INTRODUCTION
The interest-focused model has suggested that conflict strategy is determined by people's concern over their own interests and those of others (Pruitt & Carnevale, Reference Pruitt and Carnevale1993; Rahim, Reference Rahim1983; Ruble & Thomas, Reference Ruble and Thomas1976). However, as conflicts or disputes often occur between parties with prior interactions or existing social bonds, people may take these relationships into account when choosing an approach to conflict. For example, in a pioneering work on conflict avoidance, Leung (Reference Leung1988) delineates a scenario in which a seller promises a buyer to refund his/her purchase price for an appliance if it broke within a week but then declines to do so. The two disputants are described as either friends working for the same company or as strangers. Respondents in the United States and China are asked to imagine themselves as the buyer and to decide their likelihood of suing the seller to obtain the promised refund. Leung (Reference Leung1988) finds that, if the buyer and the seller are friends, Chinese respondents are more likely than American ones to avoid conflict in trying to resolve the matter.
Subsequent researchers have usually defined the relationship between the parties to a conflict in the scenario studied. For example, someone's ideas are stolen by his/her colleague or boss during a meeting (Friedman, Chi, & Liu, Reference Friedman, Chi and Liu2006), someone is bewildered by his/her classmate's failure to repay a loan (Zhang, Zhang, & Wang, Reference Zhang, Zhang, Wang, Leung, Chiu and Hong2011), and an employee is considering whether to raise questions about his/her supervisor who made an inappropriate decision (Zhang, Wei, Chao, & Zheng, Reference Zhang, Wei, Chao and Zheng2017). In such situations involving the potential for conflict, individuals usually consider whether their relationships with one another will be affected. In other words, before conflict becomes overt, the anticipation of conflict predisposes people to take actions to avoid it.
The widely adopted interest-focused theories highlight individuals’ concern over economic outcomes when determining conflict resolutions (e.g., Rahim, Reference Rahim1983; Ruble & Thomas, Reference Ruble and Thomas1976). Nevertheless, in some cultures or situations, maintaining interpersonal harmony between the parties involved is the core interest. For example, handling potential conflict is risky in East Asia, where interpersonal relationships are considered paramount (Leung, Reference Leung1988; Leung, Koch, & Lu, Reference Leung, Koch and Lu2002; Peng & Tjosvold, Reference Peng and Tjosvold2011). Similarly, Chinese people are more likely than people in Western countries to expect worse relational outcomes from direct confrontations and to prefer to remain silent in conflictual situations (Friedman et al., Reference Friedman, Chi and Liu2006).
We argue that, when interpersonal relationships are highly valued, individuals prefer avoidance to confrontation even at the risk of a financial loss, in the belief that confrontation might disrupt interpersonal harmony between the parties involved. As a result, avoidance originates out of concern for an interpersonal relationship, rather than individual interests. In other words, individuals tend to avoid conflict if they have a negative anticipation, a cognitive assessment of the cost to a relationship from engaging in conflict (Zhang et al., Reference Zhang, Zhang, Wang, Leung, Chiu and Hong2011).
In this research, we examine which factors engender negative anticipation by individuals, which leads them to adopt a strategy of avoidance in conflictual situations. We identify individual values and interpersonal contexts that are relevant to the concern over harmony and investigate their effects on negative anticipation as well as avoidance. The value placed on superficial harmony treats the maintenance of that harmony as the socially desirable method of protecting oneself from being socially sanctioned (Leung, Reference Leung, Christopher and Erez1997; Leung et al., Reference Leung, Koch and Lu2002; Wei, Zhang, & Chen, Reference Wei, Zhang and Chen2015). This notion is first developed in Chinese culture and has spread to other cultures (e.g., Leung, Brew, Zhang, & Zhang, Reference Leung, Brew, Zhang and Zhang2011; Lim, Reference Lim2009). This value reflects a defensively relational motive, which influences individual perceptions and interpretations of conflictual situations and determines the choice of appropriate approaches to conflict.
This research contributes to the existing literature primarily in two ways. First, we theorize that people avoid conflict for the sake of relational concerns, which extends the literature on conflict management. Dual-concern models generally consider avoidance the least effective strategy for addressing conflict, because it indicates individuals’ low self-concern and low other-concern. We argue that, in handling conflict, individuals consider not only their economic interests and those of the other party but also their mutual relationship. Although the existing literature examines how overt conflict is handled, our theorizing extends to people's responses before conflict becomes explicit. Because the emphasis is on maintaining the relationship, or harmony, people choose to avoid conflict because they anticipate that a direct confrontation will damage the relationship. By highlighting individuals’ social-relational concerns, this research at least provides a complementary perspective to dual-concern models in explaining conflict avoidance. Second, by strengthening the relationship between superficial harmony and negative anticipation, we extend the functions of superficial harmony in conflict management. Leung (Reference Leung, Christopher and Erez1997) suggests that superficial harmony that is driven by disintegration avoidance and genuine harmony that is driven by harmony enhancement relate to conflict resolution differently. Subsequently, Leung and colleagues (Reference Leung, Brew, Zhang and Zhang2011) confirm that disintegration avoidance is positively related to conflict avoidance, but harmony enhancement is not related to conflict avoidance. Our findings further explain why superficial harmony results in conflict avoidance. Because of their goal of maintaining superficial harmony, people tend to have a negative assessment of the relational outcomes from engaging in conflict, which leads to avoidance. In short, this research provides a social-relational perspective for understanding why people prefer to avoid conflict, rather than engage in it, focusing on the workplace.
THEORETICAL BACKGROUND AND HYPOTHESES
Avoidance as a Response to Relationship Concerns
People experience tension because of real or perceived differences in interests, perspectives, behaviors, or values, and these tensions result in latent and explicit conflict (De Dreu & Weingart, Reference De Dreu and Weingart2003). The existing research assumes that social interaction is regulated by self-interested actors scanning, processing, and developing relationships based on their economic interests. For example, the dual-concern model (Pruitt & Carnevale, Reference Pruitt and Carnevale1993; Rahim, Reference Rahim1983; Ruble & Thomas, Reference Ruble and Thomas1976) posits that how people handle a conflict is determined by concern over the outcome for oneself and the other party, which leads to five conflict-handling styles: dominating, integrating, compromising, obliging, and avoiding. Among the five approaches, avoidance is regarded as the least effective and the least socially desirable, because it reflects low self-concern and low other-concern.
The interest-focused theories highlight the economic interests of both parties and focus on conflicting parties’ assessment of the potential gain or loss of tangible resources, such as money, labor, time, or social resources such as status or position (Pruitt & Carnevale, Reference Pruitt and Carnevale1993). However, in some cultures or situations, interpersonal relationships or social harmony are at the core of individuals’ interest, and thus individuals are concerned with the maintenance of their relationship with the other party (e.g., Ohbuchi & Takahashi, Reference Ohbuchi and Takahashi1994). Because conflict often involves parties with prior interactional experience or social relationship, individuals in these cultures consider the existing social connections when deciding how to deal with conflicts. The different types of conflict, such as task conflict and relationship conflict (e.g., Jehn, Reference Jehn1995), are highly related (r = 0.54; De Deru & Weingart, Reference De Dreu and Weingart2003), so resolving a task conflict inappropriately may lead to a relationship conflict. This is particularly likely in Chinese culture, in which task and relationship conflicts are closely intertwined (Zhang & Hou, Reference Zhang and Hou2012).
