Why Are Subordinates Less Satisfied with Mediation? The Role of Uncertainty

An abundance of literature and research (Wall, Stark, and Standifer 2001; Kressel 2006; Donohue 2007) reflects the rapid proliferation of labor‐related mediation used for different types of organizational conflict, such as collective labor–management conflicts (Martinez‐Pecino et al. 2008), inter‐ and intragroup conflicts (Jones 2005), and interpersonal conflicts (Herrman 2006) between colleagues or between subordinates and supervisors. In this article, we focus on hierarchical conflicts.

Depending on their jurisdiction, workers confronted with workplace conflicts will often have the choice of resolving the conflict via litigation, by filing a grievance internally or externally with a government agency, or by choosing an alternative dispute resolution vehicle, such as mediation. Mediation has been defined as “a process in which disputants attempt to resolve their differences with the assistance of an acceptable third party who counters tendencies toward competitive win–lose strategies” (Kressel 2006: 726). In arbitration a third party provides a final and binding decision, but in mediation the parties select, under a mediator's guidance, the final resolution of their own problem (Kressel 2006; Goldman et al. 2008). Timely use of mediation can prevent or limit the negative consequences of conflict (Giebels and Janssen 2006) by helping participants recognize and manage the different perceptions, cognitions, and emotions they hold in the conflict or the mediation (Bush and Folger 1994).

Although a mediation goal is to promote the equal participation of both parties and flattened lines of communication, it is unclear to what extent this is possible when parties hold structurally different positions. Research in the context of hierarchical conflicts indicates that hierarchy and power affect perceptions and experiences in conflict as well as conflict behavior (Van de Vliert, Euwema, and Huismans 1995; Fitness 2000). Consequently, it is possible that structures and logics inherent to the organizational system also affect parties during the mediation. For example, is a supervisor willing to listen to a subordinate commenting on his or her leadership style? And is a subordinate taking the risk of speaking up during the mediation, when this could have repercussions when parties are back at work?

To date, empirical field research on mediation in labor conflicts and especially on the influence of their hierarchical positions on parties' experiences and perceptions is lacking. To help fill this void, this article seeks to explore the influence of the party's hierarchical position on his or her satisfaction with the mediation and on his or her uncertainty about the mediation. Is mediation able to obtain a satisfactory outcome for both parties? Do both parties experience, to the same extent, uncertainty during the mediation, and does this uncertainty have the same consequences for subordinates and supervisors?

Research in work settings has already indicated that subordinates and supervisors differ with regard to the amount of situational uncertainty they experience (Chen, Brockner, and Greenberg 2003). This might be particularly true in conflict situations because facing hierarchical conflicts is more common for supervisors than for subordinates (Euwema 1992), and supervisors might feel less constrained in their behavior (Keltner, Gruenfeld, and Anderson 2003). Is this also true, however, when hierarchical conflicts are mediated by a third, neutral party? We explore these questions using data from real mediations involving hierarchical labor conflicts in the Netherlands.

Conflicts between supervisor and subordinate, or employer and employee, can be defined as hierarchical conflicts. By definition, parties involved in this kind of conflict occupy a different hierarchical position and therefore usually differ in their positional power. Differences in hierarchical position and power have considerable implications for cognitions and emotions (Hegtvedt 1990; Lawler and Yoon 1993), as well as for behaviors and interactions (Ford and Johnson 1998; Fitness 2000; Anderson and Berdahl 2002).

Studies focusing on hierarchical conflicts found that subordinates and supervisors handle conflict differently: whereas supervisors tend to force or to confront offenders in order to solve the situation, subordinates are likely to withdraw or to avoid (Van de Vliert, Euwema, and Huismans 1995; Fitness 2000; Aquino, Tripp, and Bies 2006). These differences are in line with the approach/inhibition theory of power (Keltner, Gruenfeld, and Anderson 2003), which states that low‐power individuals feel restricted in their behavior, whereas high‐power individuals are more likely to feel free to behave in ways they like and to fear fewer negative consequences.

