From hierarchical to matrix structure: tensions in negotiating shared leadership configurations

Abstract

Purpose

Despite much attention being devoted to shared leadership, the negotiation of such arrangements remains underexplored. In parallel, the revival of interest in matrix structures reveals their challenges but neglects the dynamics of shared leadership. In this case study, the author analyzes the tensions experienced by senior managers of a healthcare organization transitioning from a hierarchical to matrix structure as they negotiate their leadership roles in this new arrangement.

Design/methodology/approach

The author interviewed 16 senior managers, observed their meetings and analyzed documents. These data were combined with secondary data including previous interviews and observations of this top leadership team. The author then conducted an inductive data analysis.

Findings

The author's analysis reveals that the tensions experienced by senior managers as they negotiate their roles reflect the co-existence of leadership surpluses (too much leadership) and deficits (too little leadership) in matrix organizations. The author argues that surpluses and deficits are not mutually exclusive but are interrelated and shows how leadership surpluses can create leadership deficits.

Practical implications

The author’s findings suggest that in contexts of leader abundance, actors should explore leadership voids. Particular attention should be paid to incidents of intrusion and exclusion, moments of transition and intense role negotiation, as those contexts are particularly conducive to leadership deficits.

Originality/value

While previous work on matrix structures focuses on leadership surpluses, the author discusses leadership deficits. The author explores how more leaders do not necessarily mean more leadership, but instead how more leaders may result in leadership voids.

Citation

Gibeau, E. (2024), "From hierarchical to matrix structure: tensions in negotiating shared leadership configurations", International Journal of Organization Theory & Behavior, Vol. 27 No. 1, pp. 14-34. https://doi.org/10.1108/IJOTB-12-2022-0235

Introduction

Transitioning from a hierarchical structure to a matrix structure forces senior leaders to redefine leadership arrangements while managing often complex organizations operating at full capacity. The complexity and range of issues tackled by these leaders reflect the significant impact this negotiation process and its outcome can have (Garman et al., 2020), however, we know very little about the way these changes in leadership arrangements proceed, especially during the early stages (Chreim, 2015). We do know that the way leaders share their roles in the early stages tends to make a deep impression on the way they work together (or not) over time (Gibeau et al., 2020), highlighting the importance of better understanding how leaders negotiate the nascent formation of these arrangements.

Multiple labels refer to similar arrangements where leadership is shared by more than one individual, including but not limited to “plural”, “shared”, “distributed”, “collective” and “pooled” (Denis et al., 2012). I rely on Pearce and Conger’s (2003) definition of shared leadership as, “dynamic, interactive influence process among individuals in groups for which the objective is to lead one another to the achievement of group or organizational goals or both” (p. 1). Interest in the antecedents and consequences of shared leadership has surged lately (Zhu et al., 2018; Abson and Schofield, 2022; Lyndon et al., 2020) particularly in the quest to decipher the impact of the model on team effectiveness and performance in various conditions (Salas-Vallina et al., 2022; Wang et al., 2014). Recent research also offers insight concerning the functioning, contexts and conditions for success of various forms of shared leadership arrangements (Sergi et al., 2016; Gibeau et al., 2016).

The configuration of shared leadership arrangements, that is the division of roles among a group of leaders, has been studied in healthcare organizations (Gibeau et al., 2020; Scott et al., 2016). Authors have described how groups of leaders function, ranging from independent to synergic work (Zhu et al., 2018; Chen and Zhang, 2022). However, these descriptions neglect to inform us of the messiness involved in negotiating the shape of the configuration, focusing on the positive aspects of the arrangement. Other scholars have asked how to account for ambiguity, tensions, conflicts and power dynamics in shared leadership arrangements (Fairhurst et al., 2020; Denis et al., 2012).

Importantly, most studies examine the definition of configurations among a few individuals (in co-leadership, for instance) (Gibeau et al., 2020), the negotiation of roles in groups of leaders with stable membership (Fox and Comeau-Vallée, 2020) or how leaders adapt to (or reject) the addition of one or more new leader(s) (Beirne, 2017). In those contexts, most roles within the configurations are grounded in past practices and role definitions, either through negotiation of leadership roles in mundane interactions or by having only one or a few members carve out a role in relation to existing ones. However, contexts have seldom been explored in which all members of a leadership group at the same hierarchical level – which I call a constellation (Hodgson et al., 1965) – experience their roles being redefined simultaneously.

Arguably, such contexts can heighten ambiguity that leaves much to be negotiated by all actors. The interactive nature of shaping leadership constellations established in previous work (Clifton et al., 2020) points to the complexity of a context where completely new groups of leaders are all defining their roles in relation to one another at the same time. The implementation of matrix structures – such as my empirical setting – set the stage for such complex role negotiation, since role and accountability overlaps are inherent to the structure and create role conflicts and ambiguity (Ford and Randolph, 1992).

While designing structures and clarifying authority is a key responsibility of senior management (Ihrke, 2004), contexts in which senior management are redefining their roles and relationships while redesigning a structure are underexplored and pose a particular level of complexity. This is especially true in public administration, where “a mix of personality and ego, complex sets of rules, public pressures and political environment make managing very difficult” (Schneider, 2016, p. 60) yet reaching objectives depends on leadership (Holmes, 2009).

Given the importance of distributed leadership arrangements and our limited understanding of the early stages of how those arrangements are defined, I build on the literatures on shared leadership and matrix structures to explore the types of tension that members of nascent constellations face as they negotiate their leadership configuration. Specifically, this study is based on 16 interviews, observations and document analysis performed as a healthcare organization in Quebec is transitioning at a strategic level from a hierarchical to a matrix structure, and as a new model of shared leadership is being negotiated.

This setting presents a unique opportunity to respond to the many recent calls for a more nuanced understanding of the development of shared leadership (Edelmann et al., 2023; Zeier et al., 2021). Through the lived experience of tensions in a newly formed constellation of senior managers, this study refines our comprehension of the construction of shared leadership and counterbalances the overemphasis on the benefits of the leadership arrangement in the existing literature (Chen and Zhang, 2022). Such inductive, qualitative, case-based study allows researchers to gain a rich, in-depth perception of the lived experience of leaders and a subtle understanding of situations of great complexity such as the implementation of matrix structures (Döös and Wilhelmson, 2021; Wellman, 2007).

