internal reliabilities (Cronbach αs) were as follows: Pleasure/Schadenfreude (happy, satisfied, joyful, pleased; α = .95); Sadness (sad, depressed, upset; α = .87); Admiration (admiration, respect; α = .57); Pride (proud, fulfilled; α = .83); Resentment (resentful, feeling of injustice, indignant; α = .76); Anger (angry, annoyed, frustrated; α = .89); Regret (remorse, ashamed; α = .76); Sympathy (sympathetic, sorry, pity; α = .84). The score for each composite was the mean rating across the emotions making up the composite. Single item ratings were used for guilt, disappointment, and surprise.
Our measure of deservingness used five items that had been validated in previous studies (e.g., Feather 2008b). The items for other person for the success condition were as follows (the alternative statement for the failure condition is in parentheses): On the information given in the scenario, “Chris deserved to get the job (deserved the failure to get the job)”; “Chris got the job because he/she was a worthy candidate (Chris did not get the job because he/she was not a worthy candidate)”; “Chris got the job because he/she earned it (did not get the job because he/she did not earn it)”; “Chris got the job on his/her merits (Chris’s failure to get the job was a true reflection of his/her merits)”; and “Chris’s getting the job was justified (Chris’s not getting the job was justified)”. The same items were used in the self-related condition but modified so as to apply to self. Each item was answered by using a 1–7 rating scale with end labels (Disagree strongly, Agree strongly). A principal components analysis provided a one-factor solution that accounted for 79.34% of the variance. The deservingness score for each participant was the mean rating across the five items (α = .93).
In addition to providing the usual demographic information (age, sex), participants were also asked to rate the extent to which they thought that the other or self was directly responsible for getting the job or for not getting the job, depending on experimental condition. They used a 1–7 rating scale with end labels, Chris (or I) was not responsible for it at all, Chris (or I) was very much responsible for it. Participants also rated the extent to which they thought that the stimulus person’s behavior in the scenario was positive and appropriate (1–7 scale, end labels were: Chris’s behavior (or My behavior) was not positive and appropriate at all, Chris’s behavior (or My behavior) was very positive and appropriate), and how difficult it was for them to imagine the scenario about Chris or self and their feelings about it (1–7 scale, end labels were: It was not difficult to imagine it, It was very difficult to imagine it).
Mean ratings of emotions for self and other in relation to deserved and undeserved positive outcome
Deserved positive outcome
Undeserved positive outcome
4 vs. 1, 2, 3
2 vs. 1, 3, 4
1 vs. 2, 3, 4
4 vs. 1, 2, 3
4 vs. 1, 2, 3
3 vs. 1, 2, 4
3 vs. 1, 2, 4
The results relating to the manipulations showed that perceived responsibility was higher for high effort in applying for the job than for low effort, that high effort was judged more positively when compared with low effort, and that participants did not report difficulty in imaging themselves in either the self condition or the other person condition. The higher perceived responsibility in the high effort condition probably reflected a belief that the successful outcome was more the result of personal motivation and less influenced by external factors when compared with success in the low effort condition. The positive outcome was perceived as more deserved when it followed a positive action (high effort) than a negative action (low effort), consistent with our theoretical analysis. Importantly, this difference in deservingness retained statistical significance, F (1,73) = 38.96, p < .001, when we controlled for the effects of perceived responsibility via covariance analysis. Thus our manipulation directed toward creating conditions of deservingness and undeservingness was successful and remained so after differences in perceived responsibility were taken into account. Deservingness implied personal causation as reflected in perceived responsibility (Feather 1999b) but the appraisal of the evaluated action/outcome sequence was also crucial.
The results of the planned comparisons showed that all of the predictions in Hypothesis 1 were supported at a high level of statistical significance. Thus pleasure was lower for other’s undeserved positive outcome; pride was stronger for self’s deserved positive outcome; admiration was higher for other’s deserved positive outcome; resentment and anger were higher for other’s undeserved positive outcome; and guilt and regret were stronger for self’s undeserved positive outcome.
