Relational Leadership: New Developments in Theory and Practice
In this symposium we explore relational leadership and related concepts, highlighting new developments in theory and their implications for practice. Relational leadership is defined here as a pattern of reciprocal interrelating between workers and managers to make sense of the situation, to determine what is to be done and how to do it (Gittell & Douglass, 2012). Each party learns from the other, with workers contributing the more focused in-depth knowledge associated with their roles while managers contribute the broader less focused knowledge associated with their roles. Together they create a more integrated holistic understanding of the situation. This process of reciprocal interrelating involves communicating through relationships of shared goals, shared knowledge and mutual respect, with mutual respect as an emotional connection that heightens each party’s attentiveness to the needs and insights of the other, triggering cognitive connections in the form of shared goals and shared knowledge.
In the traditional bureaucratic organizational form, by contrast, the worker-manager relationship is defined by norms of hierarchy and power-over rather than power-with (Weber, 1920). At the same time this hierarchy is embedded in roles that provide some protection against outright domination (Weber, 1920). “Hierarchy without domination” means that a realm of autonomy exists within the confines of a worker’s job description, protected by formal rules from outright domination (Weber, 1920). Theories of street-level bureaucracy (Lipsky, 1980) as well as more recent theories of job-crafting (Berg, Grant & Johnson, 2010, Wrzesniewski& Dutton, 2001) suggest that workers do have a realm of autonomy even in traditional bureaucratic organizations, providing them discretion within the confines of their job descriptions and even enabling them to reshape their job descriptions. This realm of autonomy can be used to withhold work effort but can also be used to take actions on behalf of customers or to increase the meaning of the work. Effective use of this autonomy is limited however when workers lack understanding of the whole due to their subordinate position in the bureaucratic hierarchy and their constrained role in the horizontal division of labor.
In the pure relational organizational form, participants exercise influence based on their personal qualities rather than their roles. The upside of the relational form is that participants must earn the commitment or loyalty of other organizational participants. The downside is that the lack of role-based authority means there are no formal limits to the use of that authority, which can degenerate into despotism or nepotism as Weber argued when making his case for the superiority of the bureaucratic form.
Relational leadership differs from the leadership found in the pure bureaucratic form and the pure relational form by being both role-based and reciprocal. Relational leadership builds on Follett’s (1949) concept of reciprocal control, a form of control that is not coercive but rather “a coordinating of all functions, that is, a collective self-control” (1949: 226). Achieving this collective self-control, she argued, requires a form of leadership that is distributed throughout the organization rather than concentrated in a few positions. Follett observed organizations in which “we find responsibility for management shot all through a business [and] some degree of authority all along the line [such that] leadership can be exercised by many people besides the top executive” (1949: 183). Rather than vesting authority in one person over another based on his or her position in the hierarchy, authority is shared (Fletcher, 1999). The core characteristic of relational leadership is the embedding of authority into each role, based on the knowledge associated with it.
Distributed leadership, carried out by both formal and informal leaders throughout the organization to facilitate achievement of organizational objectives (Ancona & Bresman, 2007), has several characteristics in common with to relational leadership. Distributed leadership is a form of influence that can be exercised by participants at any level of an organization, and moreover, leaders are most effective when they can inspire others to engage in the responsibilities of leadership rather than attempting to carry out all leadership responsibilities on their own. Distributed leadership thus would appear to require facilitative leadership behaviors rather than directive leadership behaviors, and transformative leadership behaviors rather than transactional or passive leadership behaviors. Lending support to this perspective, Carson and co-authors (2007) found that supportive supervisory behaviors predict greater frontline worker engagement in shared leadership.
However, relational leadership does more than draw upon expertise and leadership from participants throughout the organization. It is a process of reciprocal interrelating through which the expertise held by different participants interpenetrates, creating a more holistic perspective that is integrative rather than additive. Relational leadership requires facilitating the interpenetration of expertise among others, which in turn requires the skills to build relationships among others, creating a safe space in which they can reciprocally interrelate with each other. According to Lipman-Blumen (1992: 184), facilitating connections among others is a key attribute of connective leadership:
Connective leadership derives its label from its character of connecting individuals not only to their own tasks and ego drives, but also to those of the group and community that depend upon the accomplishment of mutual goals. It is leadership that connects individual to others and to others’ goals, using a broad spectrum of behavioral strategies. It is leadership that ‘proceeds from a premise of connection’ (Gilligan, 1982) and a recognition of networks of relationships that bind society in a web of mutual responsibilities.