Although this study draws from previous work taking a positivist approach, the theoretical underpinning for my study lies in constructivism (Boswell and Badchuk, 2023). I mobilize existing theory as a basis to design my study, to guide my data analysis and to carve out my theoretical contribution, but I strive to let my findings emerge from the data. Additionally, I consider not only the physical reality of participants but also the meaning they ascribe to this reality (Lincoln and Guba, 2016).

Theory

In this section, I explore the literature on tensions in the development of shared leadership (Edelmann et al., 2023), the negotiation of leadership configurations (Gibeau et al., 2020) as well as studies on leadership in matrix structures (Egelhoff and Wolf, 2017). These streams of literature provide a foundation for the theoretical examination of the tensions in defining leadership configurations in nascent constellations.

Developing shared leadership

Factors that shape shared leadership development have been explored, ranging from the individual level to context (Döös and Wilhelmson, 2021). While some scholars of shared leadership emphasize it as naturally emerging (Carson et al., 2007), others focus on it as a planned phenomenon (Zhu et al., 2018) or conceive it as a combination of both planned and emergent leadership (Gibeau et al., 2016). Edelmann et al. (2023) discuss this divide between emerging and planned shared leadership, noting how emergent leadership can create confusion, uncertainty and conflicts, thereby complicating decision-making.

At the individual level, tensions within shared leadership arrangements can be triggered by a lack of transparency (Klinga, 2021), low levels of mutual respect and trust, or misaligned values, goals, personalities, skills and experience (Döös and Wilhelmson, 2021; Zhu et al., 2018; Wu et al., 2020). Resistance to sharing power or decision-making, or letting other leaders take the foreground (Eckman, 2018), as well as neglecting to recognize other leaders' expertise (Abson and Schofield, 2022) also set the stage for inter-leader tensions.

Role ambiguity, overlaps and complete disconnect within shared leadership arrangements are known to create tensions (Lyndon et al., 2020; Saxena et al., 2018). In healthcare settings, such disconnect may result from silo effects between the medical and administrative groups (Rogers et al., 2020). In this context, Boak et al. (2015) propose wide involvement in co-designing and implementing a shared leadership arrangement to foster internalization and engagement.

While the literature points out that a lack of facilitating conditions and contexts creates tensions in the development of shared leadership, we gain little insight about how their presence might play out in various contexts. Edelmann et al. (2023)’s study provides some needed nuance, yet the authors call for additional insight considering the complexity of the development of shared leadership. My empirical setting provides a unique opportunity to explore the initial stages of the implementation and negotiation of shared leadership in a rare and extraordinarily complex context.

Negotiating leadership configurations

Historically, leadership has been approached as an individual, heroic and romantic phenomenon, with “a tendency to see leaders as heroes and thus to use ‘leadership’ as a catch-all explanation for the successes and failures of organizations” (Denis et al., 2012). Reacting to this tradition of romanticizing and individualizing leadership, many researchers took a different approach, constructing leadership as a collective phenomenon (Montaudon-Tomas et al., 2023; Edelmann et al., 2023). In the literature, I note investigations of the functioning, contexts and conditions for success of various forms of shared leadership arrangements (Gibeau et al., 2016). Nonetheless, issues surrounding the delineation of roles and responsibilities remain central to scholarly discussions on shared leadership in various empirical settings (Eckman, 2018).

Some configuration typologies and associated tensions have been uncovered (Feng et al., 2017). In a literature review, Gibeau et al. (2016) identify four separate co-leadership configurations: disconnection between leaders, duplication of work, dominance by one leader and distribution – a more balanced arrangement in which the roles are different yet their work is coordinated. Disconnection creates tensions when one leader is bypassed, duplication’s tensions derive from leaders being played against one another, and dominance leads to tensions emanating from work overload for one leader. Some authors echo these configurations, noting how the degree of over-emergence, overlap, differentiation and distribution of leadership may generate tensions (Chen and Zhang, 2022; Lanaj and Hollenbeck, 2015).

Chreim (2015) studied role distribution and leadership practices during mergers and acquisitions. She put forward the notions of leadership deficits and surpluses to characterize configurations that can emerge in distributed leadership arrangements, where configurations can range from vacant to overcrowded spaces. Chreim (2015) explains that a void in leadership can be created when a leader decides to remain passive, even without the imposition of external limits on their agency. Such surpluses and deficits point to Hodgson et al.’s (1965) foundational work on executive role constellations. Hodgson et al. (1965) proposed three dimensions of collective leadership teams: specialization, differentiation and complementarity. Combinations of these three dimensions in varying proportions may create leadership surpluses and deficits.

Gronn (2002) and Chreim (2015; Chreim and MacNaughton, 2016) argue that surpluses and role overlap potentially create either redundancy that results in ambiguity and dysfunction, or positive mutual reinforcement, provided that role boundaries and accountability are clear. The authors see the shared role space - where leaders play a joint role – as an ambiguous space where leaders can expand their authority, collaborate, negotiate, contest, dominate or withdraw. Chreim’s (2015) study proposes a more radical conception of constellation membership and configuration, with constellation members located at different hierarchical levels.

From an empirical investigation of the shared role space in co-leadership arrangements in healthcare organizations, Gibeau et al. (2020) identified six configurations reflecting levels of overlap and weight given to the professional and managerial logics in dyadic work: the dyad of one, management duo, management unit, professional consulting, boundary duo and mission unit. In the first three configurations, the managerial logic remains dominant, while co-leaders work with an increasing degree of integration. In the last three configurations, both managerial and professional logics are present, mixed in separate ways, with co-leaders working with more or less synergy. The authors show how tensions in co-leadership arrangements are not synonymous with failure, but often serve to preserve the perspectives of the various leaders as they collaborate.