The results of the 2 × 2 ANOVAs supplement the results from the planned comparison analysis. They show that, as would be expected, when the outcome was positive (getting the job), ratings for the positive emotions (pleasure, admiration, pride) were much higher than for the negative emotions (e.g., sadness, guilt, disappointment). Across the set of emotions, however, there were strong and significant main effects of action and focus. The main effects of action showed that participants reported more pleasure, admiration and pride about the positive outcome (getting the job) when, in the role of self or other, they acted in a positive way to get the job than when they were more negative in their approach. As noted above, the positive outcome was also judged to be deserved when action was positive and undeserved when action was negative (Table 2). Note also that participants reported negligible sadness, sympathy, resentment, anger, guilt, regret, disappointment, and surprise when action was positive rather than negative, as one would expect given a deserved positive outcome. Table 2 also shows that these negative emotions were all rated more strongly but still at relatively low levels when the positive outcome followed negative action (low effort), a condition in which the positive outcome was judged to be undeserved. One exception to these low mean ratings was the relatively high mean rating for surprise which was associated with a positive outcome that was judged to be undeserved, supporting Hypothesis 3.
The main effects of focus were less widespread than those for action. They showed that reported pleasure and pride were stronger when self succeeded than when other succeeded. Resentment, anger, and disappointment were stronger when other succeeded than when self succeeded. These results reflect a bias in favor of self relating to the valence of the emotion. The significant interaction effects for guilt and regret showed that both of these emotions were strongest when self succeeded despite a negative action (low effort). Participants in the role of self reported more guilt and regret when they got the job in the IR firm without putting in much effort in the way they applied. Tests of simple effects (Tukey test) showed these differences were statistically significant (p < .01).
Taken together these results generally support the classification of the emotions in Table 1 for self and other when a positive outcome is judged to be deserved or undeserved. The additive combination of significant action effects and significant perspective or focus effects in the 2 × 2 ANOVAs would in most cases lead to the location of the corresponding emotions in Table 1, consistent with the results of the planned comparisons. However, some emotions showed no main effects of focus. These emotions were admiration, sadness, sympathy, and regret. As we have seen, all four of these emotions were associated with significant action effects. Reported admiration was higher for both self and other when the action was positive; reported sadness, sympathy, and regret were higher when the action was negative. And, as we just noted, regret was associated with a significant action by focus interaction effect.
Were these different effects mediated by perceptions of deservingness? To answer this question we conducted hierarchical regression analyses for each of the predicted emotions for the positive outcome condition. In these analyses action (coded negative = −1, positive = 1) was entered at the first step and deservingness as the mediator was added to the regression equation at the second step. We tested for mediation by conducting a Sobel test on the indirect effect of action on each predicted emotion with deservingness as the mediator. In keeping with the ANOVAs we also included focus (self, other) as a covariate at each step in each of the analyses.
The results showed that the mediated indirect path was significant for pleasure, z = 2.21, p = .027; admiration, z = 2.54, p = .011; pride, z = 1.95, p = .051; resentment, z = −4.10, p < .001; anger, z = −2.55, p = .011; guilt, z = −2.85, p = .004; and regret, z = −2.58, p = .01 (all tests are two-tailed). Inspection of the standardized beta coefficients showed complete mediation of deservingness for resentment, guilt, and regret, and partial mediation for pleasure, admiration, pride, and anger. In all cases the coefficient for the direct effect of action on the predicted emotion was reduced after deservingness was added as the mediator, and became nonsignificant in those cases where mediation was complete.
Mean ratings of emotions for self and other in relation to deserved and undeserved negative outcome
Deserved negative outcome
Undeserved negative outcome
2 vs. 1, 3, 4
2 vs. 1, 3, 4
4 vs. 1, 2, 3
3 vs. 1, 2, 4
3 vs. 1, 2, 4
1 vs. 2, 3, 4
1 vs. 2, 3, 4
3 vs. 1, 2, 4
In this case the results showed that perceived responsibility was higher for low effort in applying for the job than for high effort and that high effort was judged more positively than low effort. The higher perceived responsibility in the low effort condition probably reflected a belief that failing to get the job was caused by not trying hard enough in a situation where effort would obviously be important. Participants in the self-related condition perceived themselves as more responsible for the negative outcome and they also reported more difficulty in imagining the negative outcome. The negative outcome was perceived as more deserved when it followed a negative action (low effort) than a positive action (high effort), again consistent with the theoretical analysis based on Feather’s (1999a, b, 2006) model. And, as was the case for the positive outcome, this difference in deservingness retained statistical significance, F (1, 82) = 34.57, p < .001, when we controlled for the effects of perceived responsibility via covariance analysis.
