Abstract
Using the context of an unexpected airline upgrade, we examined factors that influence an individual’s reaction when they are overrewarded compared to others: guilt-proneness and relationship to the other, underrewarded, individuals. Results demonstrated that for individuals high in guilt-proneness, satisfaction with the upgrade and behavioral intent may be qualified by a feeling of existential guilt when they receive benefits that others do not, particularly if they have a close relationship with those others. Our results extend the research in advantageous inequality by showing that people high in guilt-proneness tend to have a heightened sensitivity to such injustices. Our findings also have important implications for the hospitality, airline, and travel industries: for customers high in guilt-proneness, receiving an expected upgrade may, in fact, have unintended negative results. Managers can use this information to make employees aware of the potential detrimental effects of rewarding or upgrading only one member of a party.
Introduction
Nearly anyone who has worked in a customer service capacity has heard the familiar directives: “Surprise and delight the customer,” “Exceed customer expectations,” “Create a WOW moment,” and “Go above and beyond.” Going that extra mile has obvious benefits, in that it can result in customers who are happy and satisfied. But there may be a hidden downside to giving customers unexpected benefits, upgrades, and perks: existential guilt.
Individuals can experience existential guilt when they feel that they have been the recipients of rewards, benefits, or good fortune that other, equally worthy, people do not receive. An awareness of this discrepancy between one’s own circumstances and those of others can result in feelings of guilt (Dahl, Honea, and Manchanda 2003; Montada, Schmitt, and Dalbert 1986). Reynolds and Harris (2009) note that in order for an individual to experience existential guilt, he must perceive that his own good fortune (relative to others) has not been earned, is not deserved, and is not justified by circumstances. This sense of illegitimacy is key—if an individual feels that he has been rewarded because of his own effort or achievement, he is more likely to feel pride than guilt (Branscombe 1998).
Unlike reactive guilt, where one feels guilt for having done something wrong, existential guilt can be experienced even when the individual has no responsibility for his or her advantaged situation (Harth et al. 2008). For example, those enjoying a high standard of living in first-world countries may experience existential guilt when taking a vacation and noting the poverty in a developing country (Pearce 2012). Of course, not all situations will affect people in the same manner, and different individuals may experience varying levels of emotional reactions to being overrewarded comparedto others. While a person may feel happy or satisfied with the outcome, this may be tempered by a recognition that the situation is somewhat unfair, leading to a feeling of guilt. Research has shown that people are deeply affected when they get outcomes that are better than they deserve (Adams 1963, 1965). When individuals find themselves in inequitable situations, they become distressed. As a result, the person who gets too much may feel guilt or shame (Adams 1965) and often report “feelings of unease” about getting more than they deserve (van den Bos et al. 2011; Jacques 1961).
Related findings of research show that people are only moderately satisfied with overpayment—more so than with underpayment and, interestingly, less so than with equitable payment (Peters and van den Bos 2008). These findings generally suggest that fairness concerns have a big impact on outcome evaluations (Montada 2002; van den Bos et al. 2006).
A deeper understanding of existential guilt is important in a consumer travel and tourism context, as it can impact the customer’s evaluation of the service provider and his satisfaction with the service. In this study, we examine two factors that influence an individual’s sense of justice, satisfaction, and behavioral intent when they are overrewarded compared to others by receiving an unexpected upgrade. The first factor is guilt-proneness, an individual-level factor measuring the propensity to experience guilt across a variety of circumstances (Cohen et al. 2011). The second is the relationship to the other, underrewarded, individuals: whether they are friends or strangers.
We set our study in an airline context, as airline travel is often a major part of any travel and tourism experience (Poria, Reichel, and Brandt 2010). Previous research has established the influence of the flight experience on customers’ satisfaction and loyalty (Alamdari 1999; Hanlon 2007). While research on airline travel is not new, recent studies have begun to focus more on the experiences and emotions of airline passengers (Poria, Reichel, and Brandt 2010; Fleischer, Tchetchik, and Toledo 2012). For example, Poria, Reichel, and Brandt (2010) conducted a qualitative study on the experiences of airline passengers with disabilities, and more recently Fleischer, Tchetchik, and Toledo (2012) examined airline passengers’ fear of flying. Our study adds to the literature on the emotions of airline passengers by exploring passengers’ reactions on being rewarded with an unexpected airline upgrade.
