Abstract
The dynamic equilibrium model of organizing has become an influential theoretical framework in paradox research. The model describes paradox management as tightrope-walking, as actors cope with paradoxical tensions through continuous microshifts. The underlying assumption is that once actors accept the paradox and support opposing poles in a consistently inconsistent manner, they can effectively manage organizational tensions. We argue that paradoxical tensions cannot be subsumed under managerial control in this way due to the emergent and unpredictable nature of paradoxes in organizations. Hence, in addition to needing to balance on a tightrope, the tightrope walker may find the rope suddenly pulled in an unexpected direction by a strong gust of wind. To advance theory development, we put forward the concept of dissipative equilibrium to better capture the temporary nature of balance and the continuous vigilance and interventions needed from management to (re)balance an organization in the presence of unexpected developments that are beyond management’s control.
Introduction
With environments becoming increasingly complex and dynamic, there is a growing recognition that organizations must find ways to simultaneously satisfy different competing demands (e.g., Miron-Spektor, Ingram, Keller, Smith, & Lewis, 2018; Schad, Lewis, Raisch, & Smith, 2016; Schad, Lewis, & Smith, 2018). Paradox theory has contributed to our understanding of this challenge (e.g., Fairhurst et al., 2016; Putnam, Fairhurst, & Banghart, 2016; Schad et al., 2016; Schad et al., 2018). Building on the dynamic equilibrium model (Smith & Lewis, 2011), paradox theory describes maintaining equilibrium as dynamic balancing. The image of a tightrope walker aptly captures this idea in that one must respond to the lateral forces that threaten balance by engaging in “consistent, ongoing microshifts” (Schad et al., 2016, p. 37). Similarly, coping with paradoxical tensions requires “constant motion across opposing forces” to maintain balance amid “a continuous pull in opposing directions” (Smith & Lewis, 2011, p. 386).
A central claim of current paradox theory is that, once actors accept a paradox and support opposing sides in a “consistently inconsistent manner,” they can manage organizational tensions effectively (e.g., Lewis & Smith, 2014; Miron-Spektor et al., 2018; Smith & Lewis, 2011). Paradox research acknowledges that individual and organizational forces, such as inertial pressures, can bias actors’ behavior and lead to escalating efforts to support one side of a tension over the other. At the same time, however, the literature suggests that actors’ acceptance of and comfort with tensions allows them to overcome such “vicious” forces. We build on recent criticism that has cast doubt on the assumption that such a state of sustainable balance is attainable (Cunha & Clegg, 2018; Cunha & Putnam, 2019; Putnam et al., 2016). The dynamic equilibrium model neglects the emergent nature of paradox “revealed in surprising shifts and turns” (Cunha & Clegg, 2018) that can render tensions uncontrollable (Cunha & Putnam, 2019). This unpredictability of paradoxes can cause organizational tensions to resurface in novel and unexpected ways.
We focus in this paper on the uncontrollable aspect of paradoxes and argue that despite management’s acceptance of and commitment to proactively balancing opposing forces, managers are limited in their ability to do so. Even when managers are able to overcome their personal biases, different external and internal developments can cause an organization to fall out of balance. These developments can present themselves as external “chance events” (e.g., MacKay & Chia, 2013) or as “unanticipated consequences” (e.g., Merton, 1936) of everyday decisions, actions, and interactions of actors occurring at multiple levels of an organization. We identify different types of internal drift—individual, interactional, and organizational—that can be difficult for boundedly rational managers (e.g., Puranam, Stieglitz, Osman, & Pillutla, 2015; Simon, 1957) to preemptively account for. Contrary to the dynamic equilibrium model, which assigns a central role to managerial agency and preemptive balancing actions, we argue that sustaining paradox requires continuous investment in the recognition of emerging threats and the implementation of corrective actions to (re)balance the organization.
We put forward the concept of dissipative equilibrium as a novel metaphor for paradox management to better capture the temporary nature of the balancing of competing tensions. Originating from Prigogine’s research on temporary structures in thermodynamics (Prigogine & Lefever, 1973; Prigogine & Stengers, 1984), the concept of dissipative equilibrium has been used in management research to denote an unstable equilibrium that requires continuous management attention (Brown & Eisenhardt, 1997, p. 29; Davis, Eisenhardt, & Bingham, 2009). Inspired by complexity theory, Brown and Eisenhardt (1997, 1998) referred to the dissipative equilibrium state as “the edge of chaos” from which an organization can at any point collapse into different extremes when continuous attention and energy are not expended in managing an unstable balance (see also Eisenhardt & Piezunka, 2011).
The dissipative equilibrium model of paradox contributes to the development of a more realistic understanding of the practical challenges and longer-term outcomes of paradox management. Taking a management point of view, we elaborate on how managers can deal with paradoxes while recognizing the challenges that are associated with the unexpected nature of paradoxes and the limits of managerial agency. In contrast to the dynamic equilibrium model that focuses on external environmental factors that render tensions salient and internal factors that spur vicious or virtuous cycles, we investigate internal challenges to the ongoing management of paradox and the limits of managerial agency to preemptively address them.
We will next briefly review the research on the dynamic equilibrium model and its key assumptions, outline the main criticisms, and introduce the dissipative equilibrium model as a means to address some of these criticisms and move research forward.