The results of the planned comparisons showed that all of the predictions in Hypothesis 2 were supported at a high level of statistical significance. Thus pleasure (or schadenfreude) was higher for other’s deserved negative outcome: sadness was lower for other’s deserved negative outcome; resentment and anger were stronger for self’s undeserved negative outcome; and guilt and regret were stronger for self’s deserved negative outcome.
Again the results from the 2 × 2 ANOVAs supplement the results from the planned comparison analysis. They show that, as would be expected, when the outcome was negative (not getting the job), ratings for the negative emotions (e.g., anger, guilt, regret) were much higher than for the positive emotions (pleasure, admiration, pride). Across the set of emotions, however, we found strong and significant main effects of action and focus. The main effects of action showed that participants reported more pleasure, guilt, and regret about the negative outcome when, in the role of self or other, they were more negative in the way they applied for the job than when they were more positive. As noted above, the negative outcome was also judged to be deserved when it followed a negative action (Table 3). The results also showed that participants reported more admiration, sadness, resentment, sympathy, and disappointment when the negative outcome (failing to get the job) followed positive action (high effort). In these cases, the negative outcome was judged to be undeserved (Table 3). Surprise was also associated with an undeserved negative outcome, supporting Hypothesis 3.
The main effects of focus showed that pleasure was stronger when other failed to get the job compared with self’s failure. Sadness, resentment, anger, guilt, regret, and disappointment were stronger when the focus was on self’s failure when compared with other’s failure.
Note, however, that the main effects of action and focus that we have just described were qualified by two significant interaction effects, one for pleasure, the other for guilt. Participants reported more pleasure when other’s negative action led to a deserved negative outcome, thereby reflecting schadenfreude about other’s negative outcome, consistent with the placement of this emotion in Table 1. They also reported most guilt when self’s negative action led to a negative outcome that was judged to be deserved. Tests of simple effects (Tukey test) showed these differences were statistically significant (p < .01).
Taken together these results are generally consistent with the placement of the emotions in Table 1 for a deserved or undeserved negative outcome, especially for those emotions where the main effects of both action and focus were statistically significant and where additive effects consistent with the results of the planned comparisons would occur. Pride was associated with no significant main effects, as one would expect for a negative outcome. Other emotions (admiration, sympathy, and surprise) showed no main effects of focus but were associated with positive behavior on the part of the job applicant. Anger was associated with focus (more anger about self’s failure) but not with the action variable.
To test for the mediation of perceived deservingness we again conducted hierarchical regression analyses for each predicted emotion, this time for the negative outcome condition. Action (positive, negative) was entered at the first step of the analysis, deservingness was added to the regression equation at the second step, and focus (self, other) was entered as a covariate at each step.
Sobel tests showed that the mediated indirect path was significant for sadness, z = 2.30, p = .022; resentment, z = 3.27, p = .001; anger, z = 2.90, p = .004; disappointment, z = 3.30, p = .001; and close to significance for sympathy, z = 1.89, p = .051 (all tests are two-tailed). Inspection of the standardized beta coefficients showed complete mediation of deservingness for sadness, sympathy, resentment, and disappointment. In each case the coefficient for the direct effect of action on the predicted emotion was reduced and nonsignificant after deservingness was added as the mediator.
The present study adds in an important way to our previous research on deservingness and emotions by incorporating a perspective manipulation. This extension enables a more complete test than hitherto (Feather and McKee 2009) concerning the location of the emotions in Table 1 for self and other and their relation to deserved and undeserved positive or negative outcomes. We assumed that people will appraise an action, appraise an outcome, and appraise the conjunction of an action and its outcome. The latter action/outcome relation was assumed to be a prime influence on the perceived deservingness of the outcome and reported emotions were then assumed to vary according to whether they applied to self or other and to whether positive or negative outcomes were perceived to be deserved or undeserved.