Background Literature
Guilt Proneness
While guilt is an emotion familiar to all of us, the propensity to feel guilt varies with the individual. This propensity to feel guilt across situations, known as “guilt proneness,” is a stable individual-level characteristic that can serve as a predictor of experienced guilt levels for an individual. For example, a highly guilt-prone person is likely to experience guilt not only when they commit a transgression, but also when they were not to blame for the negative outcome, when they could have done nothing to prevent it, or even when most would argue that the victim deserved the negative outcome. The Guilt and Shame Proneness scale (Cohen et al. 2011) was developed to measure individual differences in the propensity to feel guilt and shame in varying situations. People high in guilt-proneness tend to have a heightened sense of morality, justice, and right versus wrong (Tangney and Dearing 2002). Given this, it follows that when these individuals find themselves to be the recipients of benefits that they have not earned and that others do not receive, they would be highly likely to be bothered by the unfairness of the situation and experience existential guilt. The happiness and satisfaction that they may feel due to the positive outcome they have received will be tempered by these feelings of guilt and a sense that they have unfairly profited.
Relationship Status
As guilt is primarily an interpersonal emotion (Baumeister et al. 1994), it is often experienced as the result of a disruption in social relationships. Guilt has been linked to feelings of loneliness, a sense of social rejection, and interpersonal conflicts. People also report that when feeling guilty, they feel less secure and less valued by others (Baumeister et al. 1995). The results suggest that feelings of guilt may be linked to fears of loss of social acceptance. This disruption of social relationships appears to be germane to feelings of guilt (Baumeister et al. 1994).
Previous research has suggested that the way an individual responds to being in an advantaged position may be partially dependent on the individual’s relationship with the disadvantaged person (Smith 1993). Research in corporate layoff conditions has suggested that the closeness ofthe interpersonal relationship between the advantaged person (who still has a job) and disadvantaged person (who was laid off) can affect the survivor’s attitudinal and behavioral reactions toward both the laid-off colleague and the company (Brockner et al. 1987; Heider 1958)
Clark and Mills (Clark and Mills 1993; Mills and Clark 1982) made an important distinction between different types of relationships: communal relationships and exchange relationships (Peters and van den Bos 2008). Communal relationships lead to a feeling of mutual responsibility for one another’s needs (e.g., friends). On the other hand, in exchange relationships, people are not motivated to act for the needs of others (e.g., strangers). Research has shown that people who interact with friends, compared to those who interact with unknown others, attend more to the others’ needs (Clark, Mills, and Corcoran 1989), help the other more (Clark et al. 1987), and do not keep track of individual inputs into the relationship (Clark et al. 1989). In line with this research tradition, Peters and van den Bos (2008) found that when interacting with friends (vs. strangers), people are motivated to attend to the needs of their friends, and therefore, they are not only concerned with their own outcomes but also with their friends’ outcomes. As a result, receiving unexpected rewards is less satisfying when this is done relative to friends as opposed to strangers.
In addition, a sense of identification with the disadvantaged person or group also appears to be linked with feelings of guilt. Montada and Schneider (1989) found that advantaged people may feel empathy and sympathy for the disadvantaged group, and that such emotions were highly correlated with existential guilt. It follows that for groups with which one feels most closely identified, such as family, friends, or close colleagues, these feelings would be magnified. Hence, we predict the following:
Hypothesis 1: The relationship status of othersnot receiving the upgrade (friends vs. strangers) moderates the impact of guilt-proneness on the subject’s levels of satisfaction with the upgrade and behavioral intent.
Specifically, we predict that when the others not receiving the upgrade are friends as opposed to strangers, subjects will be less satisfied with the upgrade, less satisfied with the company’s handling of the upgrade, less happy about the upgrade, less likely to spread positive word of mouth, and less likely to do business with the airline again. Furthermore, we wished to explore the psychological mechanism by which this happens. We propose that the customer’s sense of fairness is key to understanding his satisfaction with the upgrade and subsequent behavior.
Fairness
The relationship between fairness or justice and customer happiness, satisfaction, and subsequent behavior has received much attention in the tourism literature. For example, Chung and Petrick (2012) found a relationship between consumer perceptions of the fairness of ancillary airline fees and their emotions and behavioral intent toward the airline. Chunget al. (2011), in a study about national forests, found that perceived fee fairness impacted customer willingness to pay the levied fees. Oh (2003) examined the impact of price fairness on overall customer quality and value judgments in a hotel context, and perceived justice has also been demonstrated to influence customer satisfaction, word of mouth, and return intent (Kim, Kim, and Kim 2009). Alexander et al. (2012) explored the concept of fairness and justice regarding multilevel queues in theme parks and found evidence suggesting that a perception of unfairness resulted in negative emotional responses and adverse behavioral intentions.