Theoretical Background
Paradox theory has substantially enhanced our understanding of the nature, management, and outcomes of paradoxical tensions in organizations (Fairhurst et al., 2016; Schad et al., 2016). Consistent with prior work, we define paradoxes as “contradictory yet interrelated elements that exist simultaneously and persist over time” (Schad et al., 2016, p. 10) and organizational paradoxes as organizational demands that are “contradictory but must be pursued interdependently to sustain long-term organizational performance” (Knight & Paroutis, 2017, p. 403). We consider paradoxes both inherent and socially constructed (Hahn & Knight, 2021; Smith & Lewis, 2011). Paradoxes are inherent to organizations due to the “real” organizational tensions that exist prior to and beyond the will or power of organizational actors (e.g., Schad & Bansal, 2018). However, we also recognize that social interactions play an important role in how paradoxes become salient and are enacted (e.g., Hahn & Knight, 2021; Luscher & Lewis, 2008; Putnam et al., 2016).
The dynamic equilibrium model: The foundation of extant paradox theory
While scholarly recognition of organizational paradoxes dates back several decades (e.g., Cameron & Quinn, 1988; Lewis, 2000; Poole & Van de Ven, 1989), the dynamic equilibrium model of organizing introduced by Smith and Lewis (2011) has over the past ten years become a foundational building block of current theory (Cunha & Putnam, 2019). The metaphor of dynamic equilibrium emphasizes a core assumption of the model—the idea that management can dynamically balance the different sides of a paradox “by adapting to a continuous pull in opposing directions” (Smith & Lewis, 2011, p. 386). Since its publication, the model has inspired a number of conceptual and empirical studies that have built on and deepened its insights (e.g., Knight & Paroutis, 2017; Raisch, Hargrave, & van de Ven, 2018; Smith, 2014; Smith & Besharov, 2019).
The dynamic equilibrium model suggests that depending on actors’ acceptance of paradox, the dynamics of paradox play out as either vicious or virtuous cycles. Building on psychological and cognitive research, the model holds that actors tend to prefer consistency and react defensively to contradictory demands. Actors may seek to resolve their discomfort by supporting only one side of a paradox. Such behavior can, however, spur vicious cycles of escalating efforts toward one extreme or another (Lewis, 2000). Hence, the model argues that management’s ability to overcome its own bias and support both poles of a paradox are key to long-term success and provide a vital foundation for virtuous cycles (Lewis & Smith, 2014; Smith & Lewis, 2011).
Once a paradox has been accepted, paradoxical resolution requires management to constantly shift between alternative strategic agendas to support competing demands, in equal amounts. The model describes paradoxical resolution as unfolding over time through continuous iterations between complementary practices of differentiating and integrating (i.e., “iterating between splitting and integration,” Smith & Lewis, 2011, p. 394). Differentiating can be defined as the separation of the competing poles of a paradox to provide dedicated attention to them, and integrating as the search for synergies and linkages between these poles (e.g., Lewis & Smith, 2014; Smith, 2014). Accordingly, differentiating enables management to alternate its support between the competing poles and integrating allows it to periodically bring them together. Hence, both differentiating and integrating are needed to sustain paradoxes (e.g., Smith, 2014).
The conceptual insight of the importance of supporting competing poles equally is rooted in prior research on the management of competing tensions. For example, Smith and Tushman (2005) argued that cognitively shifting “back and forth” (p. 532) between exploration and exploitation enables top management teams to sustain commitments to different strategic domains. Such cognitive differentiation allows top management to recognize the needs of each domain, whereas cognitive integration makes it possible to recognize the synergies between domains.
Relatedly, Andriopoulos and Lewis (2009) found that managers were able to manage tensions by alternating between differentiating and integrating tactics within and across projects. Higher-level managers created the context by building a diversified portfolio while cultivating a paradoxical vision that stressed the need for synergies. Within the context of this broader vision, lower-level managers relied on temporally and spatially separating their work practices while purposefully improvising to find synergies between extant routines and emerging opportunities.
Smith (2014) observed how differentiating practices such as allocating domain-specific roles or comparing domains to raise novel distinctions allowed top management teams to value the needs of distinct strategic agendas. Integrating practices, such as allocating integrative roles or emphasizing overarching goals, helped leaders recognize synergies between agendas. These practices enabled decision makers to fluidly shift support between strategic domains in response to key issues and to sustain their commitments to multiple strategic goals simultaneously.
In addition to outlining effective management strategies, the dynamic equilibrium model suggests that organizations can learn to manage tensions better over time “in an upward spiral of peak performance and sustainability” (Lewis & Smith, 2014, p. 137). Although the concept of learning spirals has inspired further conceptual work (e.g., Raisch et al., 2018), there is limited empirical research on how organizations manage paradoxes over longer periods of time (Lewis, 2018; Tsoukas & Cunha, 2017). One recent exception is Smith and Besharov’s (2019) study of a social enterprise, in which leaders shifted between competing agendas to achieve both social and business missions. They observed that over time managers learned to manage tensions between their dual missions more effectively and efficiently; thus, their enactment processes became shorter and swings toward the different poles of the paradox became less extreme (Smith & Besharov, 2019, p. 26).
Limitations of the dynamic equilibrium model
While the dynamic equilibrium model has contributed significantly to our understanding of paradox management, it builds on a rather strong assumption regarding managerial agency concerning the ability of management to resolve and preemptively sustain paradoxes. The model assumes that paradoxes can be subsumed under management’s control once managers are able to overcome their own biases. Hence, the dynamic equilibrium model tends to treat “paradoxes as ‘things’ to be solved or controlled” (Cunha & Putnam, 2019, p. 100), ignoring the inherent complexity and instability associated with them (Cunha & Clegg, 2018). The organizational reality is, however, “chaotic, complex, fluid, sometimes random, frequently messy, and often surprising” (MacKay & Chia, 2013, p. 225). Even when senior management is able to overcome its own biases and support both poles of the paradox in a consistently inconsistent manner, it is impossible for them to account for the biases of all other organizational members and to anticipate the consequences of their everyday decisions, actions, and interactions (Merton, 1936).