Our results provided good support for this analysis and they make a significant contribution to the limited literature linking emotions to deservingness and to perspective. We should note than in our analysis there are some emotions that are not restricted to a single cell in Table 1 (e.g., pleasure, sadness, guilt, regret, resentment, surprise) and the results were consistent with this assumption. We should also note that we would expect some transfer of emotions to occur across the focus conditions, although some may more clearly relate to self than to other person or vice versa (e.g., Tangney and Fischer 1995; Tangney et al. 2007). Thus, one can feel pleasure and pride about another person’s achievement as well as pleasure and pride in one’s own success, and admiration about self or other’s positive behavior regardless of the outcome. One can also feel sadness and disappointment when another person suffers a negative outcome as well sadness and disappointment about one’s own failure. Guilt and regret, however, are more restricted to a self-related negative action, and surprise to outcomes that are perceived to be undeserved.
Our study used a hypothetical scenario that was designed to provide a realistic context for the judgments that were made. Participants responded in terms of how they would feel in this hypothetical situation rather than reporting on actual emotional experiences. The use of hypothetical scenarios has been criticized by Parkinson and Manstead (1993) because they may tap into stereotyped or normative views about what emotions might be expected rather than emotions that would actually be experienced in a particular context. This criticism may be overdrawn because normative judgments about emotions presumably have some basis in reality. Moreover, some research suggests that findings from studies that use hypothetical scenarios tend to be consistent with findings from studies that investigated actual emotional experiences (Robinson and Clore 2001).
How does the present analysis relate to other theoretical ideas from appraisal theory that concern justice variables and discrete emotions? At the conceptual level deservingness can be distinguished from concepts such as perceived legitimacy, accountability, and norm-compatibility that have been discussed in the appraisal literature (e.g., Smith and Lazarus 1993; Roseman 1984, 2001; Scherer 1984; Scherer et al. 2001). Perceived legitimacy (Roseman 1984, 2001) relates not only to deservingness but also to perceived entitlements that encompass rights, regulations, rules and social norms (Feather 2008b). So it is a higher-order concept. Self or other-accountability (Smith and Lazarus 1993) is similar to perceived responsibility (Weiner 1995, 2006) which, as we noted previously, is conceived as a variable that moderates deservingness reflecting personal causation (Feather 1999a, b, 2006). Norm-compatibility (Scherer 1984; Scherer et al. 2001) would influence the way actions and outcomes are evaluated. As noted previously, we would expect the prevailing social norms within a situation to be important influences on how people define actions and outcomes as good or bad. These norms are situational influences that affect evaluations in addition to the influence of a person’s own needs and values. The structure of these action/outcome evaluations would then affect perceived deservingness according to a consistency principle.
Also relevant is Weiner’s (1986, 1995, 2006) analysis of emotions that relates different emotions to causal attributions and perceived responsibility and that also takes account of emotions relating to other person and to self (see Feather 2006, pp. 60–62 for details). In Weiner’s analysis the self-directed emotions of guilt and shame following a negative event are related to controllable causality and uncontrollable causality respectively, and the other-directed emotions of anger and sympathy following a negative event are also related to these respective types of cause. Weiner’s analysis also allows for outcome dependent effects and in his later formulations (Weiner 1995, 2006) judgments of responsibility play an important role. But deservingness does not feature as a key variable in his analysis. Judgments of deservingness are further along in the chain of events and they presume that a person is responsible to some degree for the action that he or she performed and the outcome that followed it.
In the present study judgments of responsibility were higher when success followed hard work and also when failure followed low effort. They were lower when success followed low effort and also when failure followed high effort. Thus these judgments mirrored judgments of deservingness. It is probably the case that responsibility and deservingness judgments often share a similar profile. But, as noted previously, there would also be cases when these judgments are discrepant. In some cases causal attributions for an action and its outcome may diminish the role of personal causation when a situation suggests that perceived causes may operate that are external and uncontrollable from the person’s point of view (e.g., too many or too few applicants for the job in the situation used in the present study). Such attributions would undermine personal responsibility for the outcome with consequent effects on perceived deservingness. At the extreme, deservingness would be zero when responsibility was zero.