Parties involved in hierarchical conflict not only behave differently, they also perceive and experience conflict in a different way (Fitness 2000), demonstrating strong asymmetries between the parties. When involved in conflict and after an anger episode, supervisors feel in general satisfied, perceiving the conflict as resolved, thereby experiencing low levels of resentment toward their offenders. Subordinates, in contrast, often feel dissatisfied and unacknowledged by their supervisors. In addition, they perceive the conflict as unresolved, experiencing moderate to high levels of resentment toward their offenders (Fitness 2000). A possible explanation for this difference might be that subordinates, unlike supervisors, have fewer resources at their disposal — such as money, knowledge, institutional support, and social support to cope with the situation (Keltner, Gruenfeld, and Anderson 2003). In addition, subordinates are more likely to be confronted with personal humiliations (Fitness 2000). Moreover, negotiation studies reveal that low‐power people are more affected by the angry behavior of a powerful counterpart compared with the same behavior shown by a person low in power. Supervisors, in contrast, seem to be less affected by the angry behavior of a lower‐power person (Sinaceur and Tiedens 2006; Van Kleef et al. 2008).

Given this influence of hierarchy and power, we expect that parties will approach the mediation differently, with different perceptions and emotions, and, as a consequence, their expectations and perceptions of the mediation process as well as of the mediator will most likely differ. So far, it is quite unclear to what extent mediation is an effective way to solve hierarchical conflicts and whether it provides both parties with a satisfying outcome. Theoretically, mediation is implemented to balance power and to minimize asymmetries (Wiseman and Poitras 2002); however, because mediation is a temporary intervention, mechanisms underlying the daily organizational hierarchy might affect the mediation process.

Surprisingly, we did not find any empirical (field) studies addressing this issue. Based on previous research on hierarchy and power, we presume that holding these different positions will affect not only the perceptions and behaviors of the parties in the conflict but also their perception and appreciation of the mediation. As we have seen that supervisors are usually more satisfied with the outcomes of a hierarchical conflict and seem to be less sensitive to the process, this might also be true for the mediation. Although a goal of mediation is to reduce power differences between parties and give voice to both, this does not necessarily result in equal satisfaction with the mediation process and outcomes. Possibly hierarchical structures will keep on playing a role during the mediation. Knowing that conflicts are often negative and offensive and that subordinates are more affected by the interactions they encounter than supervisors, it might take more to satisfy subordinates than supervisors involved in mediation. Consequently, our first hypothesis is that subordinates involved in a hierarchical conflict will feel less satisfied with the mediation than will supervisors.

Like the pervasive influence that it has on behavior and perceptions, hierarchical position also affects the experience of uncertainty. In a hierarchical working relationship, it is likely that the person higher in the hierarchy also has more status in the relationship. In general, high‐status people are likely to experience more certainty because of their higher degrees of control, whereas low‐status people are more likely to experience uncertainty because of their dependency on higher‐status people (Fiske and Depret 1996). These feelings are reflected in the slouched protective stance of subordinates and the assertive behavior of supervisors (Fitness 2000; Theorell et al. 2000). This aligns well with research indicating that in contrast with supervisors, subordinates often use fairness information (information about distributive outcomes as well as procedural justice) to decrease their worries of being exploited by high‐status individuals (Lind and Van den Bos 2002; Chen, Brockner, and Greenberg 2003).

Being involved in a hierarchical labor conflict is usually threatening or stressful to all parties involved (Euwema 1992), giving rise to uncertainty over both relational and outcome issues (Van de Vliert, Euwema, and Huismans 1995). In addition, the unfamiliarity of the mediation setting might provoke feelings of uncertainty about the mediator, the other party, and/or the mediation process itself. We can assume that subordinates will experience more uncertainty about the mediation than will their supervisors for several reasons. The first reason is related to the potential consequences of the mediation. For subordinates the outcome of the mediation may represent the prospect of job loss and decline in financial resources, and the possible frustration of their needs related to social participation and recognition (Jahoda 1982). This is less true for supervisors. Second, supervisors might have more experience with mediation than a subordinate: whereas for a subordinate mediation may represent a new (unknown) situation stimulating feelings of uncertainty, a supervisor may participate in mediation whenever a serious conflict arises if it is a standard procedure in that company. Consequently, Hypothesis Two reads as follows: compared with supervisors, subordinates experience more uncertainty about the mediation including uncertainty about the process, the mediator, and the other party.

