Bridging the policy-practice gap: a dual challenge of organizational learning

Abstract

Purpose

This paper aims to further our understanding of policy–practice gaps in organizations from an organizational learning perspective. The authors conceptualize and analyze policy–practice gaps in terms of what they label the dual challenge of organizational learning, i.e. the organizational tasks of both adapting ongoing practices to prescribed policy demands and adapting the policy itself to the needs of practice. Specifically, the authors address how this dual challenge can be understood in terms of organizational learning and how an organization can be managed to successfully resolve the dual learning challenge and, thereby, bridge policy–practice gaps in organizations.

Design/methodology/approach

This paper draws on existing literature to explore the gap between policy and practice. Through a synthesis of theories and an illustrative practical example, this paper highlights key conceptual underpinnings.

Findings

In the analysis of the dual challenge of organizational learning, this study provides a conceptual framework that emphasizes the important role of tensions and contradictions between policy and practice and their role as drivers of organizational learning. To bridge policy–practice gaps in organizations, this paper proposes five key principles that aim to resolve the dual challenge and accommodate both deployment and discovery in organizations.

Research limitations/implications

Because this is a conceptual study, empirical research is called for to explore further and test the findings and conclusions of the study. Several avenues of possible future research are proposed.

Originality/value

This paper primarily contributes by introducing and elaborating on a conceptual framework that offers novel perspectives on the dual challenges of facilitating both discovery and deployment processes within organizations.

Citation

Martin, J., Ellström, P.-E., Wallo, A. and Elg, M. (2024), "Bridging the policy-practice gap: a dual challenge of organizational learning", The Learning Organization, Vol. ahead-of-print No. ahead-of-print. https://doi.org/10.1108/TLO-05-2023-0079

Introduction

The notion of a policy–practice gap refers to instances where there is a mismatch between a prescribed policy and the ongoing practices the policy is intended to regulate. That is, a mismatch between general and uniform policy demands, on the one hand, and the diverse conditions and needs of local practices on the other. This phenomenon has been identified and described in different research fields over quite a long time and is illustrated by seminal studies from a broad range of fields: policy implementation (Pressman & Wildavsky, 1984); public services and street-level bureaucracy (Lipsky, 2010); information sciences/knowledge management (Brown & Duguid, 1991; Pfeffer & Sutton, 2000); innovation systems (e.g. Jensen, Johnson, Lorenz, & Lundwall, 2007); and cognitive engineering/human factors (Hollnagel, 2017). It has also, more recently, been addressed in empirical work, for instance, focusing on the difficulty of implementing policy-oriented initiatives aimed at making innovation more responsible (Sischarenco & Luomaranta, 2023).

Common to these and several other studies is the idea that policy–practice gaps are generated by tensions between relatively stable and codified policies (sometimes viewed as science- or evidence-based) and fluid, informal and improvisational practices evolving through day-to-day interactions. This kind of tension can be exemplified by Brown and Duguid’s (1991) distinction between formally prescribed, canonical practices (“telling people what to do and how”) and actual, non-canonical practices that evolve in and through daily work processes. Hollnagel (2017) makes a partly parallel distinction between work-as-imagined and work-as-done. Jensen et al. (2007) discuss knowledge management practices based on codified scientific and technological knowledge versus experience-based practices based on tacit knowledge and informal interactions.

Although many studies recognize and discuss these and similar tensions as important for understanding policy–practice gaps and related phenomena (e.g. knowing-doing gaps, Pfeffer & Sutton, 2000), there are relatively few attempts to analyze and try to understand the conditions and principles that generate policy–practice gaps, and that also might be helpful for handling such gaps in a reflective and productive way. If we look closer at some of these analytical attempts, it soon becomes apparent that concepts of organizational learning (Basten & Haamann, 2018) are, in one way or another, used as a central explanatory framework. A well-known case in point is March’s (1991) discussion of exploitation and exploration as two modes of learning but also as two modes for handling and balancing the impact of existing organizational codes (policies, cultures, codified knowledge) and the search for new knowledge and ways of thinking. Other influential examples include Pressman and Wildavsky’s (1984) discussion of a planning-control mode of implementation versus an explorative, learning-based mode (fidelity vs adaptation); Zollo and Winter’s (2002) models of learning of dynamic capabilities and operating routines in organizations; and the knowledge management model proposed by Nonaka and Takeuchi (1995) based on cyclical processes of knowledge conversion through interactions between tacit and explicit knowledge.