This concern over the preservation of a relationship causes individuals to believe that handling conflict is inseparable from their interpersonal relationships. Chinese culture emphasizes interpersonal harmony, thus adopting a direct approach to conflict is usually at odds with the notion of preserving a relationship (Yang, Reference Yang1981). Compared to people in Western countries, people in Eastern countries have a greater preference for conflict avoidance so as to preserve relationships and minimize animosity (Gelfand, Nishii, Holcombe, Dyer, Ohbuchi, & Fukuno, Reference Gelfand, Nishii, Holcombe, Dyer, Ohbuchi and Fukuno2001; Leung, 1988; Morris et al., Reference Morris, Williams, Leung, Bhatnagar, Li, Kondo, Luo and Hu1998; Ohbuchi & Takahashi, Reference Ohbuchi and Takahashi1994; Trubisky, Ting-Toomey, & Lin, Reference Trubisky, Ting-Toomey and Lin1991). Similarly, Chinese people score higher than Australians regarding the intention to maintain harmony (Leung et al., Reference Leung, Brew, Zhang and Zhang2011). People who are concerned about interpersonal relationship are often sensitive to the possible cost to a relationship of a confrontation and find it to conflict with the notion of harmony maintenance, so they tend to avoid conflict (Friedman et al., Reference Friedman, Chi and Liu2006).
The influence of social-relational concerns on the tendency to avoid social costs is suggested in the socially desirable responding (SDR) theory (Paulhus, Reference Paulhus, Braun, Jackson and Wiley2002). The communal motive behind SDR is a concern over a relationship with a defensive orientation (Crowne, Reference Crowne1979), which drives people to avoid being viewed in a negative light and thus to show agreeableness, dutifulness, or self-restraint in front of others. People high in such a motive focus on preventing relationship losses and managing others’ impressions of them (Lalwani, Shrum, & Chiu, Reference Lalwani, Shrum and Chiu2009). Because engaging in conflict may be regarded as an unfriendly gesture that sours an interpersonal relationship, people avoid doing so in order to be accepted as someone agreeable (Ohbuchi & Atsumi, Reference Ohbuchi and Atsumi2010) and to prevent provoking anger/hostility in others (Burris, Reference Burris2012). Therefore, we argue that conflict avoidance stems from individuals’ high concern for interpersonal relationships. If people anticipate that engaging in conflict could cause negative repercussions that harm relationships with others, they will prefer avoidance to protect themselves from possible relational losses.
Superficial Harmony and Negative Anticipation
Individuals’ endorsement of superficial harmony influences their perception and interpretation of conflictual situations. Here is a common scenario that many people have experienced in their daily lives. If someone borrows money from us without returning it in a timely way, we may either ask him/her to pay it back directly at the risk of offending or embarrassing him/her or avoid asking for the money back at our own expense. In this situation, people usually assess the consequences of using different approaches before taking action. Most people might anticipate that if they behave assertively (e.g., by asking for the money in this scenario), the other party might respond negatively, and thus their relationship would be damaged. The term ‘negative anticipation’ is used to encompass such concerns over potential relationship costs in conflictual situations (Zhang et al., Reference Zhang, Zhang, Wang, Leung, Chiu and Hong2011; Zhang et al., Reference Zhang, Wei, Chao and Zheng2017).
Negative anticipation is individuals’ assessment of conflictual situations driven by their concern over relationship costs, and it is influenced by individual values that reflect the fundamental motive guiding individuals’ perceptions and subsequent actions (Schwartz, Reference Schwartz1992). Values that emphasize social harmony (e.g., conformity, security, benevolence, and tradition) are found to be positively correlated with individuals’ tendency to behave in a socially desirable way (Schwartz, Verkasalo, Antonovsky, & Sagiv, Reference Schwartz, Verkasalo, Antonovsky and Sagiv1997), because these values express goals of self-restraint and avoiding violation of social expectations.
Because an individual value on superficial harmony is aligned with a self-protective and defensive orientation, accompanied by concern over relationships in conflictual situations, it reasonably leads to a high level of negative anticipation. In understanding Chinese harmony, Leung (Reference Leung, Christopher and Erez1997) distinguishes two types of values of harmony with different motives. Disintegration avoidance is motivated by the desire to avoid actions that strain a relationship, whereas harmony enhancement is motivated by the wish to engage in actions intended to strengthen a relationship. Harmony colored by these two values relates to conflict resolution differently. When individuals treat harmony as a means to prevent losses, they avoid conflict; however, if they consider harmony a desire to establish a long-term and mutually beneficial relationship, they resolve conflicts proactively. Consistent with this argument, Leung and colleagues (Reference Leung, Brew, Zhang and Zhang2011) find that disintegration avoidance is positively related to conflict avoidance, but harmony enhancement is not related to conflict avoidance; while disintegration avoidance is related to integrating negatively, harmony enhancement is related to integrating positively.
In a similar vein, Huang (Reference Huang1999) independently identifies two types of harmony: genuine harmony and superficial harmony. Although individuals interested in genuine harmony view others in a positive light and are active in their interaction, using sincerity, trust, and support, those interested in superficial harmony adopt a negative view of other people's reactions and tend to be cautious, such as isolating oneself from potential conflict (Huang, Reference Huang1999). In essence, disintegration avoidance reflects valuing Huang's superficial harmony with a defensive orientation that focuses on the negative consequences of a strained relationship (Leung, Reference Leung, Christopher and Erez1997; Leung et al., Reference Leung, Koch and Lu2002). Therefore, superficial harmony, not genuine harmony (i.e., harmony enhancement), fosters conflict avoidance because it leads to individuals’ negative anticipation.
In conflictual situations, superficial harmony motivates individuals to evaluate the interpersonal consequences of engaging in conflict in a negative way. People with a high level of superficial harmony are aware of potential tensions, negative emotions resulting from engaging in conflict, and possible relationship costs. Research suggests a positive relationship between superficial harmony and sensitivity to interpersonal risks as well as a tendency toward conflict avoidance (e.g., Huang, Reference Huang1999). Studies conducted in China, Australia (Leung et al., Reference Leung, Brew, Zhang and Zhang2011), Malaysia, and India (Lim, Reference Lim2009) also find that people with higher levels of superficial harmony are more likely to avoid conflict in order to protect themselves from relationship disruption and loss.