Scott et al. (2016) propose different configurations, reflecting the balance of homogeneity/heterogeneity, network density and the degree to which leaders are active-influential or passive-receptive. The authors note that tensions emerge when too many leaders with overlapping specialized expertise do not have the space to exert influence. Power struggles are also expected when active-influential leaders in dense networks seek unilateral control over decisions, situations that sometimes involve psychological territorial loss or violation (Chen and Zhang, 2022).

When reflecting on leadership arrangements between formal and informal leaders, Beirne (2017) concludes that appointed leaders can work quickly “to recover control, to co-opt ideas or embed distributed activity within traditional organizational norms”, making leadership by non-appointed leaders difficult and fleeting. In those studies, the focus is overwhelmingly – implicitly or explicitly – on situations in which traditional hierarchies may be threatened by the emergence of new leaders. Some individuals occupy a more formal or relatively fixed role, while others strive to make space for themselves as leaders. However, researchers have yet to fully understand contexts in which all leaders of a group are appointed to formal positions at the same hierarchical level yet face significant role ambiguity.

Some studies provide initial insights on groups of appointed leaders who must share leadership in conducting a change process. In those cases, the roles of these formally appointed leaders are mostly explicitly defined despite their having to negotiate their new role in implementing the change. Focusing on the tensions experienced by such a group of leaders, Ho and Ng (2017) expose the existence of competing, overlapping or conflicting visions, priorities and responsibility boundaries between leaders that give rise to tensions grounded in the boundary-spanning nature of their work.

The current literature provides insight into some tensions associated with the negotiation of leadership in existing groups or with the co-existence of emergent and mandated leadership. However, the dynamic in nascent leadership constellations where all members negotiate their role in the configuration is not well understood. Yet research suggests that early configurations profoundly impact the way leaders subsequently share leadership over the longer run (Gibeau et al., 2020). Additionally, in the context of top management teams who lead change and simultaneously manage complex organizations, the process and outcome of the leadership role negotiation process can reasonably be expected to have a considerable impact on many aspects of the organization, from governance to operations. In this study, I explore the early stages of constellation creation when all roles are somewhat up in the air and remain to be clarified, leaving much opportunity for negotiation within the space.

Leading in matrix structures

Matrix structures were the object of much attention in the seventies and eighties, when scholars focused particularly on the definition and adoption of the structure, as well as its pros and cons (Davis and Lawrence, 1978; Galbraith, 1973). Matrix structures are “a type of organisational structure that is built around two dimensions such as functions, products, or regions and in which people have two bosses” (Galbraith, 2009, p. 3). The structure’s rationale allows organizations to respond simultaneously to multiple priorities, such as managing both change and operations (Burns and Whorley, 1993).

Along with the traditional top-down line of command, matrix structures possess a lateral line of command of functional authority, creating dual intersecting lines of authority. For instance, a functional manager in charge of a special project may temporarily access and direct personnel who work in a traditional department in order to complete the project. Matrix members hence have two bosses (Egelhoff and Wolf, 2017), contradicting Fayol’s (1949) fundamental management principle of unity-of-command that each worker must have only one supervisor.

While matrix structures are flexible and accommodate economies of scale (Galbraith, 2009), major confusion and recurring power struggles can result from the dual lines of command (Galbraith, 2009; de Laat, 1994). Managers may fight to establish priority for their activities, while employees may feel pulled from all sides, not knowing to whom they should devote their time. Burton et al. (2015) describe the high-tension climate of matrix structures, highlighting that individuals at the junction point report that they receive too much information, heavy workloads and conflicting instructions. The authors note that “it is ‘what is happening’ at the junction points that will make or break the [matrix] organization” (p. 42), highlighting the importance of how leaders coordinate goals, clarify roles, explain authority lines, ensure the flow of information and bridge silos – or preserve them.

These role conflicts are inherent to the matrix, as is the ambiguity that stems from unclear expectations, nebulous accountability and overlapping reporting lines (Ford and Randolph, 1992). Galbraith (2013) argues that the matrix structure requires “boundary spanning leadership” by executives able to lead a “diverse, sometimes dissonant orchestra” (ibid: 7).

Previous research on leadership in matrix organizations emphasizes the skills, behaviors and approaches that make individual leaders appropriate for matrix structures (Sy and Cote, 2004; Sy and D’Annunzio, 2005; Wellman, 2007). While it is widely acknowledged that matrix structures set the stage for struggles among executives, early efforts to decipher adoption of matrix structures or more recent work to explore individual leadership traits and behaviors conductive to matrix contexts mostly overlook the negotiation of shared leadership roles among senior executives, especially at early formative stages. More generally, matrix structures naturally encourage the co-existence of multiple leaders, yet leadership as a shared endeavor in matrix organizations has not attracted much research attention. Notably, recent work on matrix structures highlights the lack of rigorous study of multiauthority structures in general (Levinthal and Workiewicz, 2018) and on leadership in matrix arrangements (Sahlmueller et al., 2022), supporting the need for the present research. Recently, multiple authors called for investigations of shared leadership in various contexts to gain a more nuanced understanding of the phenomenon (Edelmann et al., 2023; Chen and Zhang, 2022).

The implementation of a matrix structure triggering the development of shared leadership in the senior management team represents an ideal setting to respond to this call. This study of a healthcare organization implementing a matrix structure bridges the literature on matrix organizations and shared leadership to gain insight into the sharing of leadership in nascent constellations as senior executives strive to manage operations, implement a matrix structure, bring about change in the organization and negotiate their roles.

Method

Taking a constructivist approach in which both the physical reality of participants and the meaning they ascribe to this reality are central (Lincoln and Guba, 2016), I conducted an inductive qualitative case-based study of one healthcare organization in Quebec involved in a major transition from a hierarchical to a matrix structure (Boswell and Badchuk, 2023). Following Charmaz’s (2006) approach, I used the literature as a basis to identify a research gap, construct a research question, design the study, select sensitizing concepts to guide my data analysis and anchor my discussion of the contributions. However, I let themes and contributions emerge from the data, grounding these contributions in the constructed reality of participants (Koleva, 2023). Throughout the research process and as required with this inductive approach, I avoided letting the theory or my own positioning as researcher shape results (Boswell and Badchuk, 2023).

