There are two primary conclusions from the research investigating whether team conflict has a positive or negative effect on team performance: (1) task conflict can be, but is not always, beneficial to team performance, but procedural and relationship conflict rarely are; and (2) correlations between task, relationship, and procedural conflict tend to be high, suggesting that teams rarely experience one type of conflict in isolation—rather different types of conflict co-occur (e.g., Dreu and Weingart 2003; Wit et al. 2012). Yet, despite widespread recognition that types of conflict covary, theory and research has only recently begun to look at how conflicts interact with one another to impact team performance (e.g., Greer et al. 2008; Jehn and Chatman 2000; Simons and Peterson 2000). Moreover, very little attention has been paid to how the interaction between conflict types impacts team conflict management processes as a dependent variable, rather than team performance.

Our research addresses this gap in the intra-group conflict literature. We develop a theoretical explanation for how and why teams manage conflict more or less successfully when different types of conflict co-occur. Our research asked: can theorized benefits of task conflict to team performance also apply to developing effective strategies for managing relationship and procedural conflict? Given that relationship and procedural conflict have uniformly poor implications for team performance, does their co-occurrence with task conflict keep a team from benefitting from the potential positive implications of task conflict? We propose that when task conflict co-occurs with procedural or relationship conflict, task conflict will act as a catalyst facilitating the effective management of procedural and relationship conflict. In contrast, we propose that when procedural or relationship conflict co-occurs with task conflict, procedural or relationship conflict will suppress the likelihood that task conflict is managed effectively. By “effective management” we mean that teams choose strategies for managing conflicts that are consistent with theoretically and empiricially prescribed “matches,” as found in the literature (which will be detailed below), between type of conflict and conflict management strategy. We contend that co-occurring task conflict is a catalyst of effective management of procedural or relationship conflict, because it motivates the team to confront conflict in a constructive fashion and provides a model for doing so. We contend that co-occurring procedural or relationship conflict has a suppressing effect on the effective management of task conflict, because it distracts the team from the task, and does not provide an effective model for managing task conflict.

We focus on conflict management as our dependent variable rather than team performance for several reasons. First, it is a more proximate outcome of conflict than performance. Second it is an important moderator of the conflict to performance relationship (Behfar et al. 2008; Dreu 1997; DeChurch and Marks 2001; DeChurch et al. 2007; Greer et al. 2008; Tekleab et al. 2009). Third, a deeper understanding of the interactive effects of different types of conflict on conflict management extends knowledge by identifying the conditions under which different types of conflict are and are not managed appropriately. Our research is inductive. We start from the premise that whether or not a team uses the theoretically prescribed “appropriate process” to manage one type of conflict depends on what other type of conflict is co-occurring. We anticipated three possible patterns: (1) the co-occurring conflict could act as a catalyst, making it more likely another type of conflict would be managed appropriately; (2) the co-occurring conflict could act as a suppressor, making it less likely that another type of conflict would be managed appropriately; or (3) the co-occurring conflict could have no effect on the association between conflict and appropriate conflict management. Our theory sections below develop the reasoning underlying our premise. We explain why we anticipated, that if any type of conflict acted as a catalyst, it would be task conflict and if any type of conflict acted as a suppressor it would be procedural or relationship conflict. We also explain our reasoning, based on prior research findings (see Sect. 3), regarding appropriate fit between type of conflict and conflict management strategy.

1 Overview of the Study

Our inductive approach began with semi-structured interviews with executives from 44 different organizations who had recently been a member of a negotiating team. A negotiating team is a group of threeFootnote 1 or more people on one side of the table who are responsible for planning and executing an across-the-table bargaining strategy. We chose to study co-occurring conflicts in the context of negotiating teams because the complex nature of their task and the deliberate selection of members with divergent expertise, underlying interests, and/or political influence within the organization mean these teams were likely to face a number of different and simultaneous intra-group conflicts. Our sampling identified male and female experts from a variety of industries, whose teams had negotiated in deal making or dispute resolution contexts, and whose teams were in the roles of buyer, seller, plaintiff or respondent. Limiting our study to a sampling of negotiating teams allows us to compare and contrast teams engaged in a generally similar task across a vast array of organizational contexts.

We recorded and transcribed the interviews and then identified passages containing accounts of conflicts and conflict management strategies. These accounts were the units of analysis for concept mapping text analysis (Jackson and Trochim 2002). The concept mapping analysis generated 12 clusters of conflicts and 15 clusters of conflict management strategies. We then used pattern coding (Miles and Huberman 1994) to merge the conflict clusters and, separately, the conflict management clusters into meta-categories grounded in the team conflict literature. This created four meta-categories of conflicts and four meta-categories of conflict management strategies. Next academic experts rated the extent to which each conflict meta-category was similar to task, procedural, and relationship conflict.

At this point we turned to the team conflict, negotiation, and organizational behavior literatures to identify a theoretical rationale to match each type of conflict with an appropriate conflict management strategy. We then assessed the likelihood our premise: whether or not a team uses the theoretically prescribed “appropriate process” to manage one type of conflict will depend on what other type of conflict is co-occurring, using Chi square analysis. Our analyses assessed the probability of the theoretically appropriate management strategy being used when its corresponding type of conflict occurs and a co-occurring conflict occurs. We also assessed the interactions of co-occurring conflict types on team members’ own perceptions of their team’s effectiveness. We consider this analysis to be a method check—or a way to provide convergent evidence—that occurrences of “appropriate” matches of conflicts and management strategies were also times when our study participants felt that the team was operating well. We label this measure “perceived effectiveness.”