Inspired by and building on these conceptualizations, the purpose of this paper is to further our conceptual understanding of policy–practice gaps in organizations from an organizational learning perspective. To advance this purpose, we will conceptualize and analyze policy–practice gaps as an outcome of an iterative learning and improvement cycle. More specifically, an actual policy–practice gap during a certain period is assumed to be a function of how the organization (i.e. its leading actors) handles the two organizational learning tasks of adapting ongoing practices to prescribed policy demands and adapting the policy itself to the needs of practice in a next step of the iterative learning cycle. We will call this the dual challenge of organizational learning in relation to policy–practice gaps.

Against this backdrop, two main research questions will be addressed. First, how can the dual challenge in handling policy–practice gaps be analyzed in terms of organizational learning? Second, how can an organization be managed to successfully handle this dual learning challenge and, thereby, also bridge policy–practice gaps in the organization?

The remainder of the paper is structured as follows. First, we present a short overview of the field of organizational learning. Second, we introduce a model intended to illustrate policy–practice gaps as an outcome of an iterative organizational learning and improvement cycle. In the third section, we address the issue of how to manage organizational learning to bridge (or ideally avoid) policy–practice gaps in an organization. Structural and cultural conditions, as well as five key principles for organizational learning, are proposed and supported by literature examples. Next, an illustrative example is given related to a Swedish debate on a trust-based approach to governance and management in public sector organizations. Finally, we conclude our findings and present some implications for further research and practice.

The field of organizational learning

Although organizational learning has been a subject of research within management and organizational studies at least since the early 1960s, interest in this concept has grown considerably in recent times (Alerasoul, Afeltra, Hakala, Minelli, & Strozzi, 2022; Elkjaer, 2021). Not least against the backdrop of the ongoing green transformation, an increased focus on social sustainability and the crises management activities prompted by the COVID-19 pandemic, organizational learning has been highlighted as instrumental in developing strategies for organizational change and resilience (Orth & Schuldis, 2021; Rupčić, 2022a; Tortorella, Narayanamurthy, & Staines, 2021).

Top-down and bottom-up processes of organizational learning

Organizational learning refers to descriptive and explanatory studies of learning processes within and between organizations (Basten & Haamann, 2018; Dixon, 2017). Processes of organizational learning furthered through an interplay between the individual, group and organizational levels, have been conceptualized in many different frameworks and models over the years (Crossan, Lane, & White, 1999; Huber, 1991; Senge, 1990). Considering some of the more well-known conceptualizations, it becomes apparent that these can be ordered into two main groups. On the one hand, there are conceptualizations that build on a top-down view of learning, that is, learning in the sense of putting new knowledge, new ideas or new policy into action. In general terms, this view of learning is akin to what March (1991) calls learning as exploitation, or what we later in this article will call adaptive learning or learning through the deployment of policies, ideas, or knowledge. This view of organizational learning could be exemplified by many interpretations of evidence-based practice, where the focus is on the implementation of codified, research-based knowledge (Nilsen, Avby, & Ellström, 2012).

On the other hand, there are also conceptualizations of organizational learning – admittedly less recognized and studied – that build on a bottom-up view of learning, that is learning driven by practice and the emergent solutions to unforeseen events and disturbances that occur continuously, that is, what March (1991) calls learning as exploration. As conceptualized by Zollo and Winter (2002), this type of bottom-up learning is generated through mechanisms of experience accumulation, articulation and codification of knowledge. We will later refer to this mode as developmental learning or learning through a logic of discovery.

March (1991) argued that there is a need to find a balance between exploration and exploitation. In line with more contemporary research (Brøns Kringelum & Brix, 2020), this is also the main assumption behind this paper. Unfortunately, however, there is a paucity of knowledge concerning what we here denote as the “dual challenge”, that is, understanding how top-down and bottom-up processes of organizational learning can work together, as dual concepts, inherently connected in a conceptual sense but still adamantly different in a processual sense.