Future research is needed to disentangle the different effects of deservingness and responsibility across a range of contexts but clearly the structural model of deservingness offers an alternative way of mapping discrete emotions, taking account of the evaluated action/outcome sequence as well as other variables such as like/dislike relations and ingroup/outgroup relations that are part of the deservingness model but outside of the scope of the present study (see Feather 1999b, 2006). These variables, along with responsibility, are assumed to moderate judgments of deservingness with consequent effects on discrete emotions. For example, we would expect that the deserved failure of a disliked person or member of an outgroup would elicit more pleasure or schadenfreude when compared to how one reacts to the deserved failure of a liked person or ingroup member. Results of research support this prediction (e.g., Feather and Sherman 2002).
Finally, the analysis presented in the current article is not meant to cover all types of situations, only those where considerations of justice come into play. As Feather (2006, pp. 64–65) noted, anger may occur in situations where justice is not an issue, often occurring as a relatively automatic response to frustrating events (Berkowitz 1962). Disappointment may occur when high expectations are violated and not confirmed by positive outcomes (Feather 1963). Sympathy may occur when a person is perceived to be a victim of forces that are beyond his or her control and where personal responsibility for the negative event is absent (Weiner 1995). Schadenfreude following another person’s failure has been shown to occur when there is resentment about the other person’s previous undeserved high achievement (Feather 2008a; Feather and Sherman 2002; Feather and Nairn 2005) and also in relation to malicious envy (Smith and Kim 2007) and, at the group level, to the pain of inferiority (Leach and Spears 2008). Emotions such as shame, guilt, and regret often relate to the social context that surrounds interpersonal interactions (e.g., Baumeister et al. 1994; Berndsen et al. 2004; Fischer et al. 2003; Hareli and Parkinson 2008; Parkinson et al. 2005; Tangney and Fischer 1995; Tangney et al. 2007). Feather (2006) proposed that, in addition to the guilt that relates to an undeserved positive outcome for self, “feelings of guilt may also occur when p’s action has negative effects on another person, especially on one who is liked or a member of p’s ingroup” (p. 244). Emotions such as contempt, anger, and disgust may occur relatively automatically as intuitive “gut” feelings and reflect respect for the social order and the violation of moral codes (Haidt 2007; Rozin et al. 1999). Emotions are also often triggered by subtle cues (Ben-Ze’ev 2000). However, we propose that beliefs about deservingness/undeservingness are an important part of this complex fabric (see also Kristjánsson 2006) and that the mapping of discrete emotions also has to recognize who performs the action that leads to the outcome (self or other).
By focusing on the evaluative structure of action/outcome relations and the appraisal of deservingness the theoretical analysis and research that we have presented opens up a new way of conceptualizing discrete emotions. The analysis incorporates a person’s own values as well as social norms as key influences on how positive or negative actions and their positive or negative outcomes are evaluated. It allows for perspective differences relating to self or other focus, and it also provides for the effects of moderating variables that influence perceived deservingness. Thus, in a number of ways, it contributes to the social psychology of justice and emotion, furthering our understanding of how beliefs about deservingness influence the way people feel about outcomes.