Particularly in workplace settings, where employees encounter problems that are created or continued by managers at a higher level, they face a latent conflict. Speaking up about the problems may undermine the supervisors’ authority (Liang, Farh, & Farh, Reference Liang, Farh and Farh2012) or cause them to become defensive (Burris, Reference Burris2012), whereas remaining silent is an avoidance strategy to prevent conflict with a supervisor (Huang, Van de Vliert, & Van der Vegt, Reference Huang, Van de Vliert and Van der Vegt2005). Whether employees decide to express their dissenting opinions depends on the anticipated consequences (e.g., Detert & Burris, Reference Detert and Burris2007), particularly their concern about the relationship with their supervisor (Wei et al., Reference Wei, Zhang and Chen2015; Zhang et al., Reference Zhang, Wei, Chao and Zheng2017). Remaining silent is more likely than exposing conflict to maintain a harmonious relationship with the supervisor or at least minimize animosity. Therefore, employees’ negative anticipation of a soured relationship with their supervisor causes their silence – a manifestation of conflict avoidance in voice situations.
Superficial harmony shapes employee perceptions of the risk from expressing dissent and from voicing any warnings about dysfunctional practices. Because dissent and warnings may give rise to conflicts over tasks and further devolve into relationship conflicts (Burris, Reference Burris2012; Liang et al., Reference Liang, Farh and Farh2012), people who value superficial harmony perceive more negative repercussions from such behaviors (Wei et al., Reference Wei, Zhang and Chen2015). This is because they are motivated to protect themselves from potential relational losses, which serves as a lens for them to pay attention to and interpret information that is relevant to relationship costs. Taking these arguments concerning general conflict and voice situations together, we propose that the more individuals endorse the value of superficial harmony, the more costly they perceive direct confrontation to be for interpersonal relationships.
Hypothesis 1: Individual value of superficial harmony will be positively related to negative anticipation.
Negative Anticipation and Conflict Avoidance
As a cognitive assessment of the consequences of confrontation, negative anticipation predisposes individuals to adopt a conflict avoidance strategy. Whenever people are concerned about the negative repercussions of engaging in conflict, they tend to avoid potential conflict so as to conceal bad news (Welch & Welch, Reference Welch and Welch2005), remain silent about organizational problems (Milliken, Morrison, & Hewlin, Reference Milliken, Morrison and Hewlin2003), or avoid open expression of their dissent (Peng & Tjosvold, Reference Peng and Tjosvold2011). As discussed earlier, the approach to conflict can be considered a way of presenting oneself in social situations. Individuals tend to present themselves in socially desirable ways. The communal motive to be socially desirable in one's responses is often accompanied by fear of negative evaluations by others (Watson & Friend, Reference Watson and Friend1969), and this fear of social rejection (Börger, Reference Börger2013) drives people to avoid expressing their dissent, especially when interacting with those in authority, such as a supervisor at work (Wei et al., Reference Wei, Zhang and Chen2015).
In general, Chinese people consider avoidance a means of protecting an interpersonal relationship from disruption (Tjosvold & Sun, Reference Tjosvold and Sun2002). When they witness people with whom they have a pre-existing relationship misbehaving and making inappropriate decisions, individuals often worry that openly discussing the problem (e.g., correcting someone else's misbehavior or pointing out a problem) may offend the person engaging in those activities and thus harm interpersonal harmony. The negative expectation of the outcome of a direct approach to conflict leads to a higher tendency toward avoidance, compared with people with a lower negative expectation (Friedman et al., Reference Friedman, Chi and Liu2006). Recent studies demonstrate that conflict avoidance stems from individuals’ anticipation of a negative response to confrontation by the other party (Zhang et al., Reference Zhang, Wei, Chao and Zheng2017).
Studies in organizational settings also confirm the relationship between negative anticipation and avoidance. People believe that openly expressing dissent is risky, since it may embarrass others at upper levels, disrupt interpersonal relationships, and cause managerial retaliation (e.g., Detert & Edmondson, Reference Detert and Edmondson2011; Milliken et al., Reference Milliken, Morrison and Hewlin2003; Wei et al., Reference Wei, Zhang and Chen2015). Because speaking up may contradict supervisors and spark conflict, employees decide whether to voice their views or remain silent, depending on their perception of the consequences of doing so. If they anticipate that speaking up will disrupt their relationship with their supervisor and may result in personal losses, they will choose to be silent (e.g., Detert & Burris, Reference Detert and Burris2007; Detert & Edmondson, Reference Detert and Edmondson2011; Milliken et al., Reference Milliken, Morrison and Hewlin2003). Wei and colleagues (Reference Wei, Zhang and Chen2015) provide direct evidence that employees are less likely to question or warn about problematic practices if they anticipate a high risk of relationship damage from doing so. Based on these arguments, we have the following hypothesis:
Hypothesis 2: Negative anticipation will be positively related to conflict avoidance.
We argue that superficial harmony has a significant relationship with negative anticipation (H1), and negative anticipation has a significant relationship with conflict avoidance (H2). Because a significant direct relationship between independent and dependent variables is not necessary for testing a potential mediator (e.g., MacKinnon, Lockwood, & Williams, Reference MacKinnon, Lockwood and Williams2004), we posit the following hypothesis:
Hypothesis 3: Superficial harmony will have an indirect relationship with conflict avoidance, mediated by negative anticipation.
Relationship Closeness as a Moderator
Research on the motivational bases of information processing and strategy in conflict and negotiations notes the interplay between people and situations (De Dreu & Carnevale, Reference De Dreu and Carnevale2003). In particular, motives rooted in people are more likely to affect cognition and behavior in situations in which such motives are more relevant. This rationale is also supported in the literature on SDR, such that the influence of individual values on perceptions and behaviors depends on the interpersonal context. For example, although people who value interdependence, belongingness, and harmonious relationships demonstrate more self-effacement to maintain social approval, this tendency ceases in domains where agentic traits are valued (Yik, Bond, & Paulhus, Reference Yik, Bond and Paulhus1998). In organizations, the relationship between employees’ value of superficial harmony and their perceived risk from speaking up weakens when they are in a group where they can freely express doubt (Wei et al., Reference Wei, Zhang and Chen2015).
So, we propose that the relationship closeness between the parties involved in a conflictual situation moderates the effects of superficial harmony on negative anticipation and conflict avoidance. The closeness of the relationship between oneself and others influences how people react to one another in social interactions (Hwang, Reference Hwang2000). Organizational researchers define relationship closeness as a high level of trust and positive affect between partners (Chen & Chen, Reference Chen and Chen2004; Chen & Peng, Reference Chen and Peng2008). The trust in a close relationship evolves from past experience and prior interactions and is accompanied by feelings of confidence and security (Rempel, Holmes, & Zanna, Reference Rempel, Holmes and Zanna1985). These feelings decrease the salience of the goal to strive for social approval. In addition, trust embedded in a close relationship reduces interpersonal uncertainty in conflictual situations (Leung, Reference Leung1988), diminishing the need to protect oneself from negative evaluations by others. Employees who perceive trust from the supervisor feel safer about expressing their ideas that challenge superiors (Detert & Burris, Reference Detert and Burris2007).