Defining organizational learning

There are three main approaches informing definitions of organizational learning: the cognitive-behavioral approach, the sociocultural approach and the knowledge-creating approach (Paavola, Lipponen, & Hakkarainen, 2004). A basic assumption underlying the cognitive-behavioral approach is that organizational learning is about the development and change of routines through the accumulation of experience and feedback in relation to targets (March, 1991; March & Olsen, 1976). Culture and cultural context in learning organizations (Dahl & Irgens, 2022; Rupčić, 2022b) is a core concept of the sociocultural (or situated) approach with learning as participation in work practices and activities A key assumption is that learning cannot be separated from work and other social practices where it is assumed to take place or be used (Brown & Duguid, 1991) and is a matter of participation in practices (Lave & Wenger, 1991). The knowledge-creating approach is based on the view that the production, transformation and utilization of knowledge are fundamental for understanding organizational learning (Paavola et al., 2004). Learning is viewed as an interplay between intra- and inter-individual (social) processes of knowledge creation. This model assumes that knowledge creation could be understood as a cyclical process of knowledge conversion based on the interaction between tacit and explicit knowledge (Nonaka & Takeuchi, 1995).

How, then, do we understand organizational learning in this paper? Many definitions and studies of organizational learning focus on a locus of learning that is grounded in an individual level of analysis and a view on learning as an interplay between processes of cognition and action of individuals in an organizational context (Huber, 1991; Simon, 1991). Other definitions of organizational learning focus on a collective subject, that is, the group or team as the locus of learning (Edmondson, 2002; Kérivel, Bossard, & Kermarrec, 2021; Senge, 1990). In line with Crossan et al., 1999), we rather view organizational learning as furthered through a complex, continuous interplay between the individual, group and organizational levels.

To be more specific, we define organizational learning as the dual process of putting new knowledge/technology, ideas or policies into action (exploitation) and, simultaneously, adapting the new knowledge, ideas or policies to the needs, unforeseen events or problems that are likely to occur in ongoing organizational practices (exploration). Accordingly, ongoing practices may be opportunities for redesign and renewal of the new policies as originally designed and prescribed, that is, as potential drivers of practice-based innovation (Ellström, 2010; Jensen et al., 2007).

Conceptualizing policy–practice gaps

A model conceptualizing policy–practice gaps as a possible outcome of an iterative organizational learning and improvement cycle is presented in Figure 1. The model depicts a learning system (e.g. an organization or a work team) that can modify its practices in response to processes or events in its internal or external context.

A basic idea underlying the model in Figure 1, is a view of organizational learning as driven by a double duality of organizational conditions. On the one hand, there is a duality between formal and informal aspects of an organization; specifically, between prescribed policies and ongoing practices. On the other hand, there is a duality between what we call a logic of deployment and a logic of development, respectively. By integrating these two dualities into one model, we create a basis not only for a new understanding of mismatches between policy and practice but also for a theoretical integration between a rationalistic and top-down view of organizational learning and a view that also emphasizes the informal, ad hoc and practice-based character of learning in organizations.

Another fundamental idea underlying the model in Figure 1 is that tensions and contradictions between organizational practices as officially prescribed (the policy dimension) and as perceived and carried out in practice (the practice dimension) create potential for organizational learning. The extent to which these potentials for learning are realized or not, and whether the learning will be mainly adaptive or developmental in character (Ellström, 2001), will depend on the choices that are made – intendedly or unintendedly – by the management or by other significant actors concerning the balance between and the relative strength of the logics of deployment and discovery.

It is important to underline, though, that making the distinction between a policy and a practice dimension, does not necessarily imply a hierarchical relationship between the two as, for example, between the management system and the operational activities in an organization. It could just as well refer to the formal and informal aspects of, for example, a work team or a work unit within a larger organizational structure.

We will use the remainder of this section to explicate basic elements and aspects of the proposed framework.

The policy–practice duality

Let us start with the distinction between policy and practice. What we call the policy dimension comprises prescribed policies, guidelines, etc. that are part of the explicit and official image of an organization (or a team) and its operations, that is, its formal aspects. Thus, the policy dimension is in a sense, a “frozen system”, comprising a world of abstract concepts, models, visions and strategies that are intended to regulate the world of operations and work. However, as shown by empirical research, the policy dimension may in many cases be only loosely coupled to operational activities and routines (Meyer & Rowan, 1977; Weick, 1976). The policy dimension’s complex decision and interaction processes typically involve many actors with different ideas and interests, which makes it likely to expect more or less open conflicts and negotiations between concerned actors (Feldman & Pentland, 2003). The normative ideas and structures that prevail at a certain point in time can be a result of negotiations and collective learning processes (Barley & Tolbert, 1997) or of adaptations to pressures from different institutional environments to adopt “new” ideas or practices (Brunsson & Jacobsson, 2000; Meyer & Rowan, 1977).