The empirical setting

The uniqueness of the case stemming from the early implementation of a matrix structure, combined with the potential transferability of the insight gained from exploring leadership negotiation, renders this case particularly appropriate to answer the research question (Gioia, 2021). To make the structural change studied explicit, Figure 1 provides a simplified illustration of the old and new structure.

The change triggered a redefinition process for the roles in the senior management team. The previous traditional hierarchical structure was composed of administrative directorates and clinical directorates. Members of administrative directorates advised and supported members of clinical directorates. Administrative directors possessed a functional authority based on their specific expertise, such as human, financial or material management, allowing them to control certain aspects of the work of members of clinical directorates. Clinical directors were responsible for a portion of the services offered to the public by their organization. They were responsible for managing operations, but also implementing changes and tracking performance measures. Ten members of the senior management team were from the previous structure and six new members were recruited from other organizations as the matrix structure was being implemented.

The new top management team was composed of 16 directors in charge of three types of directorates: administrative, functional and clinical. These directors were all at the same hierarchical level. Directors of administrative directorates maintained similar roles before and after the change, advising and supporting members of other directorates in accomplishing their roles. In the matrix structure, these directors retained the same functional authority as in the previous hierarchical structure, able to intervene and control specialized aspects of the work of members of functional and clinical directorates according to their expertise. All administrative directors transitioned from the previous to the new structure.

Four functional directorates were introduced in the organization, one each specializing in the quality of nursing practices, in medical practices, in planning and change management, and one specializing in overall performance and quality. Functional directors also had an advisory role and functional authority, which related to their responsibility to develop specific aspects of the healthcare center. Since these directorates were new, the directors were new to their roles. Three were also new to this organization. The fourth had been previously employed in the health center, but at another level in the hierarchy.

Clinical directors were responsible for managing the day-to-day services offered to the population as well as managing the service providers. In the previous structure, they used to single-handedly manage change plus the operations of their directorate. The goal of introducing the matrix structure was to extract from clinical directorates all aspects of their work that was organization-wide, that was central to the organization yet might be neglected because these line directors were overwhelmed by and focused on their work managing operations. Now they had to adapt to having new functional directors intervene in their teams. Two of the three clinical directors were in a similar position within the organization before the implementation of the new structure, while one had been at a lower level in the organization.

Data collection

Inspired by Wellman (2007) and much previous work on shared leadership (Döös and Wilhelmson, 2021), I conducted an inductive, qualitative study to gain insight into the lived experience of senior managers. This approach allows researchers to explore in-depth processes that involve much complexity (Creswell, 2003). I interviewed all 16 members of the executive constellation (that is, the top management team). Table 1 details the participants’ profiles.

During the interviews conducted between November 6, 2007 and February 14, 2008, participants were asked about their professional history and current roles. Following the critical incident interview method (Flanagan, 1954), participants were invited to explain incidents related to role overlaps, contradictory demands, ambiguous requests and times when issues remained unaddressed (that is, no one felt responsible). This approach gives participants freedom to focus on what is important to them without biasing or imposing established constructs (Franken and Plimmer, 2019). The critical incident technique is appropriate when exploring tensions that participants experience in highly complex contexts such as the implementation of a matrix structure in which all senior management roles must be negotiated simultaneously (Wellman, 2007). Participants were then asked to explain their strategies to clarify roles, as well as whether they perceived a disequilibrium in the roles and workload of members of the top management team.

Although the verbatim transcript from interviews constitutes my primary source of data, I also attended and took notes at eight strategic meetings over the span of four months to observe the senior management team as they were interacting in their natural environment. When attending meetings, I wrote down everything that each participant said that seemed closely or remotely related to the research question, attempting to use the same words and sentence construction as participants. I also noted participants’ body language and positioning in the room, as these can reflect tensions and coalitions. Whenever possible, I would make notes of my impressions and interpretations in a separate document to guide data analysis toward what appeared to be more central, emotional or significant issues during meetings. Such day-to-day observations can contribute to a rich and nuanced understanding of the context in which leaders evolve, provide a first-hand reflection of leadership negotiation in real-life situations, enable researchers to access the discourse of participants in various contexts, and help researchers differentiate routine and situation-specific dynamics (Lincoln and Guba, 1985).

I analyzed all relevant documents distributed during this period, including reports, organizational charts, job descriptions and action plans regarding projects involving more than one member of the senior management team. The stability of documents makes their analysis particularly useful when studying sensitive situations since they remain unaffected by the passage of time, changed circumstances and the presence or characteristics of the researcher (Morgan, 2022). Documents such as meeting minutes also enable an independent reconstruction of events to corroborate participants’ narrative recounting of key exchanges during official meetings (Bowen, 2009).

I complemented these sources with secondary data, including verbatim transcripts of 21 interviews conducted with all members of the top management team between May 13, 2006 and December 3, 2007 as well as observation notes taken at 26 strategic meetings held between March 9, 2006 and August 27, 2007. All secondary data were collected for a project aimed at better understanding the implementation of new work arrangements and innovative practices in healthcare organizations. The present study on the negotiation of shared leadership was initiated as the researchers conducting the wider study on innovative work arrangements identified defining the leadership roles of the top managers as they were transitioning from a hierarchical to a matrix structure as a fundamental issue relevant to both theoretical development and practice. Although the suitability of secondary data is always a concern in qualitative research (Rodriguez, 2021), the embeddedness and alignment of both projects render this secondary data set particularly relevant to this study.

In studies of elites – “individual[s] who holds or has held some powerful position that has afforded the individual unique knowledge or information from a privileged perspective” (p. 1) – triangulation allows researchers to get a full and nuanced picture of the empirical situation and is especially valuable when investigating politically sensitive topics (Natow, 2020). Using multiple methods to collect data used in this study contributed to a rich and nuanced understanding of the complex empirical situation studied. Approval from the ethics board of the investigators and from the participants’ organization was obtained. Before the interviews and observations, participants were required to sign an informed consent statement, were presented with the project objectives and had the opportunity to ask questions.