The positive affect derived from a close relationship broadens individuals’ thought-action repertoire and enables them to be approach oriented (Fredrickson, Reference Fredrickson2001). These effects counter the defensive and protective orientations implied by superficial harmony in situations that involve potential conflict and uncertain interpersonal outcomes. For example, positive affects experienced and displayed in social interactions assuage people's concern over exposing their vulnerabilities to others, which made them less concerned about avoiding ego-threatening interactions and more likely to reach out to obtain ideas and share their own ideas with others (Anderson & Thompson, Reference Anderson and Thompson2004). Through reducing the concern over disruption to a relationship and enhancing an approach orientation in social interactions, a close relationship can mitigate the relationship between superficial harmony and negative anticipation. Therefore, we propose the following hypothesis:
Hypothesis 4: Relationship closeness will negatively moderate the relationship between superficial harmony and negative anticipation, such that the positive relationship between superficial harmony and negative anticipation will be weaker when relationship closeness is high (vs. low).
Furthermore, as negative anticipation is positively related to avoidance, it links the interaction between superficial harmony and relationship closeness to conflict avoidance. This is consistent with the rationale that individuals present themselves in social interactions based on anticipation of the outcome of their actions in a specific situation (Paulhus & Trapnell, Reference Paulhus and Trapnell2008). When interacting with a close friend (vs. a stranger), individuals demonstrate less defensive and self-protective behaviors (Korchmaros & Kenny, Reference Korchmaros and Kenny2001) and are less likely to adopt an avoidance strategy because they perceive less risk (Leung, Reference Leung1988). Rather, they ask each other more questions and are more critical of their friends’ decisions (Shah & Jehn, Reference Shah and Jehn1993). We thus have the following moderated mediation hypothesis:
Hypothesis 5: Superficial harmony and relationship closeness will interact to influence conflict avoidance through negative anticipation. Specifically, the indirect effect of superficial harmony on avoidance through negative anticipation will be weaker when relationship closeness is high (vs. low).
Our theorized moderated mediation model is summarized in Figure 1, and we conducted one pilot study and two main studies to examine this model. As our two main studies focused on employees’ lack of voice at work, we first conducted a pilot study demonstrating the convergence of the lack of voice and conflict avoidance. Study 1 was a scenario-based study using both Chinese and American samples, testing H1, H2, and H3. In Study 2, we asked executives to rate their direct subordinates’ speaking up/avoidance at their companies and asked those subordinates to indicate whether they felt negative anticipation when speaking up to a supervisor and to evaluate their level of closeness with that supervisor. The data allow us to test the full model using path analysis in a multilevel structural equation modeling (MSEM) framework.
Figure 1. A moderated mediation model of conflict avoidance
STUDY 1
Methods
Participants and procedures
One hundred and eighty-nine (50% females) MBA students at a university in northern China took part in this study. Their average age was 35 years, with an average of 4 years of work experience. They participated in the study voluntarily while taking part in a training program at a business school.
This study was conducted in two phases. In phase 1, participants completed the scale on superficial harmony (Leung et al., Reference Leung, Brew, Zhang and Zhang2011). Two weeks later, they were presented with the conflict scenario that described an implicit conflict between a supervisor and a subordinate: the chief technology officer (supervisor) announced a new policy on project team management at a meeting, but the director (subordinate) knew that the policy was not workable.Footnote [1] In this scenario, subordinates who can identify the problems with the policy face possible conflict. Participants were asked to put themselves in the role of the subordinate and estimate how their supervisor, the chief technology officer, would respond if they expressed their concerns directly. Participants were also asked to indicate the likelihood that they would try to avoid conflict in this situation.
Measures
Superficial harmony was measured using eight items from the harmony scale developed by Leung and colleagues (Reference Leung, Brew, Zhang and Zhang2011). An example is ‘If a person does you favors, you must be tolerant with them in order to protect your own interests’. Participants were asked to rate the items using a five-point Likert scale (1 = strongly disagree, 5 = strongly agree). The Cronbach α was 0.76.
Negative anticipation was measured by adapting the scale developed by Zhang et al. (Reference Zhang, Zhang, Wang, Leung, Chiu and Hong2011). Two sample items were: ‘I would offend the chief technology officer’ and ‘The chief technology officer would find me uncooperative’. Participants were asked to indicate the extent to which they agree/disagree with each of the items, using a seven-point scale (1 = strongly disagree to 7 = strongly agree). The Cronbach α was 0.88.
Reluctance to voice was measured as a proxy of conflict avoidance in the scenario used in this study. A number of research and studies suggest that conflict avoidance is indicated in prohibitive voice situations. Prohibitive voice refers to employees’ expressions of concern about dysfunctional practices or incidents (Liang et al., Reference Liang, Farh and Farh2012), which usually fall within the responsibilities of higher-level managers (Burris, Reference Burris2012). When employees express voice regarding disagreement and confrontation with a supervisor, they may be regarded as disrupting the execution of tasks and uncommitted to the organizational goals (Morrison & Milliken, Reference Morrison and Milliken2000), sparking supervisors’ defensiveness and retaliation, in terms of lower performance rating (Burris, Reference Burris2012), unfavorable assignments, and/or a decreased chance of promotion (Milliken et al., Reference Milliken, Morrison and Hewlin2003). Because speaking up brings conflicts with a supervisor to the surface, employees are concerned about relationship costs and personal losses when they decide whether to express prohibitive voice (Liang et al., Reference Liang, Farh and Farh2012; Wei et al., Reference Wei, Zhang and Chen2015). And remaining silent is thus a tactic used to avoid expected conflict (Huang et al., Reference Huang, Van de Vliert and Van der Vegt2005).
We conducted a pilot study to examine the convergence of reluctance to voice and conflict avoidance. The results showed that the two variables were highly correlated, supporting our adoption of the measures for reluctance to voice as a proxy for conflict avoidance in the two main studies (see Appendix 1 for details of the pilot study).
In this study, reluctance to voice was measured using two questions on a seven-point Likert scale: ‘How likely are you to voice your opinions to the chief technology officer during the meeting’? (reverse coded) and ‘How likely are you to remain silent’? (1 = least unlikely, 7 = very likely). The Cronbach α was 0.95.
We controlled for demographic variables, including the age and gender (1 = male, 0 = female) of the participants.
Results and Discussion
We conducted confirmatory factor analyses using Mplus 6.0. A three-factor model was estimated and showed a good fit (χ2/df = 1.63, RMSEA = 0.058, CFI = 0.96, TLI = 0.95). The items all loaded significantly on the appropriate factors with loadings above 0.50, providing initial support for the discriminant validity of the scales. Comparing the baseline model with alternative models, we found that the three-factor model provided a better fit than three alternative models, including a two-factor model (superficial harmony was combined with negative anticipation, Δχ2 = 178.92, p < 0.01), and a single factor model (Δχ2 = 395.47, p < 0.01). We thus concluded that the three-factor model best represents the factor structure of the items.