Conflicts are unpredictable, dynamic situations, and therefore usually create uncertainties. Social psychological theories and studies (Festinger 1954; Lopes 1987; Hogg 2000) show that people have a fundamental need to feel certain things about the world and their place within it. Uncertainty about one's attitudes, feelings, perceptions, as well as about one's relationship to other people is generally aversive because it deprives one of confidence in how to behave and what to expect from the physical and social environment (Lopes 1987; Hogg and Mullin 1999). Confronted with uncertainty, people feel anxious about their ability to control their immediate environment. Because the mediation setting may give rise to feelings of uncertainty about the (new) mediation setting, the mediator, and/or the other party involved, and because uncertainty itself results in more negative evaluations (McGraw, Hasecke, and Conger 2003), our third hypothesis is that high levels of uncertainty about the mediation (the mediation process, the mediator, and the other party) will relate negatively to satisfaction with the mediation.

Figure One shows the expected relations we discussed so far: in comparison with supervisors, subordinates will be less satisfied with the mediation (Hypothesis One) and will experience more uncertainty about it (Hypothesis Two). In addition, uncertainty about the mediation is negatively related to satisfaction with the mediation (Hypothesis Three). This pattern of assumptions also suggests that hierarchical position may moderate the relationship between uncertainty about the mediation and satisfaction with the mediation. Particularly because subordinates perceive themselves as having less control than supervisors (Theorell et al. 2000), often feel vulnerable and insecure (Yukl and Falbe 1990), and have a greater need for transparency and intelligibility than supervisors (Lind and Tyler 1988), uncertainty can be predicted to have a stronger negative impact on them than on supervisors. If so, an aversive response is more likely to occur, resulting in lower satisfaction with the mediation.

We would expect this to be less true for supervisors. As the occupation of a higher hierarchical position provides them with more control, choices, and alternatives (Theorell et al. 2000), the uncertainty about the mediation can be expected to affect them to a lesser extent and therefore lead to fewer negative consequences for satisfaction with the mediation. Hypothesis Four thus reads as follows: the negative relationship between uncertainty about the mediation and satisfaction with the mediation will be stronger for subordinates when compared with supervisors.

In 2007 we carried out a field study that involved participants in real mediation cases concerning hierarchical conflicts between subordinates and supervisors. Thanks to the cooperation of a Dutch mediation company, we were able to contact parties who had been involved in labor mediation, and we were given access to the transcripts of the mediation sessions, which indicated the issues at stake as well as the degree of escalation. We selected at random fifty mediation cases, five cases for each of ten experienced mediators, with one hundred participants in total. Within a minimum of four weeks after finishing the mediation, surveys were mailed to participants, accompanied with a return envelope as well as letters signed by the chief executive officer of the cooperating organization and the researchers. Participation was voluntary and confidential.

Almost fifty percent of the respondents (N = 100; 49 percent) returned the questionnaire, of which twenty‐three were supervisors (thirteen men and ten women) and twenty‐six were subordinates (twelve men and fourteen women). The mean age was forty‐six years (standard deviation = 8.50). Almost as many women (49 percent; N = 24) as men (51 percent; N = 25) participated in our study. Coding the conflicts revealed that multiple issues were involved, including personalized conflict issues, suggesting a high level of escalation. Twelve percent of the responses involved the subordinate and the supervisor from the same mediation.

In this study hierarchical position has been operationalized as the occupation of a position of authority or a certain formal position in relation to the other party involved in the mediation (Item: “What is your relation with the other party?”). The conflict relation was described either as supervisor and subordinate (in large organizations) or employer and employee (in small organizations). Other potential power bases as well as the perception of power were left out of consideration.

Satisfaction with mediation is best measured multidimensionally (Conlon 2005; Herrman, Hollett, and Gale 2006). Therefore, we selected eight items covering the two major components of satisfaction: satisfaction with the mediation outcome (McGillicuddy et al. 1991) and satisfaction with the mediation process (Wissler 2002). All these items together form one general measure of satisfaction (Cronbach's alpha = 0.91). Responses for the different items were coded on a five‐point scale (1 = strongly disagree; 5 = strongly agree), a high score indicating a high level of agreement with the statements.

The first subscale, satisfaction with the mediation outcome, was measured through four items (Cronbach's alpha = 0.80). Three items were developed by Jean Poitras and Aurélia Le Tareau (2009) (“I am happy with the solution we came to,”“The settlement of the conflict was satisfactory to me,” and “I am satisfied with the agreement we reached”). The fourth item considered the extent to which the conflict had been resolved thanks to the mediation.