Taken together, these studies suggest that when a traveler or tourist perceives that he or she has been the victim of an unfair or unjust situation, this injustice results in negative emotional and behavioral responses. Our study fills a gap by examining the customer’s reactions when they are in the opposite situation: when they are the recipient of an upgrade that others do not receive and that they did not earn. We postulate that perceived fairness will mediate the impact of relationship status on satisfaction with the unexpected upgrade and behavioral intent. Therefore, we predict the following:
Hypothesis 2: Fairness mediates the impact of relationship status on the subject’s levels of satisfaction with the upgrade and behavioral intent.
Methodology
Sample
A marketing research company was hired to obtain data via online survey. The sample was composed of adults in the United States older than age 18. A total of 152 usable responses were received. In terms of demographics, the sample was evenly split for gender (51% female, 49% male). Sixty percent of the subjects had a college degree and approximately 53% reported a household income more than $60,000. Fourteen percent of the respondents had a household income more than $100,000. Seventy-six percent of the respondents were Caucasian.
Subjects were asked about their travel experiences over the past two years. The mean number of airline trips taken was 5 (minimum = 0, maximum = 70) and the mean number of upgrades received over that time was 0.52 (minimum = 0, maximum = 12). Forty-seven percent of the respondents were members of a frequent-flyer program.
Design and Stimuli
The design was a 2 (friends vs. strangers) × 2 (guilt-proneness: high vs. low) quasi-experiment. The participants were assigned to a condition and presented with a scenario describing a recent incident at an airline ticket counter. Subject were asked to imagine that they were travelling and received an unexpected upgrade to first class while others did not. These others were either strangers standing next to them at the ticket counter or close friends that were travelling with the subject (see Appendix A). An example of the scenarios is provided below.
You and some of your friends have decided to fly to Vegas for a fun weekend getaway.
Since you are all flying from different airports, you decide to all meet at the hotel
around dinner time. You get to the airport, check in curbside, and the agent tells you
that you will need to check in with the attendant at the gate to get your seat
assigned. You go through security, grab some snacks, and head to your gate. You go up
to the counter where you and several strangers are crowded around the ticket counter
waiting to get seat assignments. The ticket agent looks at all the passengers around
the desk and says, “Who would like to go first?” No one answers, so you say, “I guess
I’ll go first.” She says, “Ok, may I have your name?” You give it to her, and as she
checks you in, she says, “Would you like to be upgraded to first class?” You are not a
member of that airline’s frequent-flyer club, so you ask the agent why you would get
upgraded. The agent replies, “No particular reason. I have one empty seat left in
first class, so you can have it if you like.” You say, “Sure!” and she gives you your
boarding pass.
After reading the scenario, subjects completed a survey regarding the scenario and demographic information was collected.
Measures
Satisfaction with the company’s handling of the upgrade was measured with three items on 7-point Likert-type scales: “How satisfied are you with the company’s handling of the upgrade?” (1 = very dissatisfied, 7 = very satisfied). “When considering the upgrade that you received, how did it make you feel?” (1 = very unhappy, 7 = very happy). “When considering the upgrade that you received, how did it make you feel?” (1 = very displeased, 7 = very pleased), (Cronbach alpha = .85). Guilt-proneness was measured on a 7-point Likert-type scale using the eight guilt items from the GASP, adapted from Cohen et al. (2011) (Cronbach alpha = .83). Fairness was measured with three items adapted from van den Bos and Lind (2001): “When considering the upgrade you received from the airline employee: I think it was fair, I think it was appropriate, I think it was just” (1 = strongly disagree, 7 = strongly agree) (Cronbach alpha = .89). Behavioral intent was measured with four items adapted from Zeithaml, Leonard, and Berry (1996): “Say positive things about the airline to others”; “Recommend the airline to others”; “Consider the airline your first choice in the service category”; “Do more business with the airline in the next few months”; Cronbach alpha = .94.). Two Likert-type questions were used as manipulation checks: “The others in line with me were good friends”; “The others in line with me were strangers”; 1 = strongly disagree, 7 = strongly agree). Scenario realism was captured via two items: “This situation is very likely to be real, this situation could happen, or has happened, to me or someone I know”; 1 = strongly disagree, 7 = strongly agree (r = .676). Household income, gender, and number of trips were used as covariates in order to rule them out as possible alternative explanations.