Table 1 presents the means, standard deviations, and zero-order correlations of all variables involved. Superficial harmony showed significant correlations with negative anticipation (r = 0.36, p = 0.000) and conflict avoidance (r = 0.37, p = 0.000 respectively). Negative anticipation was significantly related to conflict avoidance (r = 0.59, p = 0.000). These results provided preliminary support for our hypotheses.
Table 1. Means, standard deviations, and correlations for Study 1
Notes: **p < 0.01 (two-tailed)
We conducted path analysis in Mplus 6.0 to estimate all the effects and test the mediation effects. We also calculated the indirect effects and estimated the Monte Carlo confidence interval using the open-source software R. Table 2 shows the results of the path analyses. Superficial harmony was positively related to negative anticipation (B = 0.80, S.E. = 0.15, p = 0.00), supporting H1. Negative anticipation was significantly related to conflict avoidance (B = 0.75, S.E. = 0.07, p = 0.00), supporting H2. The indirect effect of superficial harmony on conflict avoidance through negative anticipation was 0.60 (S.E. = 0.13), with the 95% Monte Carlo confidence interval of [0.35, 0.84]. Thus, H3 was supported.
Table 2. Mediation analyses for Study 1
Notes: SE = standard error; CI = confidence interval.
Study 1 found that whether employees avoid possible conflict by remaining silent depends on their anticipation of negative outcomes from confronting conflict. The result is consistent with existing findings. For example, Milliken and colleagues (Reference Milliken, Morrison and Hewlin2003) find that employees remain silent when they worry about negative consequences, such as being labeled a troublemaker or complainer, disrupting interpersonal relationships, and being punished at work. Other studies show that employees who believe that speaking up leads to managerial retaliation (e.g., Detert & Trevino, Reference Detert and Treviño2010) or anticipated high risks of relationship damage and career costs (Wei et al., Reference Wei, Zhang and Chen2015) tended to remain silent.
We also found that individuals who endorsed superficial harmony were more likely to perceive that direct confrontation may sour their relationship with the supervisor, and this negative anticipation led to conflict avoidance. Although Study 1 supported our H1, H2, and H3, the scenario-based design limited the external validity of the results. Study 2 was designed to examine the full model in real work settings.
STUDY 2
Methods
Participants and procedures
We contacted 132 business executives (one in each organization) when they were enrolled in an executive education program offered by a university in northern China. These executives were invited to participate in this project on a voluntary basis. All of them were ensured that the data collected would be kept confidential, and they would be provided with a summary report based on all participants’ results. Each executive was given one white envelope (with the questionnaire for him/herself) and five brown envelopes (with the questionnaires for his/her direct subordinates). The brown envelopes were marked with a number on the back corner for matching the data reported by subordinates and the evaluations from the supervisor. All the envelopes were preaddressed and postage paid.
The questionnaire for executives instructed them to assign a number to each of their direct subordinates (middle managers) and to specify each direct subordinate's demographic information, including gender, age, and job title. Then the executives indicated the extent to which each direct subordinate avoided direct confrontation when he/she found problems in the company. Finally, the executives reported their own demographic information, including gender, age, and tenure as an executive. All the executives were asked to mail back the completed questionnaire to the executive education program office independently. At the same time, they invited their direct subordinates to participate in this survey on a voluntary basis.
The direct subordinates were each given a brown envelope with a unique number, which contained a questionnaire. They were asked about the extent to which they noticed any existing or potential problems in their organization and to evaluate their negative anticipation from speaking up. As discussed earlier, in such situations, subordinates who identify problems with their supervisor's decisions face possible conflict. Because speaking up may bring conflict with the supervisor to the surface, remaining silent is regarded as conflict avoidance (Wei et al., Reference Wei, Zhang and Chen2015).
Next, these direct subordinates indicated the degree of their closeness with their supervisor (the executive). Finally, they responded to the superficial harmony scale and provided their demographic information. They were asked to individually mail the completed questionnaire in the preaddressed and postage paid envelope.
We received completed questionnaires from 88 top executives and 410 middle managers. We matched the data from all the executives and their subordinates by checking the number on the envelopes and comparing the demographic information reported by the middle managers to that provided by their top executive. Unmatched responses and groups with fewer than three middle managers were omitted in order to meet the requirement for minimum group size for hierarchical linear modeling. The valid data included responses from 81 executives and their 337 direct subordinates. The net response rates were 61.4% and 51.1% for executives and middle managers, respectively.
The average age of the 81 executives (90% males) was 40 years (S.D. = 5), with an average tenure of 102 months (S.D. = 73.06). The number of their direct subordinates ranged from 3 to 5, with an average of 4 (S.D. = 0.84). The average age of the 337 middle managers (62% males) was 36 years (S.D. = 7.38), with an average tenure of 77.5 months (S.D. = 70.22). Among them, 27.7% had a master's degree or above, and 55.1% had a bachelor's degree. The omitted responses from the executives showed no significant difference from the valid responses in their gender, age, or tenure. Similarly, no significant differences were found between the omitted responses from the middle managers and the valid responses with regard to their gender or age, but on average the latter had longer tenure than the former (F 1, 408 = 4.07, p = 0.04).
Measures
Superficial harmony was measured using the eight-item scale developed by Leung and colleagues (Reference Leung, Brew, Zhang and Zhang2011). Sample items include ‘You should not disturb your harmonious relationships with others, in order to avoid embarrassment in future encounters’, and ‘If people do you favors, you must be tolerant with them in order to protect your own interests’ (1 = strongly disagree, 5 = strongly agree). The measure was reliable (α = 0.81).
Relationship closeness was measured using the scale adapted from Chen and Peng (Reference Chen and Peng2008). Sample items include ‘we trust each other’, and ‘we can fully communicate at work’ (1=strongly disagree, 7=strongly agree). The measure was reliable (α = 0.89).
Negative anticipation was measured using the scale for the perceived risk of speaking up by Wei et al. (Reference Wei, Zhang and Chen2015), which was adapted from Zhang et al. (Reference Zhang, Zhang, Wang, Leung, Chiu and Hong2011) to fit the studied context. The instruction is as follows: ‘Suppose that you have opinions about your supervisor's decision that are different from his/hers, and you believe that the supervisor's decision will harm the organization. Generally speaking, if you express your views in this situation, what do you expect to happen? Please indicate the extent to which you agree with the following possible outcomes’. Sample items include ‘He/she would think that I don't respect him/her’, and ‘He/she would create problems for me in future jobs’ (1 = strongly disagree, 7 = strongly agree). The measure was reliable (α = 0.92).