The second subscale, satisfaction with the mediation process, was measured through four items: two items measured procedural justice perceptions and the other two, satisfaction with the mediator (Cronbach's alpha = 0.90). Because perceptions of procedural justice occupy an important place in evaluating the mediation process (Pruitt et al. 1993; Poitras and Le Tareau 2009), we incorporated, in line with work by Jason Colquitt (2001), the following items in our scale: “Have procedures been applied consistently in the mediation?” and “Were the mediation procedures free of bias?” In addition, we used two items measuring the satisfaction with the mediator (“I assume that the mediator treated me in a consistent and predictable way” and “The motives and intentions of the mediator were good”).

Although the two constructs — satisfaction with the mediation outcome and with the process — can be differentiated meaningfully, they can also be perceived as a coherent scale because satisfaction with the mediation is determined by the mediation process as well as its outcome. This is also reflected in the intercorrelation between both scales (r[correlation coefficient] = 0.52; p[probability or significance] < 0.001). In this article, we report results for the overall scale as well as for the subscales.

Instruments to assess situational uncertainty are context specific. For example, one can experience uncertainty about performance standards and appropriate behaviors (Diekmann, Barsness, and Sondak 2004) or informational uncertainty when communication is computer mediated rather than face‐to‐face (Tangirala and Alge 2006). In the mediation context, several factors may give rise to uncertainty (the newness of the mediation context, the confrontation with the counterpart, the mediator, whom one has likely never met before). No scale is known to us measuring the amount of uncertainty about the mediation. Therefore, inspired by Kristina Diekmann, Zoe Barsness, and Harris Sondak (2004), we constructed a scale with three statements: “I felt uncertain about the mediation process,”“I felt uncertain about the behavior of the mediator,” and “I felt uncertain about the behavior of the other party.” Results show a satisfactory internal consistency of 0.78. Respondents had to indicate their level of agreement on a five‐point scale ranging from 1 (strongly disagree) to 5 (strongly agree), a high score indicating a high level of uncertainty about the mediation. Because we used real mediation cases, we did not manipulate uncertainty experimentally.

We executed data management, exploratory factor analysis, and other analyses using the Statistical Package for the Social Sciences (SPSS) 17.0. We left gender and age out of further analyses because they showed no relationship to other research variables. To test our hypotheses, we used MANOVA analysis as well as regression analyses.

To ensure that the items used in this study related to the intended constructs, we performed an exploratory factor analysis. The results (see Table One) showed an optimal structure with items loading significantly on only one single factor of the three factors obtained: satisfaction with the mediation outcome, satisfaction with the mediation process, and uncertainty about the mediation.

Table Two shows mean scores, standard deviations, and intercorrelations among research variables. We include correlations significant at p < 0.10 given the relatively small sample size (N = 49).

To test Hypotheses One and Two, we conducted a MANOVA analysis with hierarchical position as the independent variable and satisfaction with the mediation (Hypothesis One) and uncertainty about the mediation (Hypothesis Two) as dependent variables. Both hypotheses were supported. As expected, we found an effect for hierarchical position on satisfaction with the mediation (F[degree of variance][1, 48] = 3.23; p[probability or significance] < 0.10): supervisors were more satisfied with the mediation than were subordinates (mean = 3.43 versus mean = 2.87) (Hypothesis One).

Testing this relation for the subscales, we found that this effect can be explained mainly by the effect of hierarchical position on satisfaction with the mediation process: (F[1, 48] = 6.37; p < 0.05). Supervisors reported greater satisfaction with the mediation process than did the subordinates (mean = 4.08 versus mean = 3.35). Satisfaction with the mediation outcome, in contrast, was not influenced by hierarchical position (F[1, 48] = 0.83; not significant).

Second, we found that hierarchical position had a significant effect on uncertainty about the mediation (F[1, 48] = 8, 53; p < 0.01). In line with Hypothesis Two, supervisors reported experiencing less uncertainty during the mediation than did subordinates (mean = 2.22 versus mean = 3.13). To investigate the relationship between uncertainty and satisfaction with the mediation (Hypothesis Three), we regressed uncertainty on satisfaction with the mediation. Our results indicate that uncertainty about the mediation is negatively associated with satisfaction with both the mediation process and outcomes (regression coefficient = −0.39; p < 0.01). Interestingly, however, we found a stronger correlation between uncertainty and satisfaction with the mediation process (correlation coefficient = −0.40; p < 0.01) than between uncertainty and satisfaction with the mediation outcome (correlation coefficient = −0.30; p < 0.05).