Hence, while the dynamic equilibrium model represents the “concept of order,” the continuously evolving reality may be better described through the “concept of disorder” (Cunha & Putnam, 2019; Putnam et al., 2016). Instead of focusing on ways to reach equilibrium, we need theories on how to deal with disequilibrium states (Cunha & Putnam, 2019). Although Smith and Lewis (2011) also note that when managers engage in paradox resolution, “the threat of vicious cycles persists, requiring managers to remain vigilant” (2011, p. 393), they do not develop the idea further. Instead, they conclude that “managing paradoxical tensions via dynamic, purposeful, and ongoing strategies of acceptance and resolution fosters sustainability” (2011, p. 394). We argue that sustaining such control over paradoxes is more difficult to achieve than the dynamic equilibrium model would seem to suggest.
We will next outline how moving from the dynamic equilibrium metaphor to the dissipative equilibrium metaphor can help advance our understanding of paradox management by enabling us to better account for the disorderly and unpredictable nature of paradoxical tensions and the difficulties associated with sustainable paradox resolution.
Dissipative equilibrium metaphor
The theory of dissipative equilibrium was originally introduced by Ilya Prigogine, a Russian-Belgian Nobelist in chemistry. Prigogine found that although lack of order and unpredictability always prevails in the neighborhood of thermodynamic equilibrium, dissipating energy to a system can result in the emergence of new, out-of-equilibrium structures (Prigogine & Lefever, 1973). While such dissipative structures are temporary and often highly unstable by their nature (e.g., a tornado), they can be sustained when energy is continuously expended to them.
Prigogine’s discovery provided the basis for the study of the dynamics of open, self-organizing systems (Prigogine & Stengers, 1984). His ideas have since been widely applied outside thermodynamics and increasingly also in organization research to explain how organizations transform and how new structures emerge (Gemmill & Smith, 1985; Macintosh & Maclean, 1999). For example, Macintosh and Maclean (1999) used the concept of dissipative structures to explain the process of organizational transformation from one structural archetype to another.
Brown and Eisenhardt (1997) applied the idea of dissipative equilibrium to structures of continuously transforming high-velocity organizations that must manage the balance between too much and too little structure (Brown & Eisenhardt, 1998). Building on the metaphor, the authors put forward the concept of “semistructures.” Accordingly, a company in a semistructured state is in a dissipative equilibrium that requires continuous vigilance and attention from management to prevent the company from “slipping into pure chaos or pure structure” (Brown & Eisenhardt, 1997, p. 29; Davis et al., 2009). While Eisenhardt and her colleagues did not use paradox theory in their theorization, the concept of semistructures is inherently paradoxical due to the need to cope with the tension between too much and too little structure.
The concept of dissipative equilibrium is also highlighted in Anderson’s (1999) portrayal of complex adaptive systems, in which the task of management is not to have a superior ability to “foresee the future or to implement enterprise-wide adaptation programs, because nonlinear systems react to direction in ways that are difficult to predict or control.” Instead, the role of management is to “set constraints upon local actions, observe outcomes, and tune the system by altering the constraints, all the while raising or lowering the amount of energy injected into the dissipative structure they are managing” (Anderson, 1999, p. 222).
Also, Burgelman’s research on organizational evolution (e.g., Burgelman, 1991, 1994, 1996) has been influenced by Prigogine’s insights on the unpredictability of complex adaptive systems (Burgelman, 2010; Prigogine, 1980). Building on Prigogine, Burgelman (2010) noted that organizational evolution should be seen similarly, as a dynamic, nonlinear process of becoming that progresses through “mutations” and “punctuations” (e.g., Gersick, 1991; Romanelli & Tushman, 1994; Tushman & Romanelli, 1985), rather than as a predictable, linearly progressing organizational evolution (Burgelman, 2010; Burgelman, McKinney, & Meza, 2017).
Although the dynamic equilibrium model was similarly inspired by thermodynamics and biology (Smith & Lewis, 2011), it builds on somewhat different assumptions regarding system dynamics. The model assumes that a system can be kept under control by adjusting the magnitude of opposite “forward” and “backward” reactions. Accordingly, when opposing reactions in the system are equal in magnitude, the effects cancel each other out and enable stable uniform conditions to prevail. For example, when there are equally many chemical reactions occurring in both directions, a chemical system is in dynamic equilibrium. Following a similar logic of reasoning, the dynamic equilibrium model of organizing suggests that as long as management provides consistent support to the opposite poles of a paradox, the effects cancel each other out and the organization is in dynamic equilibrium (Smith & Lewis, 2011).
Hence, an important assumption underlying the dynamic equilibrium model is that the influence of the opposing poles is equal and that the pull in the opposing directions is continuous (Fairhurst et al., 2016; Schad et al., 2018). In contrast, in the dissipative equilibrium model, the system is assumed to be inherently unpredictable and able to change nonlinearly around the equilibrium state (Burgelman, 2010; Prigogine, 1980; Prigogine & Stengers, 1984). Therefore, supporting both poles of the paradox continuously in an equal manner will not yet suffice. Destabilizing forces can appear unexpectedly from either within or outside the system and asymmetrically from either of the different poles, requiring corrective rebalancing actions.
To further concretize the conceptual difference between the dynamic and dissipative equilibrium metaphors, Figure 1 shows how different types of equilibria relate to each other according to their continuous (linear) vs. discontinuous (nonlinear) and static vs. dynamic nature. These two dimensions enable us to distinguish between static equilibria, dynamic equilibria, punctuated equilibria, and dissipative equilibria. Accordingly, in static equilibrium, the system is in a stable state of balance with no movements. In dynamic equilibrium, opposing movements cancel one another out and create an orderly stable state of dynamic balance (e.g., hot and cold air that enters a room at the same rate keeps the room temperature stable). In punctuated equilibrium, periods of equilibrium are punctuated by sudden radical changes followed by subsequent periods of stability. Hence, a system moves from one stable state to another through occasional punctuations that disrupt the stable state. Finally, in dissipative equilibrium, punctuations and subsequent corrective actions occur continuously and in an unpredictable manner. Keeping the system from falling out of the equilibrium state requires continuous dissipation of energy into the system in response to the observable system dynamics.