Lack of prohibitive voice was measured as a proxy for conflict avoidance, by adapting the five-item scale for speaking up about organizational problems by Liang et al. (Reference Liang, Farh and Farh2012). Sample items include ‘Speak up honestly about problems that might cause serious loss to the work-unit, even if you dissent from the supervisor’, and ‘Dare to voice your opinions on things that might affect efficiency in the work-unit, even if that would embarrass the supervisor’ (1 = strongly disagree, 7 = strongly agree). As our pilot study showed the convergence between lack of prohibitive voice and conflict avoidance (see the Appendix for details), we reversed all items to indicate the level of avoidance. The scale was reliable (α = 0.89).
At the individual level, we controlled for subordinates’ gender and tenure. As subordinates may remain silent because they lack ideas about the discussed issues, we controlled for the extent to which they have ideas. Having ideas was measured using three items adapted from Frese, Teng, and Wijnen (Reference Frese, Teng and Wijnen1999) (α = 0.84). A sample item is ‘I often have ideas about inappropriate aspects of supervisory decisions’ (1 = strongly disagree, 7 = strongly agree). Meanwhile, to exclude the possibility that negative anticipation results from low levels of genuine harmony instead of high levels of superficial harmony, we controlled for subordinates’ value of genuine harmony, measured by the 13 items developed by Leung and colleagues (Reference Leung, Brew, Zhang and Zhang2011) (α = 0.81). A sample item is ‘Maintaining harmony with others, you are able to broaden your view of the world’.
Analytical Strategy
Due to the nested nature of the data (several middle managers have the same executive), we followed the recommendations of Preacher, Zyphur, and Zhang (Reference Preacher, Zyphur and Zhang2010) and used path analysis in a multilevel structural equation modeling (MSEM) framework in Mplus 6.0 to analyze the data. It allows for simultaneous estimation of the parameters in the moderated mediation model, offering more robust estimates of standard errors of parameters than piecemeal approaches. We ran ANOVA on negative anticipation and lack of prohibitive voice and found significant between-group variances for both variables, in that F 80, 256 = 2.04, p = 0.000, ICC (1) = 0.21 for negative anticipation, and F 80, 256 = 4.35, p = 0.000, ICC (1) = 0.46 for lack of prohibitive voice. These results justified MSEM as the appropriate analytical technique. We estimated the MSEM model with random intercepts and fixed slopes. At the individual level, middle managers’ negative anticipation was modeled as a function of their value of superficial harmony, their closeness with the executive, and the interaction of the two variables. Lack of prohibitive voice was modeled as a function of negative anticipation, and whether they have ideas about the discussed issues.
With regard to the first-stage moderation, we conducted multilevel simple slope analysis (Preacher, Curran, & Bauer, Reference Preacher, Curran and Bauer2006) to estimate the simple slopes at high (1 SD above mean) and low (1 SD below mean) levels of the moderator. To examine the moderated indirect effect, we calculated the indirect effects at high and low levels of the moderator and the difference in their effect, and estimated the Monte Carlo confidence intervals using the open-source software R.
Results and Discussion
Table 3 presents descriptive statistics and individual-level correlations among all variables. At the individual level, superficial harmony was significantly related to negative anticipation (r = 0.56, p = 0.000), and negative anticipation was also significantly related to lack of prohibitive voice (r = 0.47, p = 0.000). These results provided preliminary support for most of our hypotheses.
Table 3. Means, standard deviations, and correlations in Study 2
Notes: N = 337; *p < 0.05, **p < 0.01, ***p < 0.001 (two-tailed).
Table 4 presents the parameter estimates in the MSEM model described earlier. The fixed slope for superficial harmony on negative anticipation was significant (B 2 = 0.75, S.E. = 0.09, p = 0.000), supporting H1. The fixed slope for negative anticipation on lack of prohibitive voice was significant (B 6 = 0.36, S.E. = 0.09, p = 0.000), supporting H2. The indirect effect of superficial harmony on lack of prohibitive voice through negative anticipation was 0.27 (S.E. = 0.08), at the 95% bootstrapping confidence interval (CI) of [0.11, 0.43]. Thus H3 was supported.
Table 4. Unstandardized coefficients of MSEM Model
Notes: Level 1, N = 337; Level 2, N = 81. MSEM = multilevel equation modeling; SE = standard error; CI = confidence interval.
The interaction between superficial harmony and relationship closeness was significantly related to negative anticipation (B 4 = −0.17, S.E. = 0.05, p = 0.00). We calculated and plotted the slopes in Figure 2. Superficial harmony was significantly related to negative anticipation when relationship closeness was lower (B = 0.87, S.E. = 0.13, p = 0.000), significantly stronger than the simple slope when relationship closeness was higher (B = 0.39, S.E. = 0.10, p = 0.00). The difference in the effect between high and low levels of relationship closeness was 0.48, at the 95% CI for the difference in effect of [0.15, 0.80]. These results supported H4 (the first-stage moderation).
Figure 2. The moderating effect between superficial harmony and relationship closeness on negative anticipation
Regarding the moderated mediation effect, we found that the indirect effect of superficial harmony on lack of prohibitive voice via negative anticipation was significant when relationship closeness was low (indirect effect = 0.31, 95% CI [0.14, 0.50]) and also significant when relationship closeness was high (indirect effect = 0.14, 95% CI [0.05, 0.24]). Yet the difference in the effect between the two conditions was −0.16, at the 90% CI [−0.35, −0.002], suggesting that the indirect effect was weaker when closeness is high than when it is low. These results supported H5.
In addition, we had argued that it is superficial harmony, rather than genuine harmony, that leads to negative anticipation that results in conflict avoidance. Nevertheless, we found a significant bivariate correlation between genuine harmony and negative anticipation (r = 0.22, p = 0.000) at the individual-level. On the other hand, in the MSEM analysis, the relationship between genuine harmony and negative anticipation was not significant (B = −0.06, S.E. = 0.10, p = 0.56), when all the parameters were simultaneously estimated. A supplementary analysis also showed non-significant indirect relationship between genuine harmony and lack of prohibitive voice via negative anticipation (indirect effect of −0.02, S.E. = 0.04, p = 0.57). Putting together, these results reveal the connection between genuine harmony and superficial harmony on the one hand, and their distinction in predicting conflict avoidance on the other hand.
DISCUSSION
This research examines why employees choose to avoid confronting others even when they find problems in the workplace. We demonstrate that conflict avoidance is highly correlated with reluctance to voice and lack of prohibitive voice. The evidence for the convergence of these variables enables us to use reluctance to voice and lack of prohibitive voice as proxy variables for conflict avoidance. In Study 1, participants with working experience responded to a voice scenario, and in Study 2 executives rated their subordinates’ prohibitive voice behaviors in the workplace. The results of both studies indicate that employees choose an avoidance approach because they have concerns about their relationship with their supervisor. In other words, if they anticipate that using a direct approach might hurt the relationship or hamper the development of the relationship with the supervisor, they will avoid, rather than engage in, possible conflict. This negative anticipation captures individuals’ concern over their social-relational outcomes. Both studies show that individual value of superficial harmony engenders negative anticipation, which in turn leads to conflict avoidance. We further find that the relationship between superficial harmony and negative anticipation, and the indirect relationship between superficial harmony and conflict avoidance, is negatively moderated by the closeness of the relationship between the two parties. Our research contributes to the literature on conflict management and provides implications for practitioners.