Finally, to examine the moderating role of hierarchical position in the relationship between uncertainty about the mediation and satisfaction with the mediation (Hypothesis Four), we used multiple regression analysis. To correct for possible multicollinearity, we centered each of the independent variables by subtracting the sample mean from each variable before generating an interaction term (Aiken, West, and Reno 1991). The interaction term was formed by multiplying the two centered predictor variables: uncertainty about the mediation and hierarchical position. Table Three shows a significant and positive interaction term, implying that the hierarchical position one occupies affects the strength and/or direction of the relation between the uncertainty about the mediation and satisfaction with the mediation.

Table Three indicates also that next to the influence of the interaction term, satisfaction with the mediation is affected by the level of uncertainty. When including successively and separately satisfaction with the outcome and satisfaction with the mediation process as dependent variables, the results remain comparable. For both scales, the interaction term is significant: satisfaction with the outcome (regression coefficient = 0.25; p < 0.10) and satisfaction with the process (regression coefficient = 0.24; p < 0.10), indicating the moderating role of hierarchical position.

Figure Two shows the relation between uncertainty about the mediation and satisfaction with the mediation for subordinates and supervisors. Whereas subordinates' satisfaction was affected in a negative way by the amount of uncertainty experienced about the mediation, supervisors' satisfaction was not. This pattern is also suggested by the correlations: while uncertainty and satisfaction with the mediation are correlated for subordinates (correlation coefficient = −0.51; p < 0.01), we did not find this to be the case for supervisors (correlation coefficient = 0.04; not significant). These data suggest that it is necessary to take into account the amount of uncertainty about the mediation and the hierarchical position of the parties when evaluating both parties' satisfaction with the mediation and the success of the mediation itself.

This article explores the influence of hierarchical position on satisfaction with the mediation and uncertainty about the mediation, and its potential moderating role on the relationship between satisfaction and uncertainty. The study has deviated from existing literature and research in our attempt to combine insights from three theoretical perspectives — mediation literature, theories of hierarchy and power, and theories of uncertainty — into one research model. We have tested this model using real mediation cases dealing with hierarchical labor conflicts. Consequently, the present research offers several new insights and adds to both conflict management theory as well as mediation theory and practice in a variety of ways.

First, in our study supervisors as well as subordinates expressed satisfaction with the mediation, although supervisors were more satisfied with the mediation than were their subordinates. Interestingly, this difference was not related to the mediation outcomes but to the mediation process. Second, hierarchical position affected the way the parties experienced and perceived the mediation. In addition to feeling less satisfied with the mediation, subordinates also felt more uncertain about the mediation. This aligns well with research indicating that compared with higher‐status workers, lower‐status workers or subordinates feel more uncertain and vulnerable (Fiske and Depret 1996).

The most interesting outcome might be the interaction: when confronted with uncertainty about the mediation, subordinates' satisfaction with the mediation especially was affected in a negative way. A possible explanation for this interaction effect can be found in the different appraisal of the mediation by subordinates and supervisors (Lazarus and Folkman 1984). In comparison with supervisors, subordinates have, in general, less access to resources to cope with the situation, which encourages a perception that they possess little control (Theorell et al. 2000). The confrontation with a (new) mediation situation may give rise to even more feelings of emotional turbulence or uncertainty, affecting their satisfaction with the mediation in a negative way because they do not have (enough) resources to buffer the impact of this uncertainty. This may be especially true in (potential) exit‐mediations during which the job and related resources are at stake (Jahoda 1982).

For supervisors, uncertainty about the mediation is less prevalent and has less negative consequences. This might be because of their greater access to resources and higher degrees of control. In this sense, occupying a higher hierarchical position — and the resources or control associated with this position — seems to buffer the negative effects of uncertainty about the mediation on the supervisor's satisfaction with the mediation.

These outcomes are important for further development of mediation theory. In the field of conflict at work, hierarchical differences are often present in mediation. This study underscores that the mediation process is different for higher‐status and lower‐status parties, and this should be taken into account when it comes to understanding particularly the mediation process and parties' satisfaction with it.

By indicating that hierarchical position keeps on playing a role during mediation, affecting perceptions of the mediation process and its evaluation, the results give us a better understanding of the psychological processes related to occupying a particular hierarchical position within an organization. These results have several practical implications.