Typology of Different Types of Equilibria.
A Dissipative Equilibrium Model of Paradox
Building on the understanding of the different types of equilibria, we put forward the dissipative equilibrium model of paradox. Similar to the dynamic equilibrium model of Smith and Lewis (2011), the dissipative equilibrium model focuses on the management of organizational paradoxes. These include paradoxes of learning (e.g., exploration vs. exploitation), performing (e.g., financial vs. social goals), organizing (e.g., central vs. decentral structure), and belonging (e.g., individual vs. collective) (Smith & Lewis, 2011). While we believe that the dissipative equilibrium model applies to these different types of organizational paradoxes, we exemplify its functioning through the organizing paradox since choices regarding organization design also tend to influence how learning, performing, and belonging paradoxes occur in organizations.
According to the model, shown in Figure 2, once management has accepted the paradox and is managing it by alternating its support between the competing poles of the paradox, it can reach a dynamic equilibrium. While such preemptive balancing may sometimes suffice to sustain the paradox, as illustrated by the arrow in the center of the figure, different types of unanticipated drift can disturb the balance. Figure 2 exemplifies this shift away from dynamic equilibrium to the opposing poles of the paradox with the two boxes in the upper and lower parts of the figure. Therefore, in addition to preemptively alternating its support to the different poles of the paradox, management must also invest continuous effort to recognize the different threats to balance and to develop appropriate corrective rebalancing practices. Depending on the pole to which the organization is drifting, the rebalancing practices must be matched accordingly to counter the force pulling the organization out of balance.
Dissipative Equilibrium Model of Organizing.
While management may be able to anticipate some of the different types of internal drift, the differences between the dynamic equilibrium model and the dissipative equilibrium model stem not only from management’s ability to anticipate the occurrence of different types of drift but also from the ability to preemptively account for them. The dynamic equilibrium model assumes that continuously supporting both poles of the paradox, preemptively and in an equal manner, suffices for managing the paradox (e.g., Schad et al., 2018). In contrast, in the dissipative equilibrium model, even preemptive actions should be seen as corrective because they are aimed at supporting one pole of the paradox asymmetrically to prevent an anticipated drift toward the other pole.
In the next sections, we elaborate on the different internal drifting tendencies and corrective rebalancing practices through which management can reestablish a balance between competing tensions. While external chance events can also cause an organization to fall out of balance (e.g., Jarzabkowski, Bednarek, Chalkias, & Cacciatori, 2021), we focus here on different types of internal drift driven by unanticipated organizational dynamics and which can be equally difficult for management to anticipate and control as external “black swan” events (MacKay & Chia, 2013). This approach differs from Smith and Lewis (2011), who theorized on the determinants of paradox salience and the acceptance of a paradox but not the effects of internal organizational dynamics on the ongoing management of competing poles.
In line with the understanding that tensions are both inherent (Schad & Bansal, 2018) and socially constructed (Putnam et al., 2016), we consider the different types of internal drift inherent in system-level changes that are driven by common everyday decisions, actions, and interactions occurring in an organization. Such changes can easily go unnoticed, as they take place on multiple levels of hierarchy and can foster a slow, albeit growing, drift until it becomes evident that the paradox is being pulled out of balance.
Internal drifting tendencies
Following the tradition of multilevel theorizing in organization research (e.g., Klein & Kozlowski, 2000; Nightingale & Toulouse, 1977), we identify internal drifting tendencies at three levels: individual, interactional, and organizational.
First, individual tendencies to drift denote the tendency of individuals throughout an organization to unintentionally privilege one pole of a paradox over another through their different decisions and actions. This type of drift is driven by general human dispositions, such as preferences for certainty over uncertainty or for short-term gains over longer-term gains (e.g., Kahneman & Tversky, 1979; Levinthal & March, 1993), and poses a continuous latent threat to balance. In the context of organizing paradoxes, uncertainty avoidance, for example has been found to relate to the preference for a higher degree of centralization and control at the expense of decentralization and autonomy (Levinthal & March, 1993; Smith & Tushman, 2005).
While Smith and Lewis (2011) also discuss individual-level biases such as the “cognitive and behavioral drive for consistency,” “emotional anxiety,” and “defensiveness” and their effects on senior managers’ acceptance of the paradox, we highlight how the biases of all other organization members can lead to an internal drift toward either pole of the paradox. We argue that even when senior managers themselves are able to overcome their own biases, the individual-level tendencies of all other actors at different levels and in different parts of the organization are difficult for senior managers to continuously account for. For example, in a successful organization, the fear of failing to extend a past success can manifest itself in individual managers’ decisions and actions without senior management noticing, causing an organization to drift over time toward an increasingly exploitative internal focus, at the expense of a more exploratory external focus (Vuori & Huy, 2016).
Second, we identify interactional tendencies to drift, which represent the unanticipated outcomes of the interactions of different organizational actors and entities (MacKay & Chia, 2013). Such outcomes can lead to self-reinforcing dynamics (Es-Sajjade, Pandza, & Volberda, 2021) that cause an organization to drift out of balance. Interactional tendencies to drift are often associated with coordination or collaboration deficits that can unleash unexpected dynamics in an organization across interdependent actors and entities (Kretschmer & Puranam, 2008). Merton describes this phenomenon as the “interplay of forces and circumstances are so complex and numerous that prediction of them is quite beyond our research” (Merton, 1936, p. 900). Foss’s (2003) study of Oticon’s “spaghetti organization” provides a good example of how coordination deficits can lead to an uncontrolled autonomy and the need for corrective rebalancing actions.