Data analysis

Following a constructivist view (Boswell and Badchuk, 2023), this study contemplates data as a source of information about the participants’ physical reality as well as about the meaning they ascribe to this reality (Lincoln and Guba, 2016). An inductive approach to data analysis informed by grounded theory (Franken and Plimmer, 2019) was therefore used. Of the various approaches to grounded theory (Koleva, 2023), this study is closer to Charmaz’s (2006) approach of “social reality [as] multiple, processual and constructed.” (p. 264). Hence, the data analysis process was bracketed by two literature reviews, and characterized by consistent memoing to guide iterative linking between data and theory (Koleva, 2023).

Specifically, I initially coded all primary and secondary data for first-order codes looking for any excerpt referring to tensions in negotiating configurations. I used sensitizing concepts extracted from the literature to guide this first coding round (ambiguity, overlap, role conflict and neglected issues), but I intentionally retained quotes reflecting tensions that would not fit in those categories. I then grouped the first-order codes into second-order codes reflecting groupings of similar tensions, which included role overlap, transitional ambiguity, disconnect between planning and operations, competing accountability, blind management, perceived interference, lower-level paralysis and stagnation within a constellation. In the final step – theoretical coding – I arranged the second-order codes into aggregate dimensions reflecting the tensions emerging from the data (Koleva, 2023). Planning/operations, intrusion/exclusion and inward/outward reflected those tensions.

Throughout the analysis and theory development process, I paid special attention to all data contradicting existing theories and emerging models. Of note was the recurring theme of paralysis and leadership voids in this context known for its abundance of leaders, which will be detailed in the results section. Following Charmaz and Thornberg’s (2021) guidance to ensure the quality of qualitative research, codes and categories were consistently treated as provisional until alignment between the theoretical and empirical foundations of this study emerged. Table 2 summarizes the data structure formed.

Results

In this study of leadership configuration negotiation during structural transition, most tension incidents occurred between functional and clinical directors. This tendency may be explained by the comparatively stable nature of administrative directors’ roles over the transition, and the shared understanding of those administrative roles among all directors.

The planning/operations tension

The first type of tension, planning/operations, refers to incidents of transitional ambiguity, a lack of clarity surrounding the handover of responsibility for specific projects as they progress over time from planning to operations management. At the healthcare center, the CEO’s new organizational chart assigned planning responsibility for projects to a functional director, and operational management responsibility associated with these same projects to relevant clinical directors. However, clinical directors’ experience in the previous hierarchical structure gave them responsibility for all project phases, including planning, implementation and managing operations. The CEO explains his vision for assigning planning and operations to different individuals, “Planning has been neglected too often. We have to take project planning out of clinical directorates and give it to someone inside or outside our organization. We have someone on the inside.” (E). The transfer of responsibility and leadership associated with the passage from planning to operations phase represented a considerable challenge because the exact moment and content of this transition was unclear to every director involved: “I hear comments like planning, ‘where does it stop? Where does planning end? When do we start?’ There is a fine line. It’s difficult to say: ‘planning is until this point, and then it’s operations … ’ We can’t do that. Clinical managers say, ‘I want to be involved in planning.’” (E).

This confusion caused tensions, since the clinical directors who were expected to be responsible for the operations felt accountable for the outcome yet powerless in shaping the project during its planning phase: “the planning problems, we have to fix them afterwards – we are not associated with the planning process, but we have to fix the problems after. (C)” Another clinical director explains how this division of responsibility during different phases creates a disconnect between the plan design and the reality of operations, making implementing the project quite challenging: “Functional directors are not aware of our day-to-day problems and constraints. (…) That’s why there needs to be a link. We can’t detach planning from operations, even during the planning process. We have to find a way to adjust the plan to our reality.” (C) This clinical director finds it especially difficult as she is accountable for the delivery of services resulting from the project, “I feel accountable for the services – it’s in my hands!” (C). An executive who witnessed an incident related to planning/operations transitional ambiguity comments on its impact:

In a project led by a functional director (…), we have to create work groups. During a meeting, I said, ‘where are we going to put the professionals?’ The functional director told me, ‘It’s an operational concern.’ I said, ‘no. When you plan a project like that, offices are a success condition, so it should be taken care of during the planning phase.’ (…) We get to a point where nothing is being done. (E)

Planning/operations tensions hence refers to tensions resulting from handing over responsibility and leadership during the lifespan of a project. These tensions lead to much confusion and negotiation, which consequently somewhat paralyzed project progress and completion.

The inward/outward tension

The inward/outward tension refers to the focus of directors’ attention as they were negotiating their nascent leadership configuration. An inward focus means that members of a configuration focus on each other, spending much of their energy and time on defining each others’ roles. An outward focus involves members of a nascent configuration who pay attention to actors outside the constellation but neglect fellow members of the constellation.

In an exemplary incident, many functional and clinical directors claim leadership for a specific project, arguing that they are accountable and therefore require leadership authority. One functional director claims leadership based on his responsibility for improving performance measures, another functional director’s mandate includes project planning, two functional directors’ roles involve managing the quality of care offered while two clinical directors are accountable for the delivery of services to patients. Similar perspectives from a clinical director and a functional director are reflected in the two quotes:

Everyone claims their part, we all feel accountable. A functional director is accountable by virtue of the law. A functional director is accountable by virtue of his mandate. But where does it all happen? As a clinical director, I also have a responsibility. An incredible confusion that prevents us from moving forward (C).

Some clinical directors believe and insist that they are responsible for this project, but the law says clearly which director is responsible and accountable and must exert control. (F).

More importantly, as directors negotiate their roles, others highlight the vacuum and stagnation created outside the constellation, expressed as follows:

I work with a functional director [on a project] and we have specific performance measures to track. Another functional director is responsible for tracking the performance measures, so he wants to be involved. I don’t understand what the second director has to do with the project. While those two need to clarify what is going on, there is a vacuum. I expect them to feed me some direction. (C)

The roles are unclear, and it creates a lot of ambiguity. And then there are things that don’t get done. (C).