2 Methods

2.1 The Sample: Interview Participants

We used purposive sampling (Kemper et al. 2003; Strauss and Corbin 1998) to recruit managers who could talk about their recent experience as a member of a negotiating team (Seidler 1974). All managers (from here forward referred to as participants) had prior executive education training in negotiation. We asked each of them to describe their own experiences and their own views of how the team responded rather than to represent the opinions and feelings of all members of their team (Seidler 1974). Thus, our participants provided structural, rather than strictly social-psychological, information about team functioning (Seidler 1974). To recruit participants, the authors emailed mid or high-level managers who had been in their negotiation executive education classes at two American universities. This identified people who had recent experience as a negotiating team member and were willing to be interviewed. As the email lists were not under the direct control of the authors, we estimate that approximately 500 emails were sent. The final sample size was 44, because we were unable to schedule a telephone interview with 13 of the 57 total enrollees during the five month data collection.

2.2 Interview Protocol and Questions

We used a standard set of interview questions to maintain consistency in the type of information elicited (Johnson and Turner 2003; Seidler 1974). After receiving consent to record the interview, we asked participants to describe a single negotiating team experience. We collected information about the team (e.g., size, members’ functional departments; industry; the team’s role in the negotiation (buyer, claimant, etc.); and the duration of the negotiation. The 44 participants were senior managers from a variety of industries,Footnote 2 the negotiating teams described ranged in size from 3 to 15 people, with an average team size of 5 members (\({\hbox {SD}}=2.5\)). In terms of role in the negotiation: 55 %, or 24 of the participants represented a seller role, 31 % (14 participants), a buyer role, and the remaining 14 % (6 participants) in a claimant, defendant, or employer role. On average, participants were involved with the negotiating team for 6 months (\({\hbox {SD}}=7.8\)), the median duration was 3 months, the minimum was one month, and the maximum duration was 3 years. Thirty-seven (84 %) of the participants were male; 7 (16 %) were female. Our sampling succeeded in identifying negotiating team members from a variety of industries, who played different roles in the negotiation, so that our study was not restricted by the negotiation context or demographics of the participants.

We asked: What was your most challenging experience when negotiating with this team? The follow-up was: What did the team do to manage this challenge? We purposely asked participants to describe the “challenges” (rather than “conflicts”) they experienced on the negotiating team in order to capture a broad array of oppositional experiences and to avoid any priming that might be associated in their minds with describing the experience as a conflict. Previous research has demonstrated that participants responding to interview questions react and interpret the word “conflict” with a wide range of views (Behfar et al. 2011; Bendersky et al. 2014; Schwab 1980). We therefore used the word “challenge” to best capture all situations that participants viewed as oppositional not just severe conflicts. The interview continued until the participant indicated that all of the team’s challenges and the management strategies used to address those challenges had been described.

In order to assess the participants’ views of their team’s effectiveness (“perceived effectiveness”), at the end of the interview we asked them to respond on a Likert-rated scale (\( 1 = {\hbox {strongly disagree;}} 5 = {\hbox {strongly agree}}\)) to the following questions (scale \(\alpha \,=\,.74\)): “My negotiating team was effective at developing a cohesive strategy; During negotiations, my team worked together as a true team; Our team made effective use of each team member’s skills and knowledge; I would work with these teammates again if given the opportunity.”

2.3 Phase I: Coding the Interviews

2.3.1 Concept Mapping Coding

We used concept mapping to code the interview data. Concept mapping is a text analysis technique that is a hybrid between traditional content analysis and semantic mapping techniques (see Jackson and Trochim 2002). The coding consists of five steps that combine card sorting and statistical analysis (multidimensional scaling and cluster analysis) to produce clusters of units. This method is appropriate because it classifies the data according the participants’ experiences and perspectives. It allowed us to identify the patterns of association between co-occurring conflicts and conflict management as participants experienced them (rather than how an a priori coding scheme might classify them). Later in the analysis we will return to the task, relationship, and procedural conflict framework. As a first analysis step, however, we were interested in retaining participants’ perspectives.

The study’s raw data consisted of 204 units of conflicts and 134 accompanying units of management strategies. Each unit was a passage (typically 2–3 sentences) from an interview transcript. Two coders read transcripts and independently identified passages describing conflicts and conflict management strategies. 77 % of the conflict units and 74 % of the management strategy units were initially identified by both coders. When a unit was identified by just one coder, it was discussed before being included in the data set. The set of conflict units and the set of conflict management units were given to 20 MBA student coders. Their task was to sort the conflict units and then separately the conflict management units (or in reverse order) into similarity categories of their own choice. This similarity data was entered into a multidimensional scaling analysis and then a cluster analysis. Two of the coders then used the output of the cluster analyses (a dendogram) to choose the appropriate number of clusters and give the clusters labels. This process of concept mapping was done separately for conflicts and management strategies, and produced two separate visual maps: one of 12 conflict clusters (see Fig. 1); and the other of 15 conflict management clusters (see Fig. 2). Table 6a–6c in “Appendix” presents representative statements illustrating each conflict cluster; Table 7a–7d in “Appendix” presents representative statements illustrating each management strategy cluster. The relevant information to interpret the maps in Figs. 1 and 2, is the distance, or spatial relationship between clusters, derived from the multidimensional scaling analysis. Conflicts or management strategies formed a cluster because coders judged them to be similar. The shape and size of a cluster indicates whether it is a broad or narrow conceptual area, but does not allow for meaningful interpretation (e.g., the size of a cluster does not represent the number of statements in a cluster).