The practice dimension, on the other hand, represents the implicit (tacit), usually hidden and seldom exhibited aspects of an organization (Brown & Duguid, 1991; Hollnagel, 2017), that is, what is often referred to as the informal organization (Mintzberg, 1975) or the shadow system (Stacey, 2007). This dimension concerns how organizational routines and strategies are perceived and performed in practice, in contrast to how they are prescribed and intended to be performed according to official standards and policies. Thus, the practice dimension concerns how organizational operations are subjectively understood and carried out by different actors in different roles. If the policy dimension as stated above can be aptly characterized as a frozen system, the practice dimension may be characterized as a floating system in the sense of a dynamic, constantly changing system of actors and work processes.

There are often significant differences between how an operation appears from the point of view of the policy and the practice dimension, respectively (Brown & Duguid, 1991; Brunsson & Jacobsson, 2000; Meyer & Rowan, 1977). The practice dimension can, to some extent, be viewed as having a “life of its own” with a certain autonomy in relation to formal structures and prescribed processes and tasks (Brix, 2019). This relative autonomy – also understood as loose couplings between the two dimensions (Weick, 1976) – may be manifested as improvisations and deviations in relation to formally prescribed routines or processes. There is for example often considerable creativity and an ability to improvise when it comes to finding solutions to unexpected problems that arise at work (Brown & Duguid, 1991; Feldman & Pentland, 2003; Miner, Bassof, & Moorman, 2001). However, as is also noted by these and other authors, this creativity occurs mainly unofficially and implicitly – as a part of what happens “behind the scenes” – and it may therefore not be highlighted or paid attention to, but rather run the risk of becoming forgotten. By making such “invisible” processes visible and exploring their potential value for the organization, they may survive and become valuable and established practices within the operations (Miner et al., 2001).

The organizational learning duality

The second basic duality in Figure 1, is that between what we call the logics of deployment and discovery, respectively. These concepts has their roots not only in learning theory, in particular, theories of adaptive/reproductive and developmental/innovative learning (Ellström, 2001; Engeström, Miettinen, & Punamäki, 1999) but also in management theory and the previously mentioned distinction between exploitation and exploration as proposed by March (1991). It is assumed that the relative strength of these two processes affects the balance and interplay between policy and practice. In particular, the extent to which practices are adapted to prescribed policies (the degree of implementation fidelity) or, conversely, the extent to which the policies are adapted to better fit the demands of practice. To align policy and practices through these two types of learning processes is what we in this paper call the dual challenge of organizational learning. In the broader field of implementation research, these dual processes of adaptation have been described as mutual adaptation (Leonard-Barton, 1988; McLaughlin, 1976), a strategy that has been shown to be effective in accomplishing successful policy implementation in organizations.

Now, looking more closely at the two logics, the logic of deployment focuses on how elements of the policy dimension are reproduced and realized in ongoing practices. Thus, it covers activities that aim to implement and maintain officially prescribed policies, structures and practices. To achieve this, there is a strong emphasis on goal consensus, standardization and stability. In several respects, this logic is related to what March (1991) refers to as processes of exploitation in organizational learning. However, there are also clear similarities with well-known Tayloristic principles of production (Braverman, 1998). As understood here, the logic of deployment is conceptualized as a process of adaptive/reproductive learning (Ellström, 2001, 2005). This mode of learning has a focus on establishing and maintaining well-learned and routinized action patterns. A basic condition of adaptive learning and, thus, of the logic of deployment, is to reduce variation – within and between individuals – regarding the perception and performance of a task. Example of organizational measures to reduce variation include formalization, restricted autonomy and the formulation of clearly specified tasks and goals.

Contrary to this orientation, the logic of discovery focuses on organizational practices as a source of new thinking and knowledge development, that is, on promoting renewal in ways of defining and carrying out organizational tasks. It is assumed that this renewal will be based on the variation that always exists in the performance of organizational processes and routines, a variation that may lead to development of new ideas and practices. Under certain conditions, it is also assumed that processes of exploration and renewal may lead to redefinitions and transformations of prescribed policies or structures in the policy dimension. According to this logic, then, rather than reducing variation, variation in ongoing practices is used as a source to analyze and improve organizational processes (Edmondson, 2012).