Results
Manipulation Checks
Participants in the friend condition showed a significantly higher level of agreement with the question tapping into others being friends than their counterparts in the stranger condition (M = 5.34 vs. M = 3.25, t = 7.86, p < .001). Similarly, participants in the stranger condition had higher levels of agreement with the question reflecting others in line as strangers than participants in the friends condition (M = 5.76 vs. M = 3.87, t = 7.24, p < .001). Taken together, these results indicate that our manipulations were effective. The mean rating for scenario realism was 4.74 and there were no differences across the two conditions.
Participants were categorized into high and low guilt proneness based on a median split on the guilt items from the GASP scale (minimum = 3.13, maximum = 7.0, mean = 5.66, standard deviation = .95).
Hypothesis Testing
The two dependent variables, satisfaction with the upgrade and behavioral intent, are highly correlated (r = .695) and therefore a multivariate analysis of covariance (MANCOVA) was applied to control the experimental-wise error rate. Results from a MANCOVA table indicate significant main effects for relationship status (Wilks Lambda = .879, p < .001) and for guilt proneness (Wilks Lambda = .955, p < .001). However, these main effects were qualified by a significant two-way interaction (Wilks Lambda = .950, p < .05). Household income, gender, and number of trips were used as covariates in order to rule them out as possible alternative explanations, and these variables were not significant. Descriptive means are shown in Table 1.
Descriptive Means.
Univariate ANCOVA of satisfaction reveals a significant main effect of relationship status, F(1, 151) = 24.84, p < .000, and a marginal effect of guilt-proneness, F(1, 151) = 4.55, p = .06. However, these effects are qualified by the expected interaction with relationship to the disadvantages (friends vs. strangers), F(1, 151) = 4.89, p < .05. This interaction is visualized in Figure 1. Planned contrasts showed that participants high in guilt-proneness were significantly less satisfied with the upgrade when the incident took place in front of their friends rather than strangers, M = 5.30 versus 6.41, t = 5.28, df(78), p < .001. Conversely, for participants who are low in guilt-proneness, their satisfaction with the upgrade did not differ across the two relationship conditions, that is, friends versus strangers, M = 5.33 versus 5.77, t = 1.54, df(70), ns.

Estimated marginal means of satisfaction.
Moreover, a univariate ANCOVA table reveals a significant interaction effect of guilt-proneness and relationship status on behavioral intentions, F(1, 151) = 7.21, p < .01. The interaction is visualized in Figure 2. The results from planned contrasts indicated that in the high guilt-proneness condition, participants were more likely to use the airline for future travel and to recommend the airline to others if the upgrade happened in front of strangers as opposed to their friends, M = 6.30 versus 5.30, respectively; t = 4.57, df(78), p < .001. For participants low in guilt-proneness, the effect of relationship status disappeared, M = 5.32 versus 5.43 for friends and strangers, respectively; t = 0.377, df(70), ns.

Estimated marginal means of behavioral intent.
Fairness
A mediational analysis was conducted to assess whether consumer’s fairness perceptions mediated the effects of relationship status on satisfaction with the upgrade and behavioral intentions. See Appendix B. Following Hayes’s (2012) procedures, the results for both dependent variables support the mediating effect (the 95% bootstrap confidence interval for satisfaction with the upgrade (.043, .41); the 95% bootstrap confidence interval for behavioral intent (.054, .40, number of bootstrap samples = 5000 for both). Since the confidence intervals for indirect effects are above zero, the mediation is significant. These results are congruent with hypothesis 2.
Discussion
The relationship between fairness, satisfaction, and subsequent behavior has been widely researched in the context of airlines (Chung and Petrick 2012), national forests (Chung et al. 2011), hotels (Oh 2003), and theme parks (Alexander et al. 2012). The results of our study add to this body of literature by examining the customer’s reactions when they are the recipient of an upgrade that others do not receive and that they did not earn. Traditional wisdom suggests that the objective of customer service should be to “surprise and delight.” Common sense would suggest that giving a customer unexpected perks and upgrades should enhance their experience, resulting in satisfied customers who will return. Indeed, customers perceive first-class seats on an airline to be an improvement over coach seats (Makens and Marquardt 1977), and it follows that a customer who receives an upgrade should be delighted. Our results, however, suggest that for individuals high in guilt-proneness, these effects may be qualified by a feeling of existential guilt when they receive benefits that others do not, particularly if they have a close relationship with those others. Existential guilt arises when one is the recipient of benefits or rewards that are unearned or undeserved while others do not receive these same benefits (Montada, Schmitt, and Dalbert 1986; Dahl, Honea, and Manchanda 2003). When individuals experience advantageous inequality, there is a conflict between the tendency to accept a desirable outcome and doing what is fair, or in other terms, a conflict between what they want and what they believe to be right (Peters and van den Bos 2008). From this proposition follows that when people are overrewarded, there is a source of positive affect and a source of negative affect. The positive source is the egoism-based pleasure of receiving a relatively good outcome, whereas the negative source is the fairness-based feeling of being unfairly advantaged (van den Bos et al. 2006; van den Bos and Lind 2001). Theoretically, our results extend the research in advantageous inequality by showing that people high in guilt-proneness tend to have a heightened sensitivity to such injustices (Tangney and Dearing 2002).