Fig. 1
figure 1

Concept map of conflicts overlaid with conflict meta-categories

Full size image
Fig. 2
figure 2

Concept map of management strategies overlaid with meta-categories

Full size image

2.3.2 Meta-Category Coding: Grounding Participant Perspective in Theory

The concept maps represent the variety of conflicts and management strategies that our participants described. However, as researchers grounded in the existing teams and negotiations literatures, we could see conceptual similarities between clusters identified by the concept mapping analysis. For example, on the right side of Fig. 1, there are several clusters about differences of opinions, goals, and interests. Such conflicts are labeled as “conflicts of interest” in negotiation theory and “task conflict” in the intragroup conflict literature. Therefore, in order to create theoretically meaningful meta-categories out of the clusters, we did two things. First, we used the pattern-coding methodology to merge regions of “loosely held chunks of meaning” (Miles and Huberman 1994: 70) into overarching themes, or meta-categories grounded in theory. Since the proximity and distance between clusters provide a statistical basis to identify regions of empirical similarity, we could not draw boundaries that conflicted with the coders’ perspectives of how the statements were organized (Kane and Trochim 2007). Second, the authors agreed on the number of meta-categories for each map, and then worked together to name them. The definitions of the meta-categories and their association with the original clusters are presented in Tables 6a–6c and 7a–7d in “Appendix” and shown as dotted line shapes in Figs. 1 and 2.

Table 1 Expert ratings of how much each conflict meta-category is related to task, relationship, and procedural conflict
Full size table

2.3.3 Academic Expert Classification of Conflict Meta-Categories

Our next step was to determine the convergent and discriminant validity between our conflict meta-categories (reflecting participant experiences) and the widely used three dimensional task, procedural, relationship taxonomy of team conflict (Jehn 1995, 1997). We asked a group of 20 academic experts in the organizational behavior field to rate the extent (\(1 = {\hbox {not at all related;}} 5={\hbox {totally related}}\)) to which each meta-category was related to task, relationship, and procedural conflict (see Hinkin and Tracey 1999 for a description of this process). Results of one-way ANOVA analysis with Bonferroni contrasts, reported in Table 1, indicate that our task conflict meta-category was viewed by our academic experts as more closely related to task conflict than to procedural or relationship conflict, and that the remaining three meta-categories (disruptive contributions, logistics, and roles) were viewed as more similar to procedural conflict than to task or relationship conflict.

It is worth noting that relationship conflict, although a prevalent construct in the teams literature, did not emerge as a distinct meta-category nor was it distinctly associated by the academic expert raters with one versus another of the conflict meta categories. The means in Table 1 indicate that the academic experts rated all of the conflict meta-categories around the mid-point of the relationship conflict scale—indicating that interpersonal tension was not unique to any one conflict meta-category and that some degree of interpersonal tension might exist in all of the meta-categories (e.g., see Cronin et al. 2011; Jehn 1995; Korsgaard et al. 2008 for a similar conclusion). The highest mean rating for relationship conflict was in the meta-category of disruptive contributions. However, the expert ratings of this category, as well as a close examination of the accounts within the disruptive contributions category (see Table 6a–6c in “Appendix” for examples), indicate that the behaviors in this category, such as surprising and unauthorized breaks with team strategy, members who were impatient and behaved impulsively, and differences in negotiating styles were challenging because they disrupted the flow/process of work—not because they represented interpersonal tension. The disruptive behaviors were the source of the conflict, not general interpersonal tension. This meta-category is consistent with the “contribution conflict” dimension of procedural conflict, while the roles and logistics meta-categories reflect the “logistical conflict” dimension of procedural conflict (Behfar et al. 2011).

2.3.4 Phase I Coding: Discussion

The results of Phase I of our study supported our having used concept mapping to code participants’ experiences rather than our specifying the three types of conflict as an a priori coding scheme. Had we used the three conflict types in a pre-defined coding scheme (as would be done in more traditional content analysis), we would have misrepresented participants’ actual experiences by coding some as relationship conflict (for example, the labels of the clusters in the disruptive contributions meta-category “look like” relationship conflict). The actual descriptions of conflicts in our qualitative data clearly illustrate, however, and our expert ratings confirm, that the negotiating teams in our study were not predominantly challenged by interpersonal animosity (although it existed), but by the more process-based conflicts of coordinating busy people with different expertise, styles, availability, and with differing substantive ideas and interests about the task. Finally, we found, consistent with the call for studying multiple conflicts simultaneously (Arrow et al. 2000; DeChurch et al. 2013; Greer et al. 2008; Mooney et al. 2007) that conflicts seldom occurred in isolation. Of our 44 participants 86 % (or 38) reported their teams experienced conflicts that ended up in two or more of our meta-categories.

2.4 Phase II: Evaluating the Implications of Co-occurring Conflicts for Conflict Management

The purpose of Phase II was to evaluate the implications for conflict management when more than one type of conflict co-occurred. This evaluation had two steps. First, using the intra-group conflict, negotiations, and organizational behavior literatures, we proposed an “appropriate” theoretical match between each meta-category of conflict and each management strategy meta-category. Second, we tested whether the theoretical match was more or less likely when there was another co-occurring type of conflict.

2.4.1 “Appropriate” Conflict Management for Each Type of Conflict

Our Phase 1 coding identified four meta-categories of conflict management: negotiating, structuring communications, managing logistics, and organizing and four meta-categories of conflict: task conflict, disruptive contributions, logistics, and roles. We used theory from the negotiations, teams, and organizational behavior literatures to identify theoretical prescriptions for the most “appropriate” way to manage each of the four conflicts identified in Phase I. These prescriptions are presented in Table 2 and elaborated below.

Table 2 Summary of supporting literature for matches between conflicts and management strategies
Full size table

Managing Task Conflict Task conflicts are substantive. They occur when individuals share an understanding of the problem, but have different interests or ideas about how to reach a solution (Amason 1996; Druckman and Zechmeister 1973). The task conflict meta-category combined three concept mapping clusters: teammates are not on the same page, confusion about goals, and conflicting interests within the team. These clusters described challenges in developing a negotiating strategy due to different goals among team members and/or between the team and a members’ home function (e.g., maintain the customer relationship vs. increase prices). There were 26 (59 % of the participants) accounts in this meta-category.