These statements reflect an inward focus, where leaders devote so much time and energy to role negotiation, that actors and projects outside the constellation are neglected. The CEO commented on the impact that a different incident had on the directors’ leadership. In this case, multiple directors attended meetings on a specific project. However, the work those directors did to publicly claim their leadership (that is, an outward focus) weakened the line director’s leadership:

I was trying to imagine all the directors around the table talking to the middle managers, professionals, clinicians. As a manager, I started to wonder if that wasn’t going to weaken the clinical director’s leadership, as the line manager. Won’t that endanger her leadership in front of her team? (E)

This quote reflects a participant’s preoccupation that leadership may be rendered fragile when an abundance of individuals publicly claim leadership.

The intrusion/exclusion tensions

The third type of tension, intrusion/exclusion, refers to challenges associated with functional directors’ ability to intervene in the clinical directorates. The matrix structure positioned all middle managers and employees under the authority of one clinical director and all functional directors. This structure, as designed by the CEO, empowered functional directors to use resources managed by clinical directors to achieve their functional goals. In theory, clinical directors were to be informed of this use of resources but were not in a position to refuse functional directors’ use of employee time. In practice, clinical directors were not always informed, complicating their work of managing employees and delivering services.

These incidents caused clinical directors to feel intruded on: “The disadvantages of the structure are that when roles are not clear, we feel an intrusion. (…) Sometimes my team does work that I am not informed of.” (C). Another clinical director explains having to adjust to working in an organization in which she is not accountable for part of the work done in her directorate:

In my mind, I was accountable for the activities done by my team. Finally, [the CEO] said clearly that the functional director is accountable for these activities. For a few months, my neurons had to adjust, because I wasn’t used to working like that. It means that a functional director works directly with members of my team, and sometimes I know about it, sometimes I don’t. (C).

The director explained her laissez-faire approach to overcome these situations, “If people feel that I want to control everything and prevent them from entering my directorate, they will enter without telling me.” (C). This approach ensures that other directors do not feel excluded from her directorate and mitigates against feelings of intrusion.

A different clinical director highlights the confusion resulting from intrusion/exclusion incidents and the resulting liability to a functional director’s leadership capacity, “When directors who don’t have staff work with your staff, it creates confusion. We have to learn how to work like that without discrediting our colleagues [the functional directors].” (C) A functional director shares a similar view, shedding light on how the overlap makes her feel excluded, weakens her leadership and slows down her projects:

I manage change. It’s not simple. They don’t let me in that much. (…) It's not necessarily a conflict, but they meet with their staff to discuss a project that I am leading but don’t invite me. It’s difficult to implement change when you are not physically present. We take some steps forward; we accomplish a few things.” (F)

Echoing this view, another functional director reflects on being forgotten by clinical teams:

Functional directors, we have to learn to stand up for ourselves. Sometimes, not intentionally, clinical directors won’t consult us. They are in the action, they move forward. We are going to try to see what they are doing so we are not forgotten and left behind. We want to be involved from the start rather than manage a crisis afterwards. (…) We have hierarchical authority over no one. We have a functional authority and have to learn to live with that. But it’s difficult because we are not always in the action. (F)

These comments by functional directors reflect their feelings of exclusion as they try to conduct their work.

Discussion

Based on a case study of the senior management team of a healthcare organization transitioning to a matrix structure, which compelled them to negotiate their roles, I endeavored to gain insight into the sharing of leadership in nascent constellations. My inductive analysis shed light on three recurring tensions: planning/operations, inward/outward and intrusion/exclusion Synthesizing and building from this analysis, I offer two contributions through this study. First, I show that leadership surpluses can create leadership deficits and illustrate them. Second, I contribute to the literature on matrix structures by illustrating the existence of leadership voids in contexts of leadership abundance.

The tensions uncovered in this study echo previous work on matrix structures and leadership configurations that pointed to potential ambiguity and overlap in roles (Chreim, 2015; Davis and Lawrence, 1978; Hodgson et al., 1965). My findings reinforce the idea of leadership surplus put forward by Chreim (2015), with all the vital tensions rooted in the claiming of leadership by multiple individuals for the same activities or over the same individuals. Chreim (2015) proposes that leadership constellations can be situated in a continuum of configurations that range from vacant to overcrowded leadership space. This research responds to her call for more research on how leadership surpluses and deficits play out.

Developing this idea further, I contribute to the literature by showing that leadership deficits and surpluses not only constitute mutually exclusive opposing ends of a spectrum of configurations reflecting role division in the shared role space (Gronn, 2002) of the leadership arrangement (that is, inward) but also cause us to consider configurations in relation to actors outside the constellation (that is, outward). In my case, the tensions within the constellation mostly reflect an overcrowded space. However, the findings also suggest that an internal leadership surplus at one level can be, at other levels in the organization, a leadership deficit. As directors focused inward on role negotiation to resolve leadership surpluses within constellations, they failed to exercise outward leadership, creating a void.

These findings suggest that understanding configurations requires a refined understandings of role specialization, differentiation and complementarity (Hodgson et al., 1965) within constellations (inward) as well as of roles performed in relation to other organizational actors in day-to-day activities (outward). In this case, claiming leadership within the constellation prevailed over enacting leadership outside the shared role space.

Holm and Fairhurst (2018) and Yammarino et al. (2015) reported on the interconnections between shared and hierarchical leadership and called for more investigation. This case study shows that implementing shared leadership can trigger these new leaders to focus inward in an intense competition to lead one another within the constellation. This distraction from outward responsibility can isolate them from subordinate levels and as tensions become visible, erode their hierarchical leadership. This finding not only confirms the need for a safe and private space for role negotiation (Davis and Lawrence, 1978; Schaubroeck et al., 1993) but also constitutes initial insight into how exposure of leadership surplus-related tensions may lead to deficits in hierarchical leadership as leaders’ legitimacy is discredited.

Leadership surpluses and deficits may also coexist when constellation members’ roles overlap on some activities but are highly differentiated for other activities. Planning/operations incidents reflect this dynamic. While functional directors and clinical directors claimed leadership over planning activities, other tasks remained unaddressed, claimed by no member of the senior management team. Again, leadership surpluses and deficits are present within the same constellation, in this case within the same incident.