Third, organizational tendencies to drift can cause an organization to gradually drift out of balance as it grows and matures (Eisenhardt, Furr, & Bingham, 2010). Growth can entail changes that inadvertently skew organizational members’ behaviors toward one side of tensions or the other (e.g., Es-Sajjade et al., 2021). On the one hand, growth may lead to dynamic contingency misfits, that is, an organization design that fits a company’s early strategic needs may not be a good fit anymore at a later stage (e.g., Westerman, McFarlan, & Iansiti, 2006; Zajac, Kraatz, & Bresser, 2000). A misfit between organization structure and coordination needs can lead to a drift toward excessive autonomy and diminished efficiency. On the other hand, as organizations grow and mature, they tend to develop structural inertia (Eisenhardt et al., 2010; Schwarz, 2012; Tushman & Romanelli, 1985), which can lead to a drift toward excessive control and diminished adaptability. While Smith and Lewis (2011) noted that organizational inertia can spur a vicious cycle by slowing organizational responses to a paradox, we argue that increasing structural inertia can also influence the resurfacing of the paradox.
Corrective rebalancing practices
To counter the different internal drifting tendencies that are difficult for senior management to preemptively account for, management can deploy cognitive, processual, and structural differentiating and integrating practices to correctively rebalance the organization. Prior literature has argued that alternating between differentiating and integrating can enable managers to preemptively maintain a balance between opposing forces in an organization (e.g., Smith, 2014; Smith & Lewis, 2011). We argue that such differentiating and integrating practices can also play an important role as the means to correctively rebalance an organization when management realizes that there is an impending balance between the different poles of paradox.
While prior paradox literature has recognized the importance of cognitive and structural differentiating and integrating practices (e.g., Jarzabkowski, Lê, & Van de Ven, 2013; Smith, 2014; Smith & Tushman, 2005), it has not theorized on their distinct qualities and how they can be used to address different threats to the paradox. In the following, we discuss the characteristics of different kinds of differentiating and integrating practices, put forward the concept of processual differentiation and integration, and discuss how these practices can be used to address different types of internal drift.
Differentiating and integrating can take place along multiple dimensions. Already the pioneering contingency theorists discussed differentiation and integration in terms of organizational structures and individuals’ interpersonal, temporal, and goal orientations (Lawrence & Lorsch, 1967). While the early paradox research focused predominantly on structural differentiation and integration (e.g., Tushman & O’Reilly, 1996), later research has also examined the use of cognitive (Smith & Tushman, 2005), organizational (Andriopoulos & Lewis, 2009) and leadership practices (Smith, 2014) for differentiation and integration.
We argue that different types of drift require the use of different rebalancing practices. As elaborated above, individual tendencies to drift commonly relate to different cognitive or behavioral biases that individuals have (e.g., Das & Teng, 1999; Levinthal & March, 1993; Schwenk, 1986). Cognitive and behavioral biases may be countered with cognitive differentiating or integrating. Cognitive differentiating can be used to emphasize the separate identities of competing poles. Cognitive integrating, on the other hand, can be used to emphasize synergies between them. Whereas cognitive differentiating and integrating act on individuals’ minds and cognitive frames, processual differentiating or integrating acts on the processes that individuals follow. Prior research has found that processes can help effectively alleviate individual-level behavioral biases (e.g., Hodgkinson, Maule, Bown, Pearman, & Glaister, 2002) and to guide individuals to better differentiate the different poles of the paradox or to put more emphasis on finding integrative synergies between them.
Interactional tendencies to drift differ from individual tendencies to drift in that they are driven by interpersonal or interunit interactions of different organizational actors and entities. While the everyday interactions of individuals in an organization are influenced by the structures, systems, and processes established for collaboration and coordination, they are also influenced by a number of situational factors, the randomness of the encounters of different actors, and the power structure of the organization (e.g., Cohen, March, & Olsen, 1972). Corrective processual differentiating or integrating can help rebalance an organization by, for example, shaping the different processes and practices through which different organizational entities interact with each other. For example, in a technology company with a decentralized setup, a specific technology process may be necessary to coordinate and align the technology choices that its different organizational units continuously make (e.g., Vaast & Levina, 2006).
Similarly, structural differentiating or integrating by combining or dividing organizational entities can be used to address interactional drifting tendencies. Structural integrating can aid the realigning incentives and eliminating challenges involved in cross-unit collaboration. Structural differentiating can help eliminate the challenges of intra-unit collaboration and negative synergies (e.g., Es-Sajjade et al., 2021; Friesl, Garreau, & Heracleous, 2019). Hence, depending on the direction of drift, altering organizational processes and structures can help address different coordination and cooperation deficits in an organization.
Finally, organizational tendencies to drift can be correctively rebalanced through structural differentiating or integrating practices. Broader organizational tendencies to drift tend to emerge as a company grows (Eisenhardt et al., 2010). Growth can entail changes that inadvertently skew organizational members’ behaviors toward one side of the tensions or the other. While cognitive or processual differentiating and integrating practices can also be used to address these drifting tendencies, these practices are likely to be of only limited effectiveness since the challenges are fundamentally structural. Accordingly, dynamic contingency misfits may require either structural integration, for example, through the combination of organizational units (e.g., Barkema & Schijven, 2008; Es-Sajjade et al., 2021), or differentiation through structural separation or partial spinoffs (e.g., Friesl et al., 2019; Tushman & Nadler, 1986; Tushman & O’Reilly, 1996).