Managing task conflict within a negotiating team implies that the team is going to have to do several different things including, addressing members’ different perspectives and styles and integrating them into a coherent strategy for across-the-table bargaining. Although majority rule, or another decision generating technique, might be possible in some cases, it usually leaves substantive differences unresolved (Brett 1991) which makes generating a coherent across-the-table bargaining strategy difficult. Thus, we propose that a conflict management process where members engage in internal team negotiating is the most appropriate management strategy for task conflict. Our negotiating meta-category contained six concept mapping clusters: managing different negotiation styles, building coalitions to reach agreements, managing conflicting interests, reaching a compromise, team problem solving, and building flexibility into planning illustrating that managing task conflict is a multifaceted task. The strategies participants mentioned that ended up in this meta-category of conflict management included several mentioned frequently in the conflict management literature, for example, the use of superordinate goals to align team members (e.g., Ancona 1990; Hirokawa 1983; Moreland and Levine 1992), bringing in outside experts to provide objective opinions and provide data, using external social pressure or authorities to help team members to see things from a different perspective, engaging in stakeholder relations, using persuasion to ‘win-over’ team members’ constituents, and negotiating within the team by making concessions and trading off multiple issues (Brett 2007; Fisher et al. 1991; French and Raven 1954; Friedman 1994; Janis 1972). This meta-category of conflict management strategy directly addresses managing Task conflict.

Managing Disruptive Contributions Disruptive contributions are procedural conflicts. This category contained clusters of behaviors such as different negotiation styles, aggressive personalities, conflicting personalities, and team conflicts in front of clients. Teams facing disruptive contributions were dealing with individuals who had strong opinions about negotiation strategy, whose unwillingness to acquiesce or compromise disrupted or significantly stalled team progress, and instances of team members choosing to ignore team strategy (e.g., unexpectedly accepting an offer when not authorized to do so). These behaviors broke team discipline in front of the other party and disrupted the enactment of the team’s strategy. The negative influence of disruptive contributions on performance has been well documented in the teams’ literature (Behfar et al. 2011; Bendersky and Hays 2012; Greer et al. 2008; Greer and Jehn 2007; Jehn 1995). There were 26 (59 % of the participants) accounts in this meta-category.

Managing disruptive contributions implies that the team needs to address members’ actions that are interfering with how the team was developing and implementing its across-the-table negotiating strategy. Theoretical discussions of how to manage procedural conflict and get disruptive members back in line with team strategy refer to team members trying to directly identify the problem, communicate the problem, and work with teammates to resolve the problem (Brett 2007; Brett et al. 2014; Greer et al. 2008). This requires communication and planning within the team so that when the team is at the negotiating table, it speaks with one coordinated voice. The literature suggests strategies to work through impulsive responses (Shell 2006) such as reflective questioning (Picard 2002), clarifying expectations and boundaries (Mathieu and Rapp 2009), learning how to prevent previous team mistakes (Behfar et al. 2008), and naming emotional reactions to different team positions (Bartunek et al. 1992). Techniques in the two concept mapping clusters in our management strategies meta-category structuring communications reflected these methods: adequately preparing with teammates and managing internal communication. These clusters contained strategies such as teams rehearsing, role playing, and establishing signaling systems to communicate within the team while negotiating across-the-table. These were all strategies that created commonly shared understandings and procedures for how to communicate so that disruptions would not interfere with the team’s across-the-table strategy.

Managing Logistics Logistics conflicts implied that the team had to address members’ availability to prepare for and execute their negotiation strategy. Logistics conflicts are a form of procedural conflict. This meta-category combined three concept mapping clusters: communication inefficiencies and delays, scheduling logistics, and internal member availability. These clusters described the internal team conflicts around getting the right information and the right people to preparation and negotiation sessions. The teams literature describes procedural conflicts associated with temporal synchronization, and aligning team members’ busy schedules so that the team’s decisions meet organizational deadlines (Ancona and Chong 1999; Blount and Janicik 2002; Janicik and Bartel 2003). There were 29 (66 % of the participants) accounts in this meta-category.

Logistics conflicts require management strategies that address pressures of time and distance: team members have other full time responsibilities and are often not co-located (Hinds and Bailey 2003; Janicik and Bartel 2003). Techniques in the managing logistics meta-category did just this. Team members reported using technology to respond to evolving operating conditions (DeSanctis and Poole 1994; Earley and Gibson 2002; Finholt and Sproull 1990), which respected and adjusted for members’ time-related constraints (Ancona and Chong 1999; Blount and Janicik 2002). This meta-category contained three concept mapping clusters: efficiency managing members’ time, coordinating strategies during negotiation by stepping away from the table, and managing technology logistics. The teams’ strategies included using technology such as video conferencing to ease the time pressures of travel and time zone burdens, as well as using technology such as chat rooms to get the right information to the right people at the table in real time. They also included anticipating delays such as approval processes or the deadlines or work cycles of different members’ home departments. As such, these strategies directly addressed logistics conflicts related to timing.

Manging Roles This set of procedural conflicts stemmed from members’ multiple reporting relationships, status or power differences within the team, and functional role constraints that either limited or demotivated some members’ participation in the team. These challenges caused internal team coordination and information sharing problems. The roles meta-category combined two concept mapping clusters: unclear decision rights and conflicting team member roles. There were 25 (57 % of the participants) accounts in this meta-category.