Further contributing to the scholarly discussion on leadership configurations, my findings suggest that deficits and surpluses co-exist over time, as surpluses contain the seed for leadership deficits. Gronn (2002) discusses how overlap may lead to mutual reinforcement or redundant efforts, while Chreim (2015, p. 21) observes how ambiguous spaces promote “collaboration, negotiation, contestation, domination or withdrawal” that can lead to abundant or scant leadership being exercised. In this vein, the importance of various forms of involvement by leaders in the design, implementation, negotiation and decision-making processes in contexts of shared leadership has been discussed in previous work (Boak et al., 2015). However, my study contributes some nuance by showing that extensive deliberations within the constellation can lead to cycles of intrusion and exclusion and internal paralysis which may extend to subordinate levels, resulting in leadership voids. Hence, while role clarification endeavors may be seen as a solution (Davis and Lawrence, 1978), my findings reveal their potential paralyzing effect and the leadership voids that occur.

In this vein, my results provide some nuance to the existing literature on the consequences of shared leadership. Edelmann et al. (2023) note that formally appointed leaders in shared leadership arrangements enjoy independence, clear points of contact, accountability, improved team functioning, decisiveness, lower workload, less ambiguity, more aligned expectations and well-being. However, in my case, the constellation being developed at the top of the matrix structure, even if formally appointed, experienced the opposite consequences, from less independence to work overload and ambiguity. Furthermore, Dust and Ziegert (2016) argue that shared leadership contributes to continuity, whereas senior managers in projects in this case had misaligned visions and expectations, causing discontinuity and disruption.

The case also contributes by revealing perceptions by clinical and functional directors of intrusion and exclusion respectively. Functional directors tried to intervene in clinical teams and felt ignored and “forgotten”. Clinical directors experienced those interventions as jurisdictional intrusion and were often unaware of the functional director’s work with members of their directorate. When functional directors felt excluded, they would insist on trying to assert their leadership, feeding the clinical directors’ feelings of intrusion. The latter would then try to gain some control over the activities, or at least be informed, which exacerbated functional directors’ feelings of exclusion. These incidents indicate that if one asserts leadership too strongly as roles are being negotiated, defensive responses may be triggered, weakening one’s own relationship with co-leaders and generating a leadership void. For theory, this supports the concept that leadership surpluses contain the seeds of leadership deficits.

The literature on matrix structures established the inherent role overlaps, conflicts and ambiguity decades ago (Davis and Lawrence, 1978; Egelhoff and Wolf, 2017). The tensions across senior management levels identified in previous work reflect leadership surpluses, but deficits have mostly been ignored. Yet my findings suggest that matrix structures can create vacuums where leadership deficits can expand, many of them as a result of surpluses. I thus contribute to the literature by illustrating how leadership deficits can result from leadership surpluses in matrix structures. Although the leaders in this case did not report being submerged by a “mass of bureaucratic undergrowth” (Heller, 1991: 24), my findings suggest that the complexity of defining roles orients senior leaders inward rather than upward or downward. This, in turn, led to paralysis and outward leadership deficits.

Galbraith (2013) insists on the need for boundary spanning leadership in matrix organizations. On shared leadership, Chen and Zhang (2022) discussed the potential of psychological territorial loss and perception of violation. In my case, such dynamics led to feelings of intrusion and exclusion. To counter the exclusion and paralysis that can result from overly assertive boundary spanning, my findings point to the importance of shared understanding of boundaries and balanced assertiveness in crossing those boundaries. While Burton et al. (2015) explain that the junction points are where matrix structures “make or break” (p. 42), partly due to a lack of information being distributed downward, my results reveal that the executive team responsible for planning or managing operations may potentially be blinded or paralyzed if they lack lateral and upward information flow from the senior management level.

Conclusion

This study makes two incremental contributions to the literature. First, while previous studies assign leadership deficits and surpluses to opposite ends of a spectrum and tend to consider them as mutually exclusive, my results point to the coexistence of leadership surpluses and deficits in constellations. Among other ways in which this coexistence may take place, I noted that leadership surpluses contain the seeds that generate leadership voids. Actors outside the constellation may experience surpluses of internal leadership as deficits when leadership claims within the constellation prevail over enacting leadership outside the shared role space. Role negotiation work can paralyze projects and delegitimize leadership for outsiders. Overlapping responsibility for some aspects of leadership may lead to competitive behavior within the constellation, while other aspects may be left untended in the shared space. Cycles of intrusion and exclusion can permeate frontiers, preventing some leaders from exercising leadership in overlapping zones. In sum, these findings contribute to the scholarly discussion on the sharing of leadership by uncovering the leadership surplus conundrum. Second, this study contributes to the literature on matrix structures by reviving the investigation into the often-forgotten leadership deficits in matrix structures. In such contexts, shared understandings of boundaries and balanced assertiveness in crossing those boundaries help reduce exclusion and paralysis.

While it allows for the development of insight into shared leadership, the unique setting of this study shapes its transferability. Generally, my findings are likely to apply to situations of redesign of organizational structures involving a redefinition of the roles of all or many leaders who are working in some coordinated way, such as the implementation of new structures. Contexts with high ambiguity align with my empirical setting. In the healthcare sector, frequent reform and redesign of organizations create such environments. Contexts where arrangements such as emerging shared leadership and co-leadership are being intensely negotiated are likely to reflect similar interplay of surpluses and deficits. However, my research was not designed to explore these dynamics in mundane, stable, day-to-day contexts where surpluses and deficits may play out differently.

The empirical context allowed for an investigation of the early stages of formation of constellations in which all members at the same hierarchical level negotiate their roles. Yet my findings point to the need for additional efforts to study the coexistence of leadership surpluses and deficits over time. The design of this study focused on one case study, which allowed for depth and richness, yet it somewhat limits the clarity of the transferability of the results. Further research in other settings, perhaps with lower structural complexity, could extend insight into the negotiation of leadership configurations and the interconnections between leadership voids and surpluses.