Overall, while these rebalancing practices can help temporarily reestablish balance, the organization will remain in an unstable state “at the edge of chaos” (e.g., Davis et al., 2009). Instead of attaining a sustainable dynamic equilibrium and a virtuous cycle of success (e.g., Andriopoulos & Lewis, 2009; Lewis, 2018; Lewis & Smith, 2014; Smith & Lewis, 2011), the interplay of dissipative forces and reactive rebalancing practices will persist (Cunha & Clegg, 2018), requiring continuous investments of management’s attention and energy into the management of paradox.
Two rhythms of paradox management
The dynamic equilibrium model and the dissipative equilibrium model can also be seen to represent two different rhythms of paradox management, as shown in Figure 3. The inner loop corresponds to the original dynamic equilibrium model (Smith & Lewis, 2011). According to the model, tensions can become salient either through external environmental factors or actors’ paradoxical cognition. Environmental factors that can render tensions salient include plurality that can surface competing goals, environmental change, and resource scarcity that can exacerbate tensions. Similarly, actors’ paradoxical cognition and rhetoric can render tensions salient by focusing attention on specific tensions.
Two Rhythms of Paradox Management.
Acceptance provides the basis for paradoxical resolution. In contrast, defensive responses that favor only one pole of the paradox are expected to lead to vicious cycles. Paradoxical resolution then takes place through management’s iterating responses of “splitting and integration.” After the paradox is resolved, tensions become latent, and the paradox becomes organizationally embedded in “strategies, structures, rules, processes, and identities” as “inherent features” of the organization and agency that reproduce them (Smith & Lewis, 2011, p. 393). According to the authors, the acceptance and resolution of the paradox provide a basis for “an upward spiral of peak performance and sustainability” (Lewis & Smith, 2014, p. 137). Accordingly, the dynamic equilibrium model, represented by the inner cycle, captures well the ongoing management of salient tensions with management preemptively alternating its support between the different poles of the paradox and over time institutionalizing the management of the paradox as part of an organization’s everyday work (Knight & Paroutis, 2017; Smith & Lewis, 2011).
Such preemptive balancing actions do not yet suffice, however, to sustain a paradox that can reemerge in new, unexpected ways either through external chance events or different types of internal drift, as described above. In order to account for this, the outer loop of the figure adds the dissipative dynamics to the model by showing how external chance events and different internal tendencies to drift can lead to an unexpected resurfacing of paradox in new manifestations, which are difficult for senior management to anticipate and preemptively account for.
The difficulty to anticipate and preemptively account for the reemergence of the paradox in its new manifestations is driven by the limitations of senior management’s cognitive and information processing capacity. Many different types of drift can occur simultaneously at different levels of an organization and in different organizational entities, making it hard for senior management to be continuously aware of all of these developments. Moreover, in addition to the difficulty to make sense of new manifestations of the paradox, it is even more difficult for senior management with limited agency to subsume the organization under its control to prevent imbalances from resurfacing.
To illustrate the intuition and dynamics of the dissipative equilibrium model, we provide a short vignette of a company in which coordination deficits led to interactions that caused the organization to drift out of balance. To rectify the situation, senior management had to rebalance the organization with processual integrating. In the company, balancing the tension between central control and decentral autonomy was regarded as one of its key success factors. The management team had accepted and embraced the tension, and, over the course of thirty years, the paradox had become embedded into the organization’s strategy, structure, and processes. While the company was centrally managed, it was structured into self-organized teams, and employees had veto rights over management’s key decisions.
After having participated in strategy training at a local university, the management became interested in further democratizing the company’s strategy process and in opening it to wider employee participation (e.g., Hautz, Seidl, & Whittington, 2017; Whittington, 2019). While the employees were at first enthusiastic about this more inclusive approach to strategy, the process, based on a mixed, organization-wide group of employees working on the company’s strategic vision, gradually transformed into an inefficient, energy-consuming endeavor. The group of employees participating in the strategy process was so large and diverse that it was difficult for its members to agree on any clear direction. As a consequence, the employees did not make any headway in the process for a long period of time. Moreover, the process required significant time and effort, depleting daily resources, and causing many operational problems. Eventually, after two years, top management intervened and terminated the uncoordinated strategy process. Hence, despite management’s acceptance of the paradox, the paradox between central control and decentral autonomy ended up resurfacing in a new, unexpected form, and senior management had to correctively rebalance the organization and reassume control of the process.
To show how the dynamics of these two models play out over time, Figure 4 provides a comparison of the longer-term outcomes of the dynamic equilibrium and dissipative equilibrium model. The upper part of the figure shows the dynamic equilibrium model, in which management supports the two poles of the paradox in a consistently inconsistent manner. This behavior is depicted with an alternating “zig-zag” pattern of arrows shown across the timeline. Since the dynamic equilibrium model assumes that management becomes better at managing paradoxical tensions over time, the amplitude of preemptive balancing actions decreases over time (e.g., Smith & Besharov, 2019). Hence, the assumption in the dynamic equilibrium model is that once management learns to deal with paradoxes it can subsume the paradox under its control.
The Dynamic Equilibrium Model Versus the Dissipative Equilibrium Model.
The lower part of Figure 4 shows the corresponding dissipative equilibrium model. Even when management has accepted the paradox and is supporting its both poles equally, external chance events and internal drifting tendencies may cause the organization to unexpectedly drift out of balance. These effects are depicted in the diagram with longer dotted arrows. In contrast to the dynamic equilibrium model, where the main concern is senior management’s acceptance of the paradox (“actors in an unstable state”), in the dissipative equilibrium model the main concern is the unpredictability of the organization (“organization in an unstable state”).