According to the literature, managing the procedural conflicts in the roles meta-category implies that members need to address how to allocate authority and create a clear delineation of responsibilities (e.g., Arrow et al. 2000; Friedman 1994; Hackman 1987; Hackman and Morris 1975). Following this literature we proposed that roles conflicts would be best addressed by strategies in the organizing meta-category, which included four concept mapping clusters: assigning assertive leadership, gathering adequate resoures, clarifing roles for all team members including the leader, and defining decision rights. These strategies addressed roles conflicts by developing a more organized team structure (Greer and Kleef 2010), including assigning roles and responsibilities, clarifying authority, and generating decision rules (Cartwright and Zander 1968; Hackman 1987, 1990).

In sum the theoretical matches we propose are task and negotiating, disruptive contributions and structuring communications, logistics and logistics management and roles and organizing. We propose that these are theoretically appropriate strategic responses to the challenges of task and procedural conflict that we identified within negotiating teams. We use these theoretical matches as benchmarks against which to assess how appropriately the teams managed one type of conflict (e.g. task or procedural) when another type of conflict procedural or task co-occurred. We evaluated which of the three patterns proposed at the outset in our research question took place: (1) The co-occurring conflict could act as a catalyst, making it more likely another type of conflict would be managed effectively; (2) The co-occurring conflict could act as a suppressor, making it less likely that another type of conflict would be managed effectively; or (3) The co-occurring conflict could have no effect on the association between conflict and effective conflict management. Based on theorizing and empirical research in the intra-group conflict literature, our analysis is focused on assessing the probability that when task conflict co-occurred with procedural conflict, the task conflict would have a catalyzing influence and the procedural conflict would be appropriately managed, but that when procedural conflict co-occurred with task conflict, the procedural conflict would have a suppressing influence and the task conflict would not be appropriately managed.

2.5 Phase II Analysis: Assessing the Probability of Appropriate Conflict Management When There is a Co-occurring Conflict

The Phase II analysis involved three inductive data analysis steps. First, we tested for the possible patterns of moderation in our data by treating Task conflict as the moderator of the relationship between each of the three procedural conflicts and its theoretically matched management strategy; and then by treating each of the three procedural conflicts as the moderator of the relationship between task conflict and negotiating. In our data each team was designated as either having experienced the focal conflict or not, having experienced the moderating conflict or not, and having experienced the theoretically matched conflict management category or not. To assess whether these patterns were more likely to occur than chance we generated three way tables: the independent variable (the focal type of conflict); the moderator (the co-occurring type of conflict); and the dependent variable (the theoretical conflict management match for the focal type of conflict variable). We used a Chi square analysis. Due to our relatively small sample size we report both Chi square and Fischer’s exact test results to indicate significant effects (McDonald 2009). Our second analysis step, when there were significant moderator effects from the Chi Square analysis, was to do a close examination of case examples from the qualitative accounts to provide representative examples of the moderation effect, including confirming and disconfirming accounts. Third, we assessed the implications of the significant interactions on our participants’ evaluations of their team’s effectiveness.

2.5.1 Task Conflict as a Moderator of Appropriately Managing Procedural Conflict

Task conflict represents fundamental differences in team members’ goals; they are substantive and task-oriented. Theorizing about this type of conflict has highlighted many benefits to being forced to vet different ideas, such as exposure to different ways of thinking, consideration of alternative viewpoints, recognition of the costs and benefits of certain courses of action, and making connections between different conceptual arguments (Amason 1996; Amason and Sapienza 1997; Guetzkow and Gyr 1954; Jehn 1995, 1997). Each of these findings suggests that members of teams facing task conflict are likely to have to see issues from the perspective of other team members, and to recognize that others have legitimate differences of perspective. While alternative views on process are not inherently legitimate, alternative views on core interests are inherently legitimate. For example, if a negotiator on a team representing operations talks about the negative implications for system-wide logistics efficiency of making a promise to deliver a product extremely quickly, it is hard for a negotiator on the same team representing marketing to deny that system-wide logistics efficiency is a real issue for the company they both represent, even though logistics is not a core marketing concern. By contrast, alternative views on negotiating process choices are not inherently legitimate; if a teammate insists on making offers very quickly (or slowly) there is no inherent legitimacy to those positions. Given that task conflicts force team members to treat other members’ views as legitimate, it should result in a more open consideration and awareness of others’ views on the team. That open state of mind should be helpful when teams deal with process conflicts. Additionally, the intensity of the need to resolve task conflicts (it is very hard to proceed at all if substantive differences are not resolved) may force teams to persist in their search for a resolution of task conflicts, which then provides a model of behavior that carries over to other types of conflicts. Because experience with task conflict forces persistence in conflict resolution, and forces team members to see other views as legitimate, we expect that teams that experience task conflict may thereby be better at dealing with other kinds of conflicts.

Restated with the language of our meta-categories from our Phase I coding, this would mean:

When task conflict co-occurs (as a moderator):

  • Logistics will be more likely to be managed with Managing Logistics.

  • Disruptive Contributions will be more likely to be managed with Structuring Communication.

  • Roles will be more likely to be managed with organizing.

2.5.2 Results: Task Conflict as a Moderator

Table 3 summarizes the significant results testing the moderating influence of task conflict on the appropriate management of procedural conflict.

Table 3 Summary of significant results for task conflict as enhancing the management co-occurring procedural conflict
Full size table

Managing Logistics Conflicts Support was found for a catalyzing influence of task conflict for managing logistics conflict appropriately. Teams experiencing logistics conflicts were more likely to use the managing logistics strategy when the team was also experiencing task conflict than when it was not. The match between logistics and the appropriate managing logistics response was present when teams also faced task conflict \([\chi ^{2} (1, \hbox {N}= 36) = 6.02, {p} < .05\), Fisher’s exact test \(p=.03]\), but was not present when teams did not face task conflict \([\chi ^{2} (1, \hbox {N}= 8) = 1.6, {p} > .05\), Fisher’s exact test \(p=.46]\).