Theoretical Contributions and Practical Implications
This research advances the existing literature in several respects. First, this research sheds light on individuals’ relationship concerns in conflictual situations and extends the literature on conflict by highlighting the focus of people's concerns before they choose a conflict-handling style. While avoidance is traditionally regarded as an outcome of low concern for self and low concern for others (e.g., Rahim, Reference Rahim1983), scholars emphasize the tangible outcomes for the parties to a conflict (e.g., Pruitt & Carnevale, Reference Pruitt and Carnevale1993) and conclude that conflict avoidance is the least effective strategy. However, in some cultures and situations, individuals are more concerned about their possible relationship costs (e.g., Ohbuchi & Takahashi, Reference Ohbuchi and Takahashi1994) than their economic interests; these social-relational outcomes are not adequately acknowledged in the dual-concern model. As conflicts often occur among parties who may have prior interacting experiences, individuals often consider the continuation of a positive relationship or interpersonal harmony with others when deciding how to address a conflict.
Second, this research highlights the role of superficial harmony in understanding conflict management in general and conflict avoidance in particular. From Leung's (Reference Leung, Christopher and Erez1997) point of view, superficial harmony or disintegration avoidance motivates individuals to avoid actions that strain a relationship. Superficial harmony reflects a self-protective and defensive orientation derived from high concern over a relationship. It drives people to focus on the negative repercussions of interpersonal relationships that are dampened because of conflictual situations. This motive leads people to behave in a manner that they perceive will help maintain harmony, even if there are economic costs or losses of other tangible resources. As a result, even before conflicts occur, the anticipation of relationship costs motivates people to avoid direct confrontation. Our theorizing is consistent with the finding that a strong sense of interdependence and high communal motive for socially desirable responding make people in East Asia unlikely to deal with conflict directly (e.g., Leung, Reference Leung, Christopher and Erez1997; Ohbuchi & Atsumi, Reference Ohbuchi and Atsumi2010; Tjosvold & Sun, Reference Tjosvold and Sun2002). This notion helps explain why superficial harmony engenders negative anticipation and further leads to conflict avoidance.
Finally, although our theorizing originates from Chinese culture, which values interpersonal harmony, we believe that our model holds whenever the importance of social-relational outcomes is highly embraced. In the workplace, because the supervisor has the power to make job assignments, decide resource allocations, give performance appraisals, and grant promotions, subordinates are always concerned about the relationship with their supervisor. This explains why conflict avoidance in the workplace has also been a focus of attention in cultures outside East Asia. At a leading high-technology corporation in the United States, about half the surveyed employees did not feel comfortable speaking up at work (Detert & Edmondson, Reference Detert and Edmondson2011). Jack Welch pointed out that individuals ‘keep their mouths shut in order to make people feel better or to avoid conflict. . . .That is true in every culture and in every country and in every social class’ (Welch & Welch, Reference Welch and Welch2005: 15–16). Research in Western countries demonstrates that perceived risk is a vital determinant of the avoidance of open communications (e.g., Detert & Burris, Reference Detert and Burris2007). Our theorizing accounts for these research findings.
Our research has implications for conflict management in the workplace. Though conflict avoidance is functional in maintaining positive interpersonal relationships, it is often dysfunctional or even harmful in problem solving (Tjosvold, Hui, & Ding, Reference Tjosvold, Hui and Ding2003; Zhang et al., Reference Zhang, Zhang, Wang, Leung, Chiu and Hong2011). It is sometimes necessary to engage in task conflicts in order to promote information exchange and improve the quality of organizational decisions. Our research points to the necessity of diminishing avoidance by cultivating a direct approach to conflict even among people with high relationship concern. Our findings suggest that building closer relationships can loosen the relationship between superficial harmony and negative anticipation, which in turn reduces the tendency toward avoidance. Hence, managers are advised to develop trust within the group or organization, to mitigate employees’ concerns about offending others, and these practices can be beneficial for encouraging employees to express their ideas, dissents, and true feelings. As long as employees are not preoccupied with relationship concerns in their work, they are less likely to avoid conflict. Moreover, because individuals who endorse superficial harmony view harmony maintenance as a way to protect themselves from interpersonal conflict (Leung, Reference Leung, Christopher and Erez1997; Leung et al., Reference Leung, Koch and Lu2002), managers are encouraged to use incentives to prioritize consideration of a subordinate's performance or effectiveness, rewarding information exchange and weakening excessive attachment to harmony. Lastly, because avoidance arises from people's concerns about relational losses, it is necessary to follow the classic advice to ‘separate the people from the problem’ in handling conflicts (Fisher & Ury, Reference Fisher and Ury1981).
Limitations and Future Directions
This research has limitations that should be addressed in the future. First, this research conceptualizes silence or reluctance to voice in the workplace as one form of conflict avoidance. Although our pilot study demonstrates that conflict avoidance and lack of prohibitive voice are highly correlated, providing empirical evidence for the convergence validity of the two constructs, more work is needed to establish construct validity. Scholars examine conflict avoidance and organizational silence in parallel, accumulating rich literature in each field. Both phenomena might be driven by people's concerns about risky actions, such as speaking up or engaging in conflict, suggesting the possibility of linking the literature in the two fields. As research on silence and voice has mostly been conducted in work settings, conflict management research could benefit from identifying some organizational contexts in these findings and further examine their effects on conflict-handling practices.
Second, we propose theorizing on conflict avoidance in a way that differs from the dual-concern model, in that conflict avoidance results from individuals’ concern about the social-relational outcomes, rather than individuals’ low self-concern and low other-concern. Though we provided both theoretical arguments and empirical evidence for this relational perspective on conflict management, we did not include a direct comparison of our proposed model and the dual-concern model. Recognizing the insights from the dual-concern model, this research neither intends to disprove this model nor concludes that our theorizing is superior. Instead, we theorize that, in some cultures or under some situations, relationship maintenance or interpersonal harmony is an important concern, which influences individuals’ choice of conflict-handling strategy. In future studies, we should examine the conditions in which the dual-concern model or our theorizing better predicts conflict avoidance. To achieve this goal, we can measure individuals’ concerns over their own interests, others’ interests, and relationship losses, and compare each concern's role in determining the choice of conflict-handling strategy.
Third, as we believe that our model holds whenever social-relational outcomes are prioritized, future research can be extended to other cultures that are either similar to or different from Chinese culture, which can improve the generalizability of our theory. In addition, as situations or contexts in the same culture varies in making people's relational concerns more or less salient, future studies can be designed to explore how other individual and situational factors influence negative anticipation and avoidance behaviors.
