Towards a Practice-Based Theory of Governance Learning and Institutionalization: A Cross-Case Analysis

Abstract

In this chapter an overarching interpretation of the cases is presented. It departs from pre-existent theoretical perspectives of institutional change and organizational learning. These perspectives are used to make sense out of empirical material. The author searches for significant differences between the cases. Why were public officials able to critically reflect on their practices and eventually change the way of governing in some cases, and failed to do so in other cases? How is it possible that in some cases governance ended up as an institution reduced to a quasi opinion survey, or even worse, lead to discrediting of participatory approaches, while in other cases governance was institutionalized in the form of a relatively large infrastructure of institutions enabling public engagement in various decision-making processes?

An open form of inquiry, combining pre-existent theoretical perspectives with detailed investigation and cross-case comparison of real-life public management processes resulted in the development of a holistic practice-based framework. The framework illustrates learning of a new governance mode and the subsequent institutional change. The analysis is grounded in the new institutionalism in organizational theory and supplemented with organizational learning concepts and new theoretical insights emerging from research. Insights are organized as theoretical propositions and definitions of new theoretical concepts.

Building the linkage between micro , meso and macro levels of analysis is, among others, a major contribution to the existent body of literature on public management, governance, and governance learning. Governance practice is perceived here as a combination of top-down environmental pressures and bottom-up creative responses which, following an initial phase of relative diversity, gradually become organized into more predictive patterns. The analysis that shifts between structure (formal and informal rules regulating behaviour) and agency (individual strategic actions) enables us to capture phenomena central to the process of institutional change and learning in public administration, that is, (1) the significance of original institutional environment into which change is introduced; (2) the role of individuals in particular leaders confronted with pressures and demands for a change; and (3) institutionalization processes, within which the new mode of governance is given a specific form and is henceforth taken for granted.

The analysis of cases resulted in the development of theoretical concepts within each of these three areas. In the first area (significance of original institutional environment), it is argued that coercive pressures (from legal sources) for governance are not likely to trigger learning, unless they are complemented with mimetic and normative pressures, that is, pressures stemming from uncertainty and prompting organizations to copy what has proven effective, as well as pressures rooted in approaches and orientations of professional groups. Also within the first area related to the original institutional environment, the concept of governance void is introduced. The term draws on the notion of institutional void (Mair et al. 2012) and is understood as a space where institutional infrastructure supporting collaborative forms of governance is absent or weak, for example, there are no ready-to-use tools and procedures coordinating multilevel or multijurisdictional work or public engagement in decision-making processes.¹ Polish institutional context characterized by the dominance of hierarchical relationships and high deficits of social trust provides a useful example of governance void.

Voids become apparent to actors on the ground while they are trying to manage a public dispute, a phenomenon disrupting their routine practices. It happens when public officials discover that standard, traditional ways of proceeding do not get results. The resistance or “backtalk” (Schön 1983, 1987; Yanow and Tsoukas 2009) of material and human environment causes a range of emotional and cognitive reactions like fear, puzzlement, and frustration; for comparison see John Forester’s work on practice of planning and conflict resolution (Forester 2009, 2013). This is where the role of individuals and leaders is explored. Public officials’ reactions to disruption are analysed through the concept of reflexive practice and the phenomenon of surprise (Schön 1983, 1987; Yanow and Tsoukas 2009) , closely linked to the organizational learning perspective. Following an analysis of surprise phenomenon in a collective setting, I suggest that surprise can be understood as a cognitive state caused by a disruption of institutionalized patterns of thinking and behaviour deployed by a (public) organization in order to deal with a specific (social) problem.² Surprise is regarded as a prerequisite of governance learning, a trigger for innovation with new roles and new procedures in managing the dispute and the social problem behind it (Rządca and Strumińska-Kutra 2016) . Yanow and Tsoukas’ theorizing of practitioners’ response to surprise is supplemented with withdrawal from problem-solving. This type of response is likely to emerge when the community of practitioners and experts is homogenous in their professional orientations, which here are either hierarchical, quasi-market, or New Public Management-like approaches to governance. This kind of homogeneity reinforces the language of certainty (Yanow 2009) about the way in which to manage a certain problem, inhibits learning based on critical inquiry into practice, and limits motivation to improvise and experiment.