Since there are different types of internal drift that are caused by the everyday decisions and actions of different organizational members, it is difficult for senior management to recognize and counter all these threats before an emerging imbalance becomes inevitable. The difficulty of predicting and preparing for all possible outcomes makes it more economical for senior managers to concentrate their time and effort on the development of effective ways to counter the different types of drift when an imbalance emerges rather than investing significant time and effort to attempt to preemptively manage the paradox.
In the dissipative equilibrium model, the unpredictability of different internal and external threats to the ongoing management paradox also makes it challenging for boundedly rational managers with limited managerial agency to develop a virtuous cycle of learning (e.g., Lewis & Smith, 2014; Smith & Lewis, 2011). While such learning may take place, it will likely be disrupted by different unanticipated discontinuities and can take even longer than the dynamic equilibrium model would seem to suggest.
Discussion
We put forward in this paper the dissipative equilibrium model of paradox. In contrast to the dynamic equilibrium model, which assigns a central role to managerial agency in recognizing and resolving paradoxes, the dissipative equilibrium assigns to managers more limited “embedded agency” (e.g., Floyd, Cornelissen, Wright, & Delios, 2011). Managers are often confronted with paradoxes in an organizational reality where “change, flux, and unintended consequences” (MacKay & Chia, 2013, p. 226) play an important role in determining outcomes, making it difficult to detach the management of paradox from the internal dynamics of the organization.
These different views on managerial agency and on the role of the organization are associated with different assumptions regarding the dynamics of paradox management, as shown in Table 1. The dissipative equilibrium model differs in its assumptions regarding (1) the nature of the equilibrium—the balance is more unstable than previously suggested due to unexpected external chance events and different types of internal drift that continuously threaten the temporary balance; (2) the role of managerial agency—previously accepted paradoxes appear in new, unexpected manifestations that are often difficult for management to recognize and preemptively control for; (3) the role of tensions—the pull of the different poles of paradox is unequal and changes continuously; (4) the dynamics of paradox management—management must invest energy and be vigilant in recognizing different types of drift and to develop corrective interventions to rebalance the organization; and (5) longer-term outcomes—acceptance, paradox resolution, and embedding the paradox into structures, systems, and processes do not yet suffice in ensuring longer-term sustainability. Thus, virtuous cycles may be more difficult to achieve than prior research has suggested (e.g., Lewis & Smith, 2014; Smith & Lewis, 2011).
The Dynamics of Paradox: Comparison of Theoretical Frameworks.
The dissipative equilibrium model of paradox contributes to the development of a more realistic understanding of the organizational challenges and longer-term outcomes of paradox management. We next elaborate on how the model enhances our understanding of nonorderly paradox dynamics, the interplay of different types of internal drift and corrective rebalancing practices, and the longer-term organizational outcomes of paradox management.
Nonorderly paradox dynamics
We offer a novel explanation for why “disequilibria mark reality” (Schad et al., 2018, p. 6). We argue that the dynamics of paradoxical tensions in organizations are driven not only by actors’ deliberate intentions but also by external chance events and different types of internal drift. While the dynamic equilibrium model focuses on such “owned” dynamics under management’s control, the dissipative equilibrium model focuses on the “unowned process” (MacKay & Chia, 2013) driven by the dynamic nature of an organization. By focusing on these “unowned” processes that manifest in different types of internal drift, we contribute to a view of paradox that situates managerial agency in the context of “amorphous, disparate, and impersonal processes” (MacKay & Chia, 2013, p. 208). The dynamics of paradox are thus shaped not only by actors’ deliberate choices but also by the unintended consequences of their actions (e.g., Es-Sajjade et al., 2021).
Management research has previously used the concepts of dissipative equilibrium and dissipative structures to describe organizational change and transformation processes (e.g., Gemmill & Smith, 1985; Macintosh & Maclean, 1999), intraorganizational ecological dynamics (Burgelman, 2010), and the optimal amount of structure needed to compete in a dynamic environment at “the edge of chaos” (e.g., Anderson, 1999; Davis et al., 2009). We introduce the concept into the paradox literature to complement the more orderly view of paradox dynamics represented by the dynamic equilibrium model. The dissipative equilibrium model helps us address the recent criticism of a “dominant logic of order” attributed to the dynamic equilibrium perspective (Cunha & Putnam, 2019, p. 99; Putnam et al., 2016). Similar to the dynamic equilibrium model (Smith & Lewis, 2011), the dissipative equilibrium model conceptualizes balancing opposing poles as a dynamic process. At the same time, however, the dissipative equilibrium model also accounts for the potentially defiant nature of paradoxes. Thus, in the dissipative equilibrium model, order is a fleeting state that must be constantly recreated. Accordingly, an organization is “not in equilibrium” but “rather in ‘orderly disequilibrium’” (Brown & Eisenhardt, 1998, p. 29). By elucidating how “the interplay between order and disorder” shapes paradox dynamics (Cunha & Putnam, 2019; Putnam et al., 2016), the dissipative equilibrium model provides an alternative means of understanding what “living with” paradoxical tensions entails (Smith & Lewis, 2011, p. 385).
Much of the prior research in organization theory on the emergence of dissipative equilibria or dissipative structures has built on chaos and complexity theories (e.g., Anderson, 1999; Davis et al., 2009). This is not surprising since dissipative equilibria can be seen to represent temporary states in complex adaptive systems (Davis et al., 2009; Prigogine & Stengers, 1984). Future research on the dissipative equilibrium model of paradox could similarly benefit from building on research on complex adaptive systems. The interwoven (Jarzabkowski et al., 2013; Sheep, Fairhurst, & Khazanchi, 2017) and nested nature of paradoxes over time and space (Andriopoulos & Lewis, 2009, 2010) makes it necessary for paradox researchers to find new ways to better account for complex systems of dynamic interdependencies (e.g., Anderson, 1999; Stacey, 1995; Thietart & Forgues, 1995) involved in the management of paradoxes (Schad & Bansal, 2018).