- Adams, J. A. (1965). Inequity in social exchange. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 2, pp. 267–299). New York: Academic Press.Google Scholar
- Ben-Ze’ev, A. (2000). The subtlety of emotions. Cambridge: MIT Press.Google Scholar
- Berkowitz, L. (1962). Aggression: A social psychological analysis. New York: McGraw-Hill.Google Scholar
- Brigham, N. L., Kelso, K. A., Jackson, M. A., & Smith, R. H. (1997). The roles of invidious comparisons and deservingness in sympathy and Schadenfreude. Basic and Applied Social Psychology, 19, 363–380.Google Scholar
- Feather, N. T. (1994). Attitudes toward high achievers and reactions to their fall. In M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 26, pp. 1–73). San Diego, CA: Academic Press.Google Scholar
- Feather, N. T. (1996b). Values, deservingness, and attitudes toward high achievers: Research on tall poppies. In C. Seligman, J. M. Olson, & M. P. Zanna (Eds.), The psychology of values: The Ontario Symposium (Vol. 8, pp. 215–251). Mahwah, NJ: Lawrence Erlbaum Associates.Google Scholar
- Feather, N. T. (1999b). Values, achievement, and justice: Studies in the psychology of deservingness. New York: Kluwer Academic/Plenum Publishers.Google Scholar
- Feather, N. T., & McKee, I. R. (2009). Differentiating emotions in relation to deserved or undeserved outcomes: A retrospective study of real-life events. Cognition and Emotion, 23, 955–977.Google Scholar
- Feather, N. T., & Sherman, R. (2002). Envy, resentment, schadenfreude, and sympathy: Reactions to deserved and undeserved achievement and subsequent failure. Personality and Social Psychology Bulletin, 28, 953–961.Google Scholar
- Kristjánsson, K. (2006). Justice and desert-based emotions. Aldershot, UK: Ashgate.Google Scholar
- Ortony, A., Clore, G. L., & Collins, A. (1988). The cognitive structure of emotions. Cambridge: Cambridge University Press.Google Scholar
- Parkinson, B., Fischer, A. H., & Manstead, A. S. R. (2005). Emotion in social relations. New York: Psychology Press.Google Scholar
- Portmann, J. (2000). When bad things happen to other people. New York: Routledge.Google Scholar
- Roseman, I. J. (1984). Cognitive determinants of behavior: A structural theory. In P. Shaver (Ed.), Review of personality and social psychology: Vol 5. Emotions, relationships, and health (pp. 11–36). Beverly Hills, CA: Sage.Google Scholar
- Roseman, I. J. (2001). A model of appraisal in the emotion system: Integrating theory, research, and applications. In K. R. Sherer, A. Schorr, & T. Johnstone (Eds.), Appraisal processes in emotion: Theory, methods, research (pp. 68–91). New York: Oxford University Press.Google Scholar
- Scherer, K. R. (1984). On the nature and function of emotion: A component process approach. In K. R. Scherer & P. Ekman (Eds.), Approaches to emotion (pp. 293–317). Hillsdale, NJ: Lawrence Erlbaum Associates.Google Scholar
- Scherer, K. R., Schorr, A., & Johnstone, T. (Eds.). (2001). Appraisal processes in emotion: Theory, methods, and research. Oxford, UK: Oxford University Press.Google Scholar
- Smith, R. H. (2000). Assimilative and constrastive emotional reactions to upward and downward social comparisons. In L. Wheeler & J. Suls (Eds.), Handbook of social comparison: Theory and research (pp. 173–200). New York: Kluwer.Google Scholar
- Tangney, J. P., & Dearing, R. L. (2002). Shame and guilt. New York: Guilford.Google Scholar
- Tangney, J. P., & Fischer, K. W. (Eds.). (1995). Self-conscious emotions: Shame, guilt, embarrassment, and pride. New York: Guilford Press.Google Scholar
- Tangney, J. P., Stuewig, J., & Mashek, D. J. (2007). What’s moral about the self-conscious emotions? In J. L. Tracy, R. W. Robins, & J. P. Tangney (Eds.), The self-conscious emotions: Theory and research (pp. 21–37). New York: Guilford Press.Google Scholar
- Tracy, J. L., Robins, R. W., & Tangney, J. P. (Eds.). (2007). The self-conscious emotions: Theory and research. New York: Guilford Press.Google Scholar
- Walster, E., Walster, G. W., & Berscheid, E. (1978). Equity: Theory and research. Boston: Allyn Bacon.Google Scholar
- Weiner, B. (1986). An attributional theory of motivation and emotion. New York: Springer.Google Scholar
- Weiner, B. (1995). Judgments of responsibility: A foundation for a theory of social conduct. New York: Guilford.Google Scholar
- Weiner, B. (2006). Social motivation, justice, and the moral emotions. Mahwah, NJ: Lawrence Erlbaum Associates.Google Scholar