Fletcher’s concept of ”fluid expertise” (1999: 64) in the worker-manager relationship reflects a co-creation process consistent with relational leadership:
[P]ower and/or expertise shifts from one party to the other, not only over time but in the course of one interaction. This requires two skills. One is a skill in empowering others; an ability to share - in some instances even customizing - one's own reality, skill, knowledge, etc. in ways that made it accessible to others. The other is skill in being empowered: an ability and willingness to step away from the expert role in order to learn from or be influenced by the other.
Fluid expertise requires mutual respect, as well as the ability to be caring, responsive and closely attuned to another through the development of both cognitive and emotional connections. One characteristic of relational leadership is leading through humble inquiry, described by Schein (2009) as a form of giving, seeking and receiving help that leaders can use to establish a culture of reciprocal learning throughout an organization.
Relational leadership (worker-manager), along with relational coordination (worker-worker) and relational coproduction (worker-customer), are three processes of reciprocal interrelating that form the core of relational bureaucracy. Relational bureaucracy is a hybrid of the relational and bureaucratic forms in which reciprocal interrelating enables participants torespond to each other in knowledgeable and caring ways, while formal structures embed reciprocal interrelating into roles, thus enabling the scalability and sustainability typically associated with the bureaucratic form (Gittell & Douglass, 2012).
Structure of symposium
This symposium explores relational forms of leadership, with participants from multiple perspectives seeking to articulate the theories behind these forms of leadership as well as their implications for practice. We start with a mapping of the territory by Foldy and Ospina, arguing that relational leadership is one of several forms of collective leadership. Ancona, Backman and Parrott follow with an updated look at distributed leadership and its characteristics – distributed, decentralized and decoupled from roles, outlining similarities with and differences from the concept of relational leadership.
Gauthier and Hornstrup each explore relational leadership as a process of sensemaking in the face of complexity, recognizing both cognitive and emotional dimensions of this process. We conclude with a study by Stephens that explores how leaders foster coordination among others through meaning making, a process that involves embodying the whole for diverse participants.
As discussant, Fletcher will launch the symposium with buzz groups, asking audience members to discuss with each other what they hope to learn. She will break at two points during the symposium to allow additional buzz groups among audience members, then will present her overarching commentary at the conclusion, followed by audience discussion.
The proposed symposium explores the micro-processes of relational leadership, thus creating a potential fit with the Organizational Behavior Division. We explore the implications of these micro-processes of leadership for the organizational form itself, and the organization’s ability to achieve critical performance outcomes, thus creating a potential fit with the Organization and Management Theory Division. We explore relational leadership as a means for transforming mechanistic organizations to become more responsive to complexity and uncertainty, creating a potential fit with the Organizational Development and Change Division.
Relational leadership as collective leadership: Mapping the territory Erica Gabrielle Foldy & Sonia Ospina
Wagner School of Public Service, New York University As criticisms of traditional leadership theory and research amplify and diversify, a variety of new terms challenge the notion of leadership as a one-directional relationship between leader and follower. Scholars have referred to leadership as “shared”, “distributed,” “constructed,” “post-heroic” and “relational” among other terms (Pearce & Conger, 2003; Gronn, 2002; Hosking, 2003; Drath, 2001; Ospina & Sorenson, 2006; Fletcher, 2004; Uhl-Bien, 2006). While they all rest on a basic assumption that leadership does not automatically reside in a single, often heroic, individual, these conceptualizations of shared leadership vary widely. In this presentation, we provide a brief map of the territory -- a framework that suggests the basic dimensions that can differentiate these approaches. We then suggest “collective leadership” as an umbrella term that encompasses these conceptualizations and position relational leadership within this framework.