In terms of learning theory, the logic of discovery is understood here as based on developmental or innovative learning (Ellström, 2001, 2005; Engeström et al., 1999). That is, a mode of learning that emphasizes critical reflection and questioning of habitual routines (Dewey, 1933), unlearning (Hedberg, 1981) and development of new solutions and routines that may replace established ways of working. Thus, there is a focus on exploring and possibly transforming rather than reproducing and maintaining prevailing routines and policies.

The distinction between the two logics of deployment and discovery is in some respects parallel to the two modes of organizing that Edmondson (2012) calls “organizing to execute” and “organizing to learn”. Which of these two modes predominates in a certain organization at a certain time, depend according to Edmondson (2012) on the management’s cognitive framing of the organization. That is, the management’s beliefs, assumptions and choices concerning critical organizational parameters. Examples of such critical parameters include hiring policy – conformers and rule-followers versus problem-solvers and experimenters; training – learning before doing versus learning from doing; employee discretion and empowerment – follow the script (formalization) versus there is no script to follow (improvisation). The first mentioned choice in each pair relates to the mode “organizing to execute” (the logic of deployment), whereas the second choice in each pair relates to “organizing to learn” (the logic of discovery).

Interplay and dialogue between organizational actors

Another important element of the model in Figure 1 is what we call a space for interplay between organizational actors. This space may be temporary or permanent; formal or informal; its members may meet face-to-face or in a digital environment (a virtual space) depending on the circumstances at hand and the choices that are made. The important function of this interactive space is to be an arena for dialogue across different types of boundaries in an organization (Nonaka & Konno, 1998). Boundaries that could be based, for example, on differences in hierarchical or professional status, in organizational position or in area of expertise (Edmondson, 2012). In relation to the model in Figure 1, a focal boundary is that between actors in the domains of policy-making and practical operations, respectively. The purpose of creating this type of space would be to create a structure and procedures for identifying and exploring differences concerning perspectives, values and believes concerning important topics (e.g. a new change initiative in the organization) with the possibility in mind to reach a common understanding of the issues at hand and how to deal with these issues.

The role of context

The interplay between the policy and practice dimensions and the two logics of learning takes place within a specific contextual setting. This setting changes continuously over time, creating a dynamic learning environment for actors in the organization. Specifically, we distinguish between an internal and an external context, respectively. The internal context comprises an almost unlimited number of proximal factors in the organization that may influence (constrain or enable) processes of change and learning. These proximal factors are related to different aspects of the organization, including leadership/management, employee characteristics, organizational design, decision processes, culture, technologies, etc. However, the proximal factors are also history-dependent and related to distal factors in the external context. The external context includes other organizations, factors at a societal level (e.g. governance, policies, laws), and not least, different institutional environments that “provide” new ideas, standards, solutions, etc. (Meyer & Rowan, 1977). Influences from contextual factors are indicated in the model as external influences.

Summing up: understanding policy–practice gaps as failures of mutual adaptation

Now, let us sum up the main arguments so far and in doing so, address the first research question of this paper. According to the framework presented above, organizational learning is conceptualized as an iterative organizational learning and improvement cycle. Policy–practice gaps are understood as a failure of adapting ongoing practices to prescribed policies and/or to adapt policies to the complexities of ongoing practices. That is, what we call a failure of mutual adaptation.

To avoid this type of failure would require that the two logics of deployment and discovery are balanced and work in conjunction. If this is not the case, due, for example, to a dominating preference for managing in accordance with a logic of deployment, there will be a focus on efficiency and control, creating a policy-driven, top-down type of situation and risk for driving out local adjustment and development processes. Conversely, a too strong emphasis on the logic of discovery, would open up for a bottom-up and practice-driven organization focusing on transformations of policies to be better aligned with the demands of practice, but with less concern for the efficient implementation of prevailing policies. Both these situations would increase the risk for misalignment and thus for gaps between policy and practice.

As argued above, a lack of balance between the logics of deployment and discovery may depend, among other things, on the cognitive framing of organizational issues by the management and leading actors in the organization (Edmondson, 2012). Specifically, it could be a matter of how issues of policy implementation and organizational learning are viewed and handled in the organization and, indeed, if there at all is time and space for such considerations. What we above have described as a space for interplay between organizational actors has critical functions to play for managing policy–practice gaps, not least by providing a forum for discussions concerning how to better align policies and practices in the organization. In the next section, we will offer a possible agenda for such discussions by proposing a set of principles for handling the dual challenge of organizational learning.