Managing Logistics Conflicts: Case Examples Teams with logistics conflict often struggled with arranging meetings around members’ heavy travel schedules and bosses being unwilling to release members to participate fully in team planning sessions. Common responses to this challenge when task conflict co-occurred were for the team to proactively request more resources such as: asking for time-release for members from their daily jobs, building in time for approval delays, and re-ordering team priorities. Teams with task conflict co-occurring with logistics were also better at using technology to manage logistics than teams without co-occurring task conflict. For example, some teams with task conflict opened chat rooms during live negotiations so that team members not at the negotiating table could participate and give input as the negotiation developed. This allowed team members to save valuable travel time, but also to relay real time information between the lead negotiator and their bosses.

Teams that did not experience task conflict, but nevertheless experienced logistics conflicts, managed their logistics conflicts less effectively. For example, a team in this category had to ask the counterpart across-the-table to delay the negotiations or to extend deadlines because of unanticipated or last minute member absences or substitutions. Other teams reported having to “scramble” to revise previous positions based on new information or developments that came late because of logistical problems. Such teams reported they could not try, or did not have ample time to use, technology to ease temporal strains. In the absence of task conflict, teams experiencing logistics conflicts most frequently turned to the strategy of structuring communication. That is, rather than creating sound managing logistics strategies, teams instead rehearsed how to manage through the fact that they were unprepared in front of the team across-the-table. So although teams may have managed to overcome (i.e., hide) their logistics conflicts, they were not optimally prepared to resolve their logistics conflicts.

Managing Disruptive Contributions Conflicts Support was also found for the catalyzing influence of task conflict for appropriately managing disruptive contributions. Teams experiencing disruptive contributions were more likely to use structuring communications when the team was also experiencing task conflict than when it was not. When teams experienced task conflict, they were more likely to manage disruptive contributions with structuring communication \([\chi ^{2} (1, \hbox {N}= 36) = 3.90, p = .05\), Fisher’s exact test \(p=.04]\) than when task conflict were not present \([\chi ^{2} (1, \hbox {N}= 8) = 1.6, p > .05\), Fisher’s exact test \( p= 1]\).

Managing Disruptive Contributions Conflicts: Case Examples Having to address task conflict seemed to make teams more aware of potential vulnerabilities to team discipline. For example, in one team (that also experienced task conflict) when a sales person was not in agreement with a price position the team adopted, the team rehearsed how that person should react if directly questioned by the counterpart. Other teams developed cueing systems and role played multiple scenarios within the team to head off disruptive contributions. In this way, task conflict seemed to raise awareness and create a proactive response that led to effective management of potentially disruptive behavior.

Members of teams without task conflict reported being surprised by, and forced to react directly to, individual members’ disruptive contributions. For example, when members got emotionally upset or behaved in ways that were not consistent with team strategy, these teams became vulnerable to “divide and conquer” tactics from across-the-table. In response to disruptive contributions, these teams most frequently used organizing. This typically meant, stopping the negotiation with a caucus or recess, and then clarifying who could speak at the table and who had authority to decide a particular issue. Participants reported that while caucusing and using authority helped avoid complete team breakdown, it also signaled a weakness in team strategy that the counterpart could exploit—something the teams with task conflict generally were able to avoid by having addressed or having become aware of potential disruptive contributions before the negotiations actually began.

Managing Roles Conflicts The Chi square analysis did not support the catalyzing influence of task conflict for appropriately managing roles with organizing. Interestingly, use of the organizing management strategy had no statistically unique pattern: it was applied with equal probability to all the types of conflict. Thus, it was no more likely to occur in response to roles than other conflicts.Footnote 3 Roles conflicts were prevalent: participants described teams of peers who had never worked together before, shifting team memberships, and leaders who did not lead. The prevalent use of organizing, even when it was not theoretically predicted to be the appropriate response, is a point we will return to in the discussion.

Influence on the Team’s Perceived Effectiveness Our analysis and qualitative accounts suggest that team members more appropriately managed procedural conflicts when they co-occurred with task conflict. We next sought to determine whether or not there was “convergent evidence” (Reeve and Smith 2001) for this pattern by testing if team members also perceived (rated) their teams as more effective when task conflict co-occurred with procedural conflict than when it did not. To do so, we analyzed the interaction of task conflict and each of the other three co-occurring procedural conflicts on participants’ perceptions of the effectiveness of the team. We used the scale created by the four Likert-rated questions listed in the methods section as the dependent variable, the dichotomous present/absence of co-occurring conflict as the independent variables and ANOVA to test the interaction. These results are depicted graphically in Fig. 3a, b.

Fig. 3
figure 3

Task conflict as a moderator of the relationship between co-occurring procedural conflicts (a: logistics, b: disruptive contributions) and team members’ evaluations of team effectiveness

Full size image

Task Conflict Co-occurring with Logistics The interaction of task conflict and logistics on perceived team effectiveness was significant [\(F (1, 3) = 4.46, p=.04\)]. When there was task conflict, there were no differences in the participants’ ratings of the quality of team process due to having (\(M=4.06, SD=.60\)) or not having (\(M=4.29, SD=.65\)) Logistics conflicts. By contrast, when there was no task conflict, having logistics conflict resulted in lower perceived team effectiveness ratings (\(M=3.33, SD=.75\)) than not having logistics conflict (\(M=4.75, SD=.35\)). Thus, although it was best to have no conflict at all (as reflected by the highest perceived mean effectiveness rating), teams having logistics conflict and task conflict were judged to be as effective as teams having only task conflict. Teams having no task conflict but having logistics conflict were reported to be least effective. See Table 4 for a summary of these contrasts.