The move away from the single, heroic, leader is not new. Several decades of scholarship have explored how leadership is practiced – implicitly or explicitly -- as a joint endeavor (Hollander, 1964; Burns, 1976; Rust, 1991). However, the last decade has seen a burst of scholarship investigating this phenomenon, along with a proliferation of terminology to describe it. While many of the terms may appear similar or even interchangeable, in fact they differ significantly in what they describe. Having reviewed the relevant literature in management and organization studies, psychology and education, we suggest two basic dimensions along which the different approaches can be plotted: the “locus of leadership” and the “view of self” (see Table 1).
The locus of leadership is where leadership resides; it is the source of leadership or its “epicenter” (Hiller, Day and Vance, 2006); it is where, as researchers, we look for leadership. There are three loci: the individual, the relationship and the system. The traditional and still dominant perspective is the individual as the locus of leadership: leadership is enacted by individuals who have the appropriate traits, characteristics or styles and engage in measurable leaderly behaviors (Antonakis et al, 2003). Other work understands leadership as based in the relationship between leaders and followers: “Leadership is a concept of relationship; it assumes the existence of some people who follow one or more others… There can be no leadership if there is just one person” (Pearce, Conger and Locke, 2007, 287). A third approach is to see leadership as belonging to the collective (Drath et al, 2008) or residing in a system or context – social, organizational, even a group or team. Spillane et al, scholars of education, suggest leadership should be conceptualized as “a distributed practice, stretched over the social and situational contexts of the school” (their italics; 2004; 5). Smircich and Morgan see leadership as “enact[ing] a system of shared meaning that provides the basis for organizational action” (1982: 258).
Views of self are rooted in the researcher’s ontological and epistemological assumptions about the very nature of human beings, with consequent understandings of “the self” as individuated and autonomous or connected and co-constructed. When applied to leadership, these assumptions paint different pictures of how the relationships undergirding leadership actually work. Positivist and post-positivist approaches understand the self as a distinct entity, clearly bounded, which then engages with other, similarly autonomous beings (Uhl-Bien, 2006; Ospina&Uhl-Bien, forthcoming). In this “entity” approach, the leader and leadership are confounded (Hosking, 1988), with leadership defined as an influence relationship between two social actors -- leader and follower -- who exist as such, prior to the relationship. Leadership is explained by the neo-charismatic school, for example, as a process by which leaders “affect followers as a result of motivational mechanisms that are induced by the leaders’ behaviors” (Antonakis, 2011: 270), including “visionary behavior, positive self-presentation, empowering behaviors, calculated risk taking and self-sacrificial behavior, intellectual stimulation, supportive leader behavior and adaptive behavior” (271).
In contrast, a constructionist perspective sees the self as self-in-connection, created through interaction, with no inherent core or status independent of that which is forged through that interrelationship (Dachler& Hosking, 1995; Ospina &Uhl-Bien, forthcoming). In the constructionist approach, leadership (and those defined as leaders or followers) emerges in process as co-constructions that help advance organizing tasks (Hosking, 1988). Leadership happens in context, it does not exist prior to the relationship: "leaders must constantly enact their relationship with their followers;" they "must repeatedly perform leadership in communication and through discourse" (Fairhurst, 2007: 5). In this approach leadership is understood as relational in that it emerges only in the context of “a particular form of interaction happening at a certain time and place” (Drath, 2001: 16). In this sense, leadership is not something that the leader, as one person, possesses, as much as it is something achieved in community and owned by the group (Ospina& Sorenson, 2006; Foldy et al, 2008).
Plotting each of the two dimensions on a separate axis creates six cells which each represent a different conceptualization of collective leadership, corresponding to different degrees or types of collectivity. (In Table 1, we have suggested specific approaches and scholars whose work illustrates each cell.) We very deliberately choose the term “collective” because it can encompass all of the quite varied forms in the framework. Terms like “distributed” or “joint” leadership suggest that leadership resides in autonomous individuals who then share particular leadership tasks. Words like “processual” (Hosking, 1988) or “discursive” (Fairhurst, 2007) imply a more disembodied approach, one that investigates the process or work of leadership rather than the behaviors of individual leaders and followers. The term collective is elastic enough to provide a broad umbrella, as suggested by this definition: “involving all members of a group as distinct from its individuals”1.