Managing the dual learning challenge

We now turn to the second of the two research questions stated in the introduction, that is, how to manage the mutual adaptation and alignment between policy and practice – what we have called the dual learning challenge – and, thereby, to reduce the risk for policy–practice gaps in the organization.

Five principles of organizational learning

One possible solution that has been suggested is to promote ambidextrous leadership (Rosing, Frese, & Bausch, 2011). The idea behind this notion is to switch between the two logics over time through an interplay between what we may call restrictive and enabling leadership behaviors (Wallo, Kock, Reineholm, & Ellström, 2021). That is, between behaviors that foster adaptive learning (e.g. through reduction of variance) and behaviors that foster developmental learning (e.g. encouraging experimentation).

Another solution suggested by Adler and Heckscher (2013; 2018) is to challenge the classical contrast between mechanistic/bureaucratic and organic/non-bureaucratic forms of organization, proposing an ambidextrous and collaborative form of organization. Their version of an ambidextrous organization builds on their previous works on post-bureaucratic organizations (Heckscher & Donnellon, 1994), and not the least on the notion of enabling bureaucracy (Adler & Borys, 1996). That is, a view of formalization and standardization in organizations as enabling and developmental rather than coercive.

If we adapt the reasoning of Adler and Heckscher (2013) to our purposes and terminology, an ambidextrous and collaborative form of organization can be contrasted to other organizational types using the four-fold matrix illustrated in Table 1 below. Considering the first left-right diagonal, a mechanistic type of organization that focuses on efficiency and control contrasts with an organic type of organization that promotes creativity and innovation. Along the other diagonal, the ambidextrous, collaborative type of organization is contrasted with a traditional organization focusing on stability, and with an informal and little developed work organization.

If we focus on the ambidextrous, collaborative type of organization, Adler and Heckscher (2013; 2018) argue that this type of organization has the potential to pursue ambidexterity, and thus simultaneously promote deployment/adaptive learning and discovery/developmental learning. To realize this potential, an ambidextrous organization design should focus on the following five key principles:

• A focus on a shared purpose and the joint contribution of organizational members to a common primary task. In operational terms, interventions for promoting developments in this direction could include the use of tools for multi-stakeholder strategic dialogues and/or tools for quality function deployment, and “Hoshin Kanri Planning” (Adler & Heckscher, 2018, p. 94).

• Interactive process management. Formalization of organizational processes and work methods, but in accordance with an enabling rather than coercive mode (Adler & Borys, 1996). That is, processes and methods are adjusted through teamwork and dialogue and continuously improved by those who are using them in their daily work (Adler & Heckscher, 2013, 2018; Björk et al., 2023).

• Participative centralization. Local initiatives and decision-making combined with strong overall coordination. High degree of participation and empowerment of employees. Participation is based on capacity to contribute to the common task rather than on formal position (Adler & Heckscher, 2013, 2018).

• Teamwork based on T-shaped skill profiles. Skill formation with a focus on cultivating T-shaped skills, that is, deep knowledge in one’s own specialty combined with a breadth of knowledge in related specialties. Such a common ground is seen as critical for collaborating with and learning from others (Adler & Heckscher, 2018, p. 98).

• Collaborative trust. As opposed to other forms of trust based on submission, charismatic leadership or administrative authority and financial incentives (Weber, 1978), collaborative trust (Adler & Heckscher, 2013) and trust-based management (Bringselius, 2018; 2019) are assumed to be more open and flexible, and based on the principles mentioned above, in particular, shared purpose and interactive process management. A basic commitment is to contribute to fulfilling the purpose and goals of the organization through institutionalized dialogue concerning how to develop the organization, including, among other things, roles and working procedures (Adler & Heckscher, 2013).

Using and creating knowledge is crucial in enabling effective interaction between the policy–practice dimensions and overcoming the anticipated dual challenge. This is primarily reflected in a combination of the principles of interactive process management and teamwork based on “T”-shaped skills. By ensuring the formalization of constructive processes and collaborative work methods that are enabling rather than coercive (Adler & Borys, 1996), it is argued that processes and methods can be adjusted through teamwork and dialogue and continuously be improved by those who are using them in their daily work (Adler & Heckscher, 2013, 2018). The combination of interactive work methods together with teamwork based on relevant skills not only supports evidence- and facts-based analysis and decision-making but also allows for constructively processing the inevitable variation and complexity experienced in policy and practice dimensions.