Our results also demonstrate that the impact of guilt proneness on a customer’s satisfaction and behavior can be moderated by the relationship status of other customers who do not receive the same benefits. Peters and van den Bos (2008) found that the conflict between egoistic hedonic concerns and concern for the other person’s needs is relatively stronger in a close relationship than in a non-close relationship, and hence satisfaction judgment to overpayment is especially difficult. Our results gave support for their study and demonstrated that when customers who are high in guilt proneness received an unexpected upgrade while their friends did not, they experienced an adverse reaction to the upgrade. They tended to be less satisfied and be less likely to spread positive word of mouth and do repeat business with the company.
Furthermore, our results shed light on the psychological mechanism by which this happens. The results of our mediation analysis demonstrate that the customer’s sense of fairness mediates the impact of relationship status on the subject’s levels of satisfaction and behavioral intent. While previous research in the travel and tourism literature has examined the relationship between fairness and the customer’s emotional and behavioral outcomes (Chung and Petrick 2012; Oh 2003; Kim, Kim, and Kim 2009; Alexander et al. 2012), these studies cast the focal customer as the victim of the unfairness. Our research adds to the body of literature by demonstrating that fairness mediates the impact of relationship status on these outcomes, even when the customer is the beneficiary, rather than the victim, of the unfair situation.
Managerial Implications
Our findings, which run contrary to the conventional wisdom that an unexpected perk or upgrade is always a good thing, have important implications for the hospitality, airline, and travel industries. Unexpected upgrades happen not only in the airline industry but also in the hotel industry, where customers who have been overbooked in standard rooms are often upgraded to larger rooms or suites (Toh 1985). First, our results have clear implications for service industry employees and managers. Current practice at some airlines dictates that when a customer is upgraded, one companion may be upgraded with them, but only if that companion is traveling on the same reservation number. In addition, some airlines require that in order for a companion to be upgraded, their ticket must be of a specific revenue class (United Airlines 2012; Delta Airlines 2012). This policy does not address the needs of larger families, groups of friends, or partners traveling on different reservations. Although employees may assume that granting a customer an unexpected reward or upgrade will result in a more satisfied customer who is more likely to return, our findings suggest that there may be limitations to this assumption. We found that for customers high in guilt-proneness, receiving an expected upgrade may backfire, especially when their friends did not (as opposed to strangers). Service employees, then, would be well served to ensure that when granting upgrades or rewards, all members of a party receive the same treatment. If they do not have enough upgrades for the entire party of friends or family, it may be better to offer it to no one than to only a few members. In addition, managers can use this information to make front line employees aware of the potential detrimental effects of rewarding or upgrading only one member of a party.
Limitations and Future Research
As with any research, this study has several limitations. First, our study employed hypothetical scenarios and differentiated only between “friends” and “strangers.” Future research could employ alternative methods, such as field studies, and also further define the relationship between the upgraded subject and the other customers. Second, our study measured guilt-proneness as an individual-level characteristic, but future studies could measure subjects’ levels of experienced guilt. Third, in this study, we did not consider self-other focus and legitimacy of the unexpected rewards. Appraisal theories of emotion (Smith and Ellsworth 1985) suggest that individuals first appraise whether a situation affects the self or others. This distinction allows individuals to focus their attention on the relevant party (Harth et al. 2008). Depending on the legitimacy of the advantageous inequality, a self-focus may result in feelings of guilt or pride (Leach 2002), whereas an other-focus should lead to the emotion of sympathy. Thus, future studies should consider the focus as legitimacy is an important determinant of the specific emotions felt about advantageous inequality.
Additionally, this study focused on the emotions of the customer when he has received a reward that others did not receive. Future studies could examine the focal customer’s emotions when it is the others who receive the reward, instead.
Footnotes
Appendix A
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
Author Biographies