The place of relational leadership in the framework varies because people have used the term in different ways. For example, the definition posed for this panel is relational leadership as “a process of role-based reciprocal interrelating” between workers and managers to negotiate the work that is to be done. In contrast, Uhl-Bien (2006) defines relational leadership as “a social influence process through which emergent coordination (i.e., evolving social order) and change (e.g., new values, attitudes, approaches, behaviors, and ideologies) are constructed and produced.” (2006: 655) The first definition implies that leadership inheres in independent individuals who inter-relate across different hierarchical positions. The second locates leadership in a jointly constructed but disembodied process, not in individuals. Uhl-Bien (2006) proposes Relational Leadership Theory as an approach that can encompass both individuated and connected perspectives by explaining both the emergence of leadership relationships (drawing on traditional individuated views that focus on the nature of the relationship, such as Leader-Member Exchange), and the relational dynamics of organizing (including various constructionist views of leadership). In fact, the term “relational” has been used to refer to quite distinct understandings of leadership, each with different ontological and epistemological assumptions that result in quite distinct approaches to conducting research (Uhl-Bien &Ospina, forthcoming). This suggests the timing is right for a symposium exploring relational leadership.
Table 1: A map of collective approaches to leadership
Constructed Leadership - Drath (2001); Ospina & Sorenson (2006); Foldy et al (2008)
Discursive Leadership -Fairhurst (2007)
Processual Leadership -Hosking (1988)
Complexity Leadership Theory - Uhl-Bien, Marion &McKelvey (2007)
D-leadership and relational leadership: Beginning the conversation
Deborah Ancona, Elaine Backman & Kate Parrot
MIT Sloan School of Management The purpose of this symposium is to explore relational leadership and related concepts. In this presentation, we introduce the concept of “D-leadership” which we developed based upon intensive fieldwork in organizations operating in dynamic, highly competitive industries. D-leadership refers to leadership that is decentralized, distributed and collective, and de-coupled from organizational roles. In the presentation, we provide a brief description of our research questions and methods; present the D-leadership model; highlight our main findings; and identify three ways that the D-leadership model differs from the relational leadership model developed by the symposium organizers.
There are two largely separate but parallel literatures documenting the fact that both organizations and leadership practices have changed dramatically over the past three decades. On the one hand, macro-organizational scholars have provided evidence that organizations have become flatter; more reliant on the use of teams; less formalized with fewer work rules and less detailed job descriptions; and characterized by more porous boundaries (DiMaggio 2003). On the other hand, micro-organizational scholars have found that in many sectors, there has been a move away from “command and control” leadership in which leadership is exercised individually by those in formal positions of authority in a clearly defined hierarchy, toward “shared leadership” (e.g. Carson et al, 2002), “distributed leadership” (e.g. Gronn 2002) or “complexity leadership” (e.g. Uhl-Bien et al 2007) in which leadership is exercised by multiple leaders throughout the organization -- some in formal positions of authority and some not --working collaboratively across organizational levels and boundaries. This research brings these two literatures together by: 1) providing detailed, empirical descriptions of how collaborative leadership is being practiced in organizations considered exemplars in this new leadership form, and 2) identifying what organizational structures, practices and cultures support it. We employed a theory-building, comparative case study methodology (Eisenhart & Graebner 2007) using the following data sources:
A study of leadership practices at two R&D labs at a Fortune-500 business equipment and services company, one with a long history of collaborative leadership practices, the other which operated in a “command and control” manner. We studied two product development projects in each lab relying on extensive interview data augmented by archival data, observation of on-site meetings, and feedback sessions with study participants.
A study of leadership practices in a mid-sized, privately held company that develops and manufactures high-tech products in consumer and business markets considered an exemplar of collaborative leadership. We studied two product development teams, two process change efforts, and two strategic change initiatives, relying on extensive interview data augmented by archival data, observational data from site visits in the U.S, China and Germany, and feedback sessions with study participants.
Secondary data from five other companies known as exemplars of collaborative leadership to fill in gaps and add external validity to our case study findings.