Interplay of drifting tendencies and rebalancing practices
Since the dynamic equilibrium model assumes that a successful management of paradoxes is dependent on management’s acceptance of a paradox and its continuous preemptive support for the competing poles of the tension, research on other threats to the balance of paradox is rather limited. In addition to external “black swan” events, we identify different types of internal drift that can cause an organization to drift out of balance despite management’s continuous and equal support of the opposing poles.
Individual tendencies to drift are part of human nature. For example, cognitive inertia (Siggelkow, 2001; Tushman & Romanelli, 1985), isomorphic behavior (D’Aunno, Succi, & Alexander, 2000; DiMaggio & Powell, 1983), and uncertainty avoidance (e.g., Ahmadi, Khanagha, Berchicci, & Jansen, 2017; O’Reilly & Tushman, 2013), which influence the everyday decisions and actions of all organizational actors, can cause a drift out of balance. Interactional tendencies to drift relate to coordination or collaboration deficits that can similarly lead to unexpected organizational outcomes (e.g., Balogun & Johnson, 2005; Merton, 1936). Finally, the growth and maturing of an organization can lead to increasing structural inertia (e.g., Desai, 2008; Le Mens, Hannan, & Polos, 2015; Tushman & Romanelli, 1985) or dynamic contingency misfits (e.g., Burton, Lauridsen, & Obel, 2002; Westerman et al., 2006; Zajac et al., 2000).
Since the different types of internal drift are difficult for management to anticipate and preemptively manage, the way in which management makes sense of and reacts to early signs of drift and how it takes corrective actions to address emerging imbalances play an important role in the ability of management to sustain the balance of competing poles. Therefore, we call for future research examining how the different challenges to paradox relate to the difficulties of management to recognize them and whether, for example, prior experience enables management to become better at anticipating the different threats to balance.
We also call for future research examining the different corrective rebalancing practices that organizations can use to counter different types of drift. We identify cognitive, processual, and structural rebalancing practices that management can selectively deploy to address different types of drift. While prior literature has also distinguished between cognitive and structural differentiating and integrating practices (e.g., Jarzabkowski et al., 2013; Smith, 2014; Smith & Tushman, 2005), we go beyond the existing research in suggesting that some of these different kinds of differentiating and integrating practices may be more effective than others in addressing specific types of threats. Moreover, in contrast to the dynamic equilibrium model that portrays them as preemptive balancing practices, we view them as corrective means to rebalance an organization to prevent drift occurring in the opposing direction. Specifically, we suggest that cognitive and processual differentiating and integrating practices may be useful in countering individual tendencies to drift. Processual and structural differentiating and integrating practices may be effective in countering interactional tendencies to drift. Finally, structural differentiating and integrating practices may be effective in countering organizational tendencies to drift.
Longer-term organizational outcomes
Finally, we contribute to an enhanced understanding of the longer-term outcomes of paradox management. Prior research has suggested that maintaining dynamic equilibrium enables virtuous cycles of “peak performance and sustainability” (Lewis & Smith, 2014, p. 137; Smith & Besharov, 2019; Smith & Lewis, 2011). Accordingly, studies have suggested that when actors accept and value paradox and apply appropriate management strategies, organizations can thrive as organizational members learn to more effectively and efficiently manage tensions between competing demands, creating virtuous positive learning spirals (Raisch et al., 2018; Smith & Lewis, 2011).
We argue that the interplay between paradoxical forces and organizational members’ responses is a more precarious process. While management may be able to anticipate and proactively counter some of the different threats, even experienced managers will at times fail to recognize drift in a timely manner and have to correctively rebalance an organization. Hence, even when actors accept tensions, their understanding of the dynamics of paradox remains necessarily partial. Therefore, instead of triggering positive, self-reinforcing learning dy-namics, paradox management is likely to remain an ongoing challenge as the challenges confronting managers continue to evolve along with the organization and its environment.
By echoing the importance of adopting a more longitudinal view of paradoxes (Cunha & Clegg, 2018; Putnam et al., 2016), we join the emerging stream of paradox research that has begun to explore the temporal dynamics of paradoxical tensions in organizations (e.g., Es-Sajjade et al., 2021; Jarzabkowski et al., 2021, 2013; Sheep et al., 2017; Smith & Besharov, 2019). While some of the substreams of paradox research, such as the research on ambidexterity, have contributed to our understanding of how paradoxes can evolve and unfold over time (e.g., Luger, Raisch, & Schimmer, 2018), further work is needed to better understand the interplay of paradoxical tensions and management actions.
Conclusion
The dynamic equilibrium model has become a foundational building block of paradox theory. The model conceptualizes maintaining equilibrium amid organizational tensions as continuous balancing. Building on recent theoretical work that has challenged the dynamic equilibrium model for its lack of process sensitivity (Cunha & Clegg, 2018; Cunha & Putnam, 2019; Fairhurst, 2019; Farjoun, 2016; Putnam et al., 2016), we focus on how the dynamics of paradox play out beyond “a steady state of dynamic equilibrium” (Cunha & Clegg, 2018, p. 17). Based on our analysis, we put forward the dissipative equilibrium model as a means to better capture the temporary nature of balance and the continuous attention and interventions needed from management to (re)balance an organization. We hope that the insights provided by the model inspire further research and theory development of the nonorderly nature of paradoxes and help better elucidate why the balancing of opposing tensions remains a continuous challenge.
Footnotes
Acknowledgements
We would like to thank the Editors David Seidl and Joep Cornelissen for their constructive comments and clear editorial guidance throughout the review process. We also like to acknowledge the valuable comments of Marianne Lewis, Sebastian Raisch, and Wendy Smith on early versions of this paper.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.