A first implication is that mediating in hierarchical conflicts is potentially more complex than mediating in nonhierarchical conflicts. Consequently, within organizations a question should arise as to when and under what conditions mediation is an appropriate intervention for hierarchical conflicts. It is clear that in these cases, the complexity of a mediator's job stems especially from the fact that he or she will be confronted with two parties who perceive and experience the mediation differently and hold different expectations. The mediator should pay special attention to the power dynamics during the intake to recognize the power differences and work with these rather than ignoring them or trying to eliminate them completely (Quinn et al. 2006).

The mediator can elaborate the different views to promote mutual understanding of the relative positions and the accompanying expectations or feelings. As the experience of uncertainty seems to have the most negative effects on subordinates, it is fruitful to provide them with (extra) specific information on the mediation process, on possible outcomes, and on the role of the mediator. This can help diminish their sense that they lack control and can enhance their confidence (Spreitzer and Mishra 1999) and will also reduce feelings of uncertainty (McGraw, Hasecke, and Conger 2003).

Furthermore, a mediator should explore the different sources of uncertainty and address these. This can be done during the mediation, but preferably during the intake before the first joint session takes place. Precaucusing can be helpful in this respect, to create an open, transparent, and safe climate. During this precaucus, the mediator will be able to inform the parties in a structured manner about the mediation context, the mediator, the goals and procedures, as well as the alternatives. Also, the power relationship and fears might be addressed explicitly. Consequently, it is clear that mediators should not approach conflicts with a “one‐size‐fits‐all” approach.

Having said this, it would be a fruitful avenue for future research to examine the role and use of caucus in hierarchical conflicts. Another possible area for research is the level of trust in the mediator (Poitras 2009) as well as the level of trust between supervisor and subordinate during the mediation. Because trust is a psychological state comprising a willingness to be vulnerable based upon having positive expectations of the intentions or behavior of another (Rousseau et al. 1998), building trust could buffer the potential negative effects of uncertainty or anxiety and fear (Wu and Laws 2003).

Another area for future research would be to examine the role of procedural justice perceptions. According to the uncertainty management theory, procedural justice perceptions may buffer the negative consequences of uncertainty (Lind and Van den Bos 2002). A mediator's decision to tailor the mediation process to the specific needs of both parties involved (e.g., taking into account subordinates' higher need for information) may yield benefits not only for the organization (reduced litigation costs) but also for the parties in providing them with a satisfactory outcome.

The current study adds to earlier mediation research in several ways. First, we advanced a framework that takes into account the hierarchical position as well as uncertainty about the mediation and mediation satisfaction. Second, a methodological improvement of our study is that we designed both a multidimensional scale to measure satisfaction with the mediation and a scale to measure uncertainty associated with the mediation process. Both scales have an acceptable internal consistency.

Some limitations inherent to this study need to be considered. First, the sample is limited, by its size, the cultural context of the Netherlands, and the amount of conflict escalation (escalated conflicts). Replication in different contexts, societies, and types of (labor) conflicts is needed. Second, we used a cross‐sectional design that limits conclusions about causal relationships between research variables. Third, we were not able to distinguish hierarchical position from status or power. Although in this study the supervisor has been considered as the powerful one, it is not impossible that the person lower in the hierarchy might not actually be the more powerful party in the conflict. This is clearly an issue that requires further exploration. Fourth, we did not have a control group at our disposal. A promising line of study would be to compare our findings with research stemming from nonhierarchical conflicts. Finally, because we relied on respondents' perceptions to assess uncertainty as well as satisfaction with the mediation, common method variance problems may have been induced (McClelland and Judd 1993). A related issue is the possibility that relationships between variables may be influenced by personality factors, most notably personal uncertainty or self‐efficacy.

Although substantial attention has been paid to mediation effectiveness in terms of agreements or settlements, it is largely unclear to what extent hierarchical position affects the mediation process and the perceptions of the disputing parties involved. The strength of this study lies in demonstrating that parties who occupy a different hierarchical position experience (uncertainty) and perceive the mediation differently (satisfaction). Particularly for subordinates, the confrontation with uncertainty about the mediation is detrimental. With a growing interest among organizations in alternative ways of dealing with conflict, further research on mediation in hierarchical conflicts is worth investing in.