Next, an example is presented that illustrates how the proposed principles can be relevant to considerations on how to handle the dual learning challenge of accommodating both deployment and discovery in organizations. The example relates to a recent Swedish debate on trust-based governance and management in public sector organizations.

An illustrative example: trust-based governance

In 2016, the Swedish Government formed the Trust Delegation, comprising policy experts and researchers, to explore new governance approaches in the welfare sector. This initiative arose from the recognition that traditional governance – characterized by hierarchical, bureaucratic and competitive mechanisms – might not be the most effective way to deliver high-quality services and create positive work environments.

The Trust Delegation proposed a policy of trust-based governance as an alternative. This approach, defined as a focus on the organization’s purpose and user needs, emphasizes collaboration, a systems perspective and building trustful relationships (Bringselius, 2018; Hartman, 2018 cited in Bringselius, 2019). This idea has since been adopted – at least at a policy level – by various municipalities, health-care institutions and government organizations throughout Sweden.

At the heart of trust-based governance is the belief that public sector employees are driven to do meaningful work and help citizens. However, they often find themselves bogged down by policy mandates, such as extensive reporting and documentation, which tend to interfere with a focus on the primary tasks for which they are trained, and thereby hinder their effectiveness. Trust-based governance seeks to address this by finding a better balance between control and trusting relations, and that trust and collaboration should be combined with quality controls and transparency. The latter point is underlined also by Björk et al. (2023), arguing that a balance between reliable control and trust is essential for providing quality services and maintaining productive workplaces. However, in their critical review of the idea of trust-based governance, they also argue that trust should be a result of effective management and control, not a starting point.

Basically, the idea of trust-based governance emphasizes the purpose and primary task of the organization as a starting point for organizing activities. It also highlights the importance of understanding citizens’ experiences and aligning organizational efforts with their values (Adler & Heckscher, 2018; Bringselius, 2018). In line with this orientation, there is a focus on the core activities and the “realities” of day-to-day work practices alongside prescribed policy frameworks. By implication, this means a shift in focus, and possibly also a shift in power, from policy to practice. Dialogue and trust are proposed as a basis for the relations between the levels of policy and practice in an organization. Positive expectations and empowerment are assumed to lead to better performance than a traditional reliance on detailed directives (Bringselius, 2018).

Furthermore, it is argued that employees should be involved in policy processes and decision-making. Ideally, policy initiatives should be allowed to travel back and forth between organizational levels in what could best be described as interactive knowledge management. Teamwork and dialogue are seen as important means for establishing common understandings on tasks and missions between levels and actors in the organization. In a sense, then, to create conditions for policymakers and professionals to “share the same world” (Björk et al., 2023) (cf. Nonaka & Konno, 1998 notion of creating a “shared space for emerging relationships”).

Concluding remarks and implications for research and practice

The purpose of this paper was to further our understanding of policy–practice gaps in organizations from an organizational learning perspective. Two research questions were formulated. First, how to analyze policy–practice gaps in terms of organizational learning. Second, how to manage an organization to successfully handle and bridge policy–practice gaps.

In response to the first research question, it was argued that policy–practice gaps can be analyzed as an outcome of an iterative learning and improvement cycle (see Figure 1); specifically, as a failure of mutual adaptation. That is, a failure of adapting ongoing practices to prescribed policies and/or to adapt policies to the complexities of ongoing practices. In response to the second research question, we proposed five organizational design principles that, following Adler and Heckscher (2013; 2018), are assumed to characterize an ambidextrous, collaborative type of organization and which, if realized, may promote mutual adaptation through interplay between the two logics of deployment/adaptive learning and discovery/developmental learning.

Implications for theory and research

This paper’s main contribution lies in the development and discussion of a conceptual framework bridging the policy–practice gap and furthering a new understanding of how the dual challenge of accommodating discovery and deployment in organizations can be resolved (see Figure 1). Concerning policy–practice gaps, the analytic framework that was introduced has implications also for our understanding of organizational learning. First, rather than viewing organizational learning as primarily a policy-driven, top-down process or, conversely, as a practice-driven, bottom-up process (cf. Basten & Haamann, 2018; Crossan et al., 1999; Huber, 1991; Senge, 1990), we have argued for a cyclical view of organizational learning emphasizing processes of mutual adaptation and alignment between policies and ongoing practices. In this way, it is also possible to shed new light on the mediation between individual and organizational learning; specifically, how individual learning in and through ongoing practices in an organization may be translated into organizational learning and possibly manifested as changes in policies, routines, etc.