Since my findings are strongly rooted in the significant overlap of roles inherent to matrix structures, I believe that players in settings where there is significant overlap between roles would find the insights developed here useful. For instance, situations in which an individual previously occupied a leadership role single-handedly and who must now share their leadership role with another person in a co-leadership arrangement may be conducive to a similar interplay of leadership surplus and deficit. In matrix organizations in particular, my findings point to temporal unfolding of leadership deficits and surpluses. Although my empirical context provided a unique opportunity to explore nascent leadership arrangements, the time constraints of this study limited my ability to follow the negotiation process until some form of stability could be reached. Future research taking a process perspective could help to delve deeper into the specifics of leadership deficits in contexts of overly abundant leadership.

In this paper, I show systemic and perceived tensions that members of nascent shared leadership constellations face and identify conditions where leadership surpluses and deficits can coexist. For practice, these findings suggest that during formative stages of leadership constellations, much attention should be paid to identifying zones of overlap and ambiguity and to make each leader’s role and expectations as explicit as possible. In contexts of leadership abundance, actors should explore voids in leadership, specifically incidents of intrusion and exclusion, moments of functional transition and intense role negotiation.

Additional contributions of the study to practice suggest that role negotiation and clarification efforts should be done in private to avoid discrediting leaders, and shared understandings of roles should be jointly explained to subordinates and other outside actors. Leaders should be careful to prevent role tensions and negotiation from paralyzing work. In practice, role negotiation must be timely yet not monopolize attention. In situations of inward/outward tensions, mechanisms and processes for the use of shared resources should be agreed by all relevant leaders. Arguably, communicating with other leaders about the use of shared resources or about initiatives impacting them is key to limiting feelings of intrusion and exclusion. Healthcare organizations include various forms of shared leadership arrangements that have considerable impacts on service delivery. Frequent significant changes and reforms trigger recurrent phases of negotiation within and around these constellations. Insight into these processes and tensions can hopefully ease mutual adjustment and improve the exercise of leadership.

Figures

Figure 1

Previous and new organizational structures

Table 1

Characteristics of participants

Characteristics		Number of participants
Type of Directorate	CEO and Deputy CEOa (E)	2
	Clinical Directorate (C)	5
	Functional Directorate (F)	5
	Administrative Directorate (A)	4
Length of Experience in Management Role	0–9 years	5
	10–19 years	7
	20 years or more	4
Length of Experience at the Current Hierarchical Level	First Experience at Current Level	7
	Previous Experience at Current Level	9
Previous Role: Internal versus External	Internal	10
	External	6

Note(s): ^aWhen citing participants, the author uses the following abbreviations: CEO or deputy CEO: E; clinical directors: C; functional directors: F; administrative directors: A

Source(s): Authors work

Table 2

Data structure

Aggregate dimension	Second-order themes	First-order concepts
Planning/Operations	Role Overlap	There is a real overlap and it is not clear. (E)
		There are too many hands on the steering wheel. (F)
		We didn’t know who was supposed to do what. It was complete confusion. When I hear three people say that they are responsible. Can we define which part each person is responsible for? (F)
	Transitional Ambiguity	Is it part of planning? Maybe it is operational planning. What is planning and what is operations. (F)
		Planning, we have to define where it ends and where operations start. It was difficult to define where planning ends. When recruiting? When implementing? (C)
		Separating planning from operations management, it was trial-and-error. (E)
	Disconnect Between Planning and Operations	When functional directors who have never managed people and operations are giving orders to our staff, it’s difficult. They don’t understand how the organization runs. (C)
		Great administrators and thinkers of healthcare administration plan these things but it becomes obvious afterward that common sense should prevail and they make blunders. (F)
		The functional director works on abstract concepts and likes shovelling clouds. When it is time for concrete stuff, cement and bricks, aligning bricks to form a wall, nothing works. (F)
		I have never been asked by the functional director what could be done for the project to work and to avoid failing. (C)
Intrusion/Exclusion	Competing Accountability	Functional directors are everywhere and have the power to tell me “Study this issue, follow-up on this” so I have a lot of superiors. (C)
		When we support them, it is fine. When I make a request, it is more difficult. It is an additional workload and the loyalty goes to the clinical supervisors. (F)
		For the clinical directors, there is one boss. They are open to advice and us helping them when they need help, but when we make a request, it is very, very difficult. (F)
	Blind Management	For a project, when I asked a question, I have been told that it wasn’t my responsibility. So things happen in my team and I am not informed of it. (C)
		We won’t be consulted by clinical directors. We just want to make sure that we won’t have to pick up the pieces after. (F)
		Directors should think about the impact of what they are doing in other directorships. We have administrators who don’t know what is going on even when it affects them. (C)
	Perceived Interference	Some people play in other people’s sandboxes. They are bullies. (C)
		Some clinical directors feel threatened. They say, “is it interference in my directorate?”(E)
		Some clinical directors need to grieve their past role, and accept that functional directors are now responsible for some of the tasks they used to do. (A)
		When I think of everyone intervening in the clinical director’s hospital – it’s everyone. It can be threatening, or tiring. (C)
Inward/Outward	Lower-Level Paralysis	I can’t unilaterally decide that a member of my team will do something because there are other needs for functional directors. But it’s my staff and something needs to be done! But it’s not getting done. (C)
		Functional roles in a matrix structure are not always understood. If middle managers don’t inform me of incidents, my team can’t act on it. (E)
		No one wants to take on a group of professionals so we just throw the ball at each other. They were suffering, but now they like not having a boss. When they want a boss, they find someone who will give them what they want, and when they don’t want a boss, they just keep quiet and no one disturbs them. (F)
	Stagnation within the Constellation	Sometimes tasks are unaddressed. If there is a problem, I have a responsibility to act. Sometimes clinical directors don’t tell me so problems don’t get addressed. (F)
		We have to reduce to a minimum the number of decisions made by 16 people. It’s too many people and it’s difficult to make a decision. (A)
		Everyone feels accountable. (…) Because of that overlap, we are not getting things done and going as far as we should. (C)
		Recommendations were made. No one followed-up, and no one is following up. I am concerned - it’s in my directorship but another director is responsible. (C)

Source(s): Authors work