Lastly, it is worth investigating the outcomes of avoidance, for instance, the effects of avoidance in task and relational aspects. One study shows that people who avoid conflict are less cooperative with their counterparts in future interactions, suggesting a ‘carryover’ effect of unresolved conflicts due to avoidance (Zhang et al., Reference Zhang, Zhang, Wang, Leung, Chiu and Hong2011). Research in organizational settings also indicates that avoiding conflict is ineffective, whereas engaging in conflict contributes to strong relationships, and in turn strengthens team effectiveness and employee citizenship, even in Chinese culture, where interpersonal harmony is paramount (Tjosvold et al., Reference Tjosvold, Hui and Ding2003). Trust building and relationship maintenance are important for parties to effectively handle conflict. Is avoidance helpful or harmful for parties in building trust? This depends on how one party perceives and interprets the other's avoidant behavior. As long as the other party interprets the avoidance as a signal of good will that benefits his/her interests, the trust or relationship between them will be maintained or even improved. Future research could examine the contexts and mechanisms that lead conflict avoidance to hurt or to help problem solving and relationship building.
CONCLUSION
As conflicts often occur between parties who have prior interacting experiences or existing social bonds, people may take their interpersonal relationship into account when dealing with conflict and believe that engaging in conflict may hurt the relationship or damage interpersonal harmony between the parties involved. Results from two studies in work settings show that individual value of superficial harmony led to negative anticipation of conflict involving direct confrontation, which in turn results in avoidance. However, when the interacting parties have a close relationship, the direct relationship between superficial harmony and negative anticipation and the indirect relationship between superficial harmony and avoidance were weaker. Our research extends the existing conflict management literature by highlighting the role of superficial harmony in understanding why people avoid conflict and offers suggestions on how best to handle conflict in the workplace.
APPENDIX I
Pilot Study
The purpose of the pilot study is to examine the relationship among conflict avoidance, reluctance to voice, and lack of prohibitive voice in the workplace and to validate the match between the construct (conflict avoidance) and the measures (reluctance to voice, and lack of prohibitive voice) in our main studies.
Method
Sample and procedures
We used the online survey system ‘sojump’ to collect data. For our research purposes, we invited participants on the basis of eligibility criteria, including at least three years of work experience and being over 25 years old. We received responses from 210 participants. Among them, three respondents consistently selected the middle of the scale, showing that they tended not to take a position one way or the other. We thus omitted these three responses, leaving valid data from 207 respondents (63% females), with average age of 32 (S.D. = 4.86) and average work tenure of 99 months (S.D. = 46.32).
The survey had two parts. In the first part, respondents were presented with a conflict scenario that described an implicit conflict between a supervisor and a subordinate, and they were asked to put themselves in the role of the subordinate.
Company F was involved in communication system terminal design. You are the director of the hardware technology division. Your direct supervisor is the chief technology officer (CTO). The interaction between you and your supervisor is mainly limited to the work setting; personal interaction is rare.
Projects at Company F are traditionally completed by project teams. According to company policy, divisional managers also serve as project managers, in charge of selecting project members and assigning work tasks. In a meeting today, the CTO notifies all the divisional managers that the company has decided to change the project team management policy. Specifically, the company will no longer require project teams to be managed by the respective divisional managers. Any employees who propose creative and feasible plans will be appointed as the project team manager to carry out their own project. More importantly, their performance as project managers will have a direct impact on their performance appraisal. The CTO insists that all divisional managers implement this policy and encourage their subordinates to take the initiative to become project team managers.
However, you know that most of the subordinates, particularly the junior ones, do not have sufficient work and management experience to lead a project independently. They might also encounter difficulties in handling human resource issues. It would be extremely challenging for them to be project managers, and having their performance evaluation based primarily on their performance as project managers would create an unnecessary burden for them. Thus, you know that the company must slow down the implementation of this new policy so that the employees can become more prepared.
After reading the scenario, respondents were asked to indicate the likelihood of their expressing their concerns at the meeting and the likelihood of their avoidance of conflict in this situation.
In the second part, respondents were asked to report their prohibitive voice behaviors and conflict-avoiding style in their actual workplace.
Measures
Reluctance to voice was measured using two questions, and respondents were asked to indicate how likely they were to exhibit the described behaviors, using a seven-point Likert scale (1 = least likely, 7 = very likely). The two questions were: ‘I will voice my opinion to the chief technology officer during the meeting’ (reverse coded), and ‘I will remain silent at the meeting’. The Cronbach α was 0.89.
Conflict avoidance was measured using five items adapted from the scale measuring avoidance style developed by Rahim (Reference Rahim1983). These items were modified to fit the current conflict scenario. Respondents were asked to indicate how likely they were to exhibit the described behaviors, on a seven-point Likert scale (1 = least likely, 7 = very likely). Three of the items were: ‘I attempt to avoid being “put on the spot” and try to keep my conflict with the chief technology officer to myself’, ‘I try to keep my disagreement with the chief technology offer to myself in order to avoid hard feelings’, and ‘I avoid open discussion of my differences with the chief technology officer’. The Cronbach α was 0.93.
Prohibitive voice was measured using the five items for speaking up about organizational problems by Liang et al. (Reference Liang, Farh and Farh2012). Examples of items were: ‘Speak up honestly about problems that might cause serious loss to the work-unit, even if that dissents from the supervisor's views’, and ‘Dare to voice opinions on things that might affect efficiency for the work-unit, even if that would embarrass the supervisor’ (1 = strongly disagree, 7 = strongly agree). The Cronbach α was 0.92.
Conflict-avoiding style was measured using Rahim's (Reference Rahim1983) five items. Yet these items were adapted to target respondents’ behaviors in their actual workplace. These items include: ‘Avoid being “put on the spot” and try to keep conflict with my supervisor to myself’, ‘Keep my disagreement with my supervisor to myself in order to avoid hard feelings’, and ‘Avoid open discussion of my differences with my supervisor’ (1 = strongly disagree, 7 = strongly agree). The Cronbach α was 0.93.
We controlled for respondents’ gender (1 = male, 0 = female), age, and work tenure.
Results and Discussion
The results showed that reluctance to voice dissent was positively correlated with conflict avoidance (r = 0.79, p = 0.000). Meanwhile, prohibitive voice is negatively related to avoiding style (r = −0.68, p = 0.000). As we discussed, reluctance to voice dissent and prohibitive voice reflect employees’ conflict avoidance, which results from their concern over their relationship with the supervisor. Moreover, we found a high correlation between prohibitive voice and the general avoiding style measured using Rahim's scale.
In other words, people withhold their dissenting opinions in order to avoid potential conflicts with another party. These results support our conceptualization that reluctance to voice or a lack of prohibitive voice is a form of conflict avoidance in organizations. Therefore, we use prohibitive voice situations as an organizational context to examine conflict avoidance, as we demonstrate in Studies 1 and 2.