Table 4 Means and standard deviations for perceived team effectiveness ratings
Full size table

Task Conflict Co-occurring with Disruptive Contributions The interaction of task conflict and disruptive contributions on perceived team effectiveness was significant [\(F (1, 3) = 6.3, p = .02\)]. This interaction was due to low ratings of perceived team effectiveness when there was disruptive contributions but no task conflict (\(M=3.25, SD=.81\)). However, when co-occurring with task conflict, disruptive contributions did not inhibit perceived effectiveness ratings (\(M=4.19, SD=.63\)). When teams experienced neither conflict, perceived effectiveness was rated to be the highest (\(M=4.42, SD=.63\)). When teams did not experience disruptive contributions, but did experience task conflict, ratings of perceived effectiveness were also high (\(M=4.05, SD=.61\)). Thus, not surprisingly, teams were perceived as most effective when they had neither conflict, but if a team had disruptive contributions it was better to also have task conflict, than no task conflict. See Table 4 for a summary of these contrasts.

There was no significant interaction between task conflict and roles on perceived team effectiveness [\(F=.001 (1, 3), p >.10\)].

2.5.3 Procedural Conflict as a Moderator of Appropriately Managing Task Conflict

Although the above analysis suggests that task conflict benefits teams by catalyzing appropriate management of co-occurring procedural conflict, we next sought to assess whether procedural conflict suppresses teams’ management of co-occurring task conflict. Previous research on procedural conflict largely suggests that it has a detrimental impact on team performance because conflict over roles, resources, temporal pacing, and contributions tend to divert team attention, increase interpersonal animosity, and decrease productivity and creativity (Behfar et al. 2008; Greer and Jehn 2007; Jehn 1997; Jehn et al. 1999; Jordan et al. 2006; Kurtzberg and Mueller 2005; Matsuo 2006; Passos and Caetano 2005). As such, we were interested if this suppressing influence on outcomes might have a similar influence on the management of co-occurring task conflict. Therefore, we examined whether each meta-category of procedural conflict moderated the relationship between task conflict and negotiating. If the predictions and theorizing about conflict and team outcomes in the literature also apply to the process of conflict management, this would mean that teams would be less likely to appropriately manage their task conflicts if procedural co-occurred with task conflict than if procedural conflict did not co-occur with task conflict. Restated with the language of our meta-categories from our Phase I coding, this would mean:

When procedural conflicts co-occur (as a moderator), Task Conflict will be less likely to be managed with Negotiating.

2.5.4 Results: Procedural Conflict as a Moderator

We tested whether each of the three types of procedural conflict (disruptive contributions, roles, and logistics) acted as a moderator of the use of negotiating when task conflict was present. As the only significant results were for disruptive contributions, we only report those results in detail here and they are also summarized in Table 5.

Table 5 Summary of significant results for procedural conflicts: disruptive contributions as diminishing the management of co-occurring task conflicts
Full size table

Managing Task Conflict with Negotiating: Disruptive Contributions as a Moderator When teams with task conflict also experienced disruptive contributions, they were less likely to manage task conflict with negotiating than when disruptive contributions was absent. In the presence of disruptive contributions there was no relationship between task conflict and negotiating (\(\chi ^{2} (1, \hbox {N}= 26) = .75, p > .05\); Fisher’s exact test \(p=.62\)); in the absence of disruptive contributions there was a relationship between task conflict and negotiating (\(\chi ^{2} (1, \hbox {N}= 18) = 5.66, p < .05\); Fisher’s exact test \(p=.01\)).

Managing Task Conflict: Case Examples Teams with task conflict that also faced disruptive contributions were drawn away from managing their task conflict with the negotiating strategy because of the absorbing need to contain or manage the disruptive contributions. These teams most frequently used organizing strategies rather than Negotiating strategies to address their task conflict. Although organizing strategies contained the immediate challenge to team discipline by relying on formal authority or status of team members, it did not create a foundation for working through team members’ task conflict. So, although disruptive contributions (in the absence of task conflict) did not prevent the team from completing its task (i.e., negotiating with the counterpart), it did suppress the team’s ability to have a well-integrated within-team strategy (i.e., the theoretically ideal scenario via negotiating).

In contrast, members of teams with task conflicts, but without the added disruptive contributions, described using an internally integrative negotiating strategy. Without the distraction of disruptive contributions, these teams were able to focus on team members’ conflicting opinions, concerns, and priorities, discuss sources of resistance, and integrate conflicting interests into an across-the-table strategy.

Influence on Team Evaluations: in order to determine whether the negative effects of co-occurring procedural conflicts on managing task conflict revealed by our data analysis were consistent with participants’ perceived impressions of the effectiveness of the team’s process, we analyzed the interaction of the three procedural conflicts with task conflict on participants’ evaluations of the effectiveness of the team’s processes. As previously reported there was a significant interaction between disruptive contributions and task conflict on perceived effectiveness [F (1, 3) = 6.3, p = .02]. This interaction (depicted in Fig. 3a) was due to lower perceived effectiveness in teams in which there was disruptive contributions but no task conflict (M = 3.25, SD = .81). There were no other significant results.

3 Discussion

The goal of this research was to understand when and why teams challenged by co-occurring conflicts manage those conflicts more or less appropriately. We took an inductive approach to examine the extent to which theorizing about the “good” and “bad” influences of different types of conflict on team performance might also apply to how appropriately a team managed conflict. We focused on conflict management because it is a more proximal outcome of conflict than performance, and because it is an important moderator of the conflict to performance relationship (Behfar et al. 2008; Dreu 1997; DeChurch and Marks 2001; DeChurch et al. 2007; Greer et al. 2008). Our results provide evidence that teams’ likelihood of managing one type of conflict appropriately depends on what other type of conflict the team is also experiencing. Although more empirical work is needed, our results strongly suggest that task conflict serves as a catalyst for appropriate management of procedural conflict and that procedural conflict serves as a suppressor for appropriate management of task conflict. We think this study makes four inter-related contributions to intra-team conflict theorizing.