Second, in contrast to many conceptualizations of organizational learning as driven primarily by changes in the organization’s environment (cf. Dixon, 2017; March, 1991; Zollo & Winter, 2002), the framework proposed in this paper also emphasizes the important role that tensions and contradictions between, for example, routines and practices as officially prescribed (the policy dimension) versus as perceived and performed in practice (the practice dimension) may have as drivers of organizational learning. That is, tensions between what others have described as canonical and non-canonical practices (Brown & Duguid, 1991), or between work-as-imagined and work-as-done (Hollnagel, 2017).

Turning now to the empirical implications of the paper, we identify the following three areas where further research is needed. First, we believe that further research is needed to understand the institutional logics of the policy–practice dimensions that either enable or restrict the development of structures and agency for overcoming the dual challenge. For example, what are the institutional mechanisms that causes either teleological harmony or dialectical discord between policy and practice dimensions, thus affecting the ability to overcome the dual challenge in organizations? Second, there is a need to further our understanding of the role of leadership that is required to create a sustainable balance between discovery and deployment over time. Third, to better understand how the five proposed principles interrelate. For example, to investigate how the principles affect and influence each other and understand if all principles are equally important.

Implications for practice

The conceptualization of the policy–practice gap and principles for ambidexterity and organizational learning outlined above have apparent implications for the governance and management of organizations, especially in challenging the traditional hierarchical and bureaucratic principles of organizational design. Our illustrative example highlights some aspects of trust-based governance in current Swedish public services (Bringselius, 2018). By emphasizing organizational purpose and user needs and fostering collaboration and trust, this approach moved away from rigid policy adherence, focusing instead on the practical realities of professional work in the public sector. This Swedish case underscores the need for balancing deployment and discovery in organizations, marrying the principles of trust-based governance with reliable control to manage and govern effectively (Björk et al., 2023).

However, contextual influence and the natural evolution of organizations require different emphasis on discovery and deployment over time. Therefore, we also argue that at every given moment, one logic will, by necessity, almost always dominate the organizational learning system according to the prolific factors that are given by the specific context and situation. To effectively handle the dual challenge, it is therefore important to be aware that the balance between the two logics is achieved over time. Suppose a structure in line with our proposed framework is firmly in place. In that case, skillful management and leadership are considered critical elements to enable appropriate alternation between the two logics depending on the situation, thus maintaining a sustainable solution to the dual challenge. Specifically, we would argue that both an enabling and a restrictive leadership (Wallo et al., 2021) – that is, an ambidextrous leadership (Rosing et al., 2011) – are keys to be able to switch between the two logics of deployment and discovery and thereby handle the dual challenge.

The application of lean principles in the industrial sector further demonstrates the importance of balancing policy and practice, as outlined in our model. Originally from the Toyota Production System, lean focuses on process efficiency and waste elimination, requiring both a top-down policy initiative for deployment and a bottom-up, practical approach for effective implementation (Wallo, Martin, Sparrhoff, & Kock, 2022). While policy sets the strategic direction, the success of lean largely depends on the engagement and insights of frontline employees in their daily work. This approach illustrates the need for a balance between deployment and discovery, as emphasized in our model. The implementation of lean in the industrial sector highlights the importance of not just directing employees through policy but also empowering them to contribute to the organizational learning process, ensuring a harmonious blend of strategic alignment and operational excellence.

Our model focuses on balancing top-down deployment with bottom-up discovery and is exemplified in Sweden’s trust-based governance and the industrial application of lean principles. These cases illustrate the importance of aligning policy with practical realities, highlighting the need for both strategic guidance and practical engagement in organizational learning. This approach ensures more effective and adaptable organizational processes.

Figures

Figure 1.

A model conceptualizing policy–practice gaps as an outcome of an iterative organizational learning and improvement cycle

Table 1.

A matrix model of the logics of discovery and deployment (adapted from Adler & Heckscher, 2013. p. 37)

The relative strength of…		…the logic of discovery
		Low	High
…the logic of deployment	High	Efficiency, control	Ambidextrous, collaborative
	Low	Stable, traditional	Innovative, organic