First, our finding about the effects of managing co-occurring conflict types may help to explain why there are mixed empirical results for the benefits of task conflict for team performance. While task conflict may have benefits in some cases, in other cases it may co-occur with procedural conflict, undermining some of the benefits of task conflict; the influence of the co-occurring procedural conflict (in our study, disruptive contributions) provides negative distractions which seem to replace the focus on meaningful discussions (negotiating in our study) that should occur in response to task conflict. It seems likely then that inconsistencies in the task conflict-team performance research may be due to the presence or absence of co-occurring procedural conflicts impacting the appropriate managing of task conflict.

Our results also help explain why procedural conflict does not always have a negative impact on team performance (see Behfar et al. 2011 for a review). First, if task conflict is co-occurring, our study suggests that procedural conflict is likely to be appropriately managed, which should theoretically lead to effective performance. Second, some manifestations of procedural conflict apparently do not suppress effective management of task conflict, and so may not negatively impact performance. In our study, only disruptive contributions suppressed appropriate Task conflict management, while logistics and roles did not. The difference may be how suddenly these procedural conflicts occur. disruptive contributions typically arise without warning and presented the teams with a pressing need to contain a problem. By contrast, conflicts about roles or logistics may have been easier to anticipate since they typically occur with some degree of warning. This pattern suggest that procedural conflicts that require immediate, urgent action are the most likely to suppress appropriate management of co-occurring task conflict. This makes more understandable the inconsistent effects of procedural conflict on team process and outcomes that has been noted in the literature.

A third contribution is that our inductive approach allowed us to examine what strategies teams used regardless of whether or not they were appropriate for the situation. Our research revealed two management strategies that showed up quite often despite not being ideal for the conflict being faced. These overused strategies were organizing and structuring communication. When teams did not respond to a particular conflict appropriately, it was often because they defaulted to one of these strategies, perhaps because either approach provided teams with some surface-level control of the conflict. For example, teams developed cuing systems, specified decision rights, assigned roles, and created guidelines for participating (or not participating), grasping for a sense of order, rather than doing (or being able to do) the hard work of negotiating to resolve task conflicts or finding straightforward logistics solutions to a logistics Conflict. Organizing and structuring may have had the advantage of allowing teams to negotiate without the other side picking up on their lack of internal agreement, but neither was an optimal strategy because neither required that the team fully address team members’ conflicts of interests. Overdependence on such surface-level quick fixes might be due to a lack of time, a lack of trust between team members, role ambiguity (Tidd et al. 2004), the need to contain negative emotionality (Yang and Mossholder 2004), or low levels level of team identification (Schaeffner et al. 2014).

Finally, as a practical matter, our results provide a few suggestions for people who are managing teams. First, while it may be desirable to not have any conflicts at all, team leaders need to be prepared to manage multiple conflicts at once. They need to be “ambidextrous” in their approach to conflict. If there is task conflict, which is not unlikely given that teams are often formed for the very reason of bringing together people with different skills, interests, and goals, a leader needs to not just focus on managing that task conflict well, but also to watch for procedural conflicts that can be highly distracting—in particular those typified by our meta-category of disruptive contributions. One possibility may be to talk ahead of time about the possibility of disruptive contributions, inoculating the team from some of its effects or at least diminishing the “surprise” element of disruptive contributions. Similarly, when facing procedural conflicts such as those typified in the logistics or role meta-categories, team leaders might be well advised to harken back to the core purpose of the team and the interests represented by different team members—as occurs when there are task conflicts. This might serve as a productive catalyst to the effective management of those procedural conflicts.

Limitations and Opportunities for Future Research As with any study, there are limitations to the interpretation of our results. We relied on a small sample and team members’ own perceptions. Although team members provide one important perspective on team functioning, other constituents’ opinions (for example, those of managers overseeing the team, and/or to parties who sat across-the-table) are also relevant to the question of how effectively the team managed conflict when different types of conflict co-occurred. In addition, our results reveal associations, not cause. Future experimental research could manipulate co-occurring conflict to address causation. Future research might also test a full model showing that co-occurring conflict has an interactive effect on team outcomes and that using the appropriate management strategy mediates this effect. There may also be effects to pursue concerning the order in which conflicts occur. Does task conflict occurring before procedural conflict or procedural conflict occurring before task conflict impact the choice of appropriate management strategies? Studies, with larger sample sizes, could investigate the influence of multiple (rather than simply co-occurring) conflicts and how the severity of conflicts might influence how appropriately they are managed. Although relationship conflict did not emerge as a distinct category in our dataset, the role of emotion and interpersonal dynamics among team members is an area for future work.

In closing, the negotiating teams we studied faced significant pressure to approach the party across-the-table with a sound strategy, and to maintain a united front in the face of a dynamic and evolving negotiation. Our rich quantitative and qualitative data provided insight into the complex world of managing negotiating teams. At the same time, it provided insight into mechanisms of catalyzing and suppressing that explain how negotiating teams, and very likely other types of teams with complex tasks, manage co-occurring conflicts. We hope that this study will motivate future research that elaborates on these mechanisms that explain the complex links between task and procedural conflict and management strategies in organizational teams.

4 Appendix

See Tables 6a–6c and 7a–7d

Table 6 Representative statements for the team conflict categories: concept mapping and pattern coding analysis
Full size table
Table 7 Representative statements for the team management strategies: concept mapping and pattern coding analysis
Full size table