Further on, through reference to concepts of institutional entrepreneurship and institutional work, it is shown how double-loop learning and subsequently institutional change are triggered by individuals. For some individuals, governance voids and top-down institutional pressures create an opportunity to follow their values and expand their resources. The analysis illustrates how public organization leaders, as well as other actors representing public agencies resort to diverse institutional arrangements and use bits and pieces from local organizational, political, and community spheres to fill governance voids. I further describe how interactions of diverse actors and their relationships (e.g. low levels of social trust) result in specific patterns and forms of governance institutionalization.

In this chapter, the interpretation of cases is organized as follows. It starts with examining the role environmental pressures played in learning and in the institutionalization of a new governance mode within public agencies under consideration. Using the analytical tools of institutional perspective, I investigate the significance of changes in international and national milieus (e.g. the World Bank or the European Union, national government and its experts), which previously insisted on the economic effectiveness of public services and currently tend to promote inclusiveness and “good governance” (Kordasiewicz and Sadura 2017) . I observe how these changes are translated into the level of local governance practices through legal acts (coercive pressures), educational institutions shaping knowledge and attitudes of professionals (normative pressures), and through observing the examples of other public administration entities enacting public governance rules (mimetic pressures).

Then I link these pressures to the micro-level individual learning processes. The analysis illustrates how public administration workers couple their practices with environmental pressures, how they learn to practice governance and how the type of learning adopted by an individual is grounded in diverse configurations of institutional pressures. It is argued that workers use diverse institutional logics and local meanings creatively, depending on their professional commitments, location within the organizational structure, personal interests, power position, and interactional, on-the-ground decision-making (Binder 2007).

In the next phase of the analysis, the attention shifts from individuals per se to their interactions, within which new approaches to governance are negotiated and become gradually grounded in professional practice, organizational procedures and structures, as well as legal provisions in place. Following the ‘inhabited institutions’ approach, I picture the meso-level organizational order as emerging from social interactions, and organizations as places where people and groups make sense of, and interpret institutional vocabularies (Binder 2007).

Throughout the analysis, the meso-organizational level is the focus of attention. Yet, the analytical practice could be compared to alternating processes of zooming in and zooming out. For the analysis, it is crucial that organizations are embedded in a specific field where new governance practices are shaped in the process of interactions with external stakeholders: citizens, politicians, community representatives, NGOs, businesses, and other public agencies operating at different administration levels. All actors creatively enact new and old institutions that constitute a given field. Therefore, I begin by tracing environmental influences on the organization. The process of gradually zooming in leads to individuals and their everyday practices. I then zoom out in an attempt to capture the institutionalization process within which an activity that belongs to the individual level (such as learning) is transformed to reach the meso level through social interactions and development of shared understanding and practices.

8.1 From Legal Regulations Enabling Governance to the Practice of Governance: The Complementary Nature of Environmental Pressures for Governance

In Chap. 2, I argue that pursuant to empirical research, the governance turn is observed in many Western and CEE countries (Denters 2011) and that this simultaneous co-evolution of similar patterns of rule in public administration can be perceived as a process of institutional isomorphism (DiMaggio and Powell 1983) . The new institutionalism approach identifies general environmental factors influencing similarities in changes adopted on the organizational level. These factors, influencing change and homogenization, are grouped into three types of institutional pressures towards isomorphism : coercive, mimetic, and normative . Pressures are vehicles through which the disjunction between the values held by society and the behaviour of an organization is erased. In other words, within the perspective of new institutionalism, changes in patterns of rule can be interpreted as a result of growing expectations for more responsive and inclusive public governance. In the following considerations, I investigate the impact of these pressures on governance learning processes.

Table 8.1 presents a general overview of cases with the use of four categories central to the overall analysis: inclusiveness and effectiveness of the public management process, presence of learning, presence of different types of pressures, and institutionalization of a new collaborative governance mode. As the analysis will demonstrate, the sole existence of coercive pressure for new governance modes does not guarantee an effective learning of new governance practices. Absence of mimetic and normative pressures effectively blocks learning processes and—in the long run—hinders institutionalization of new public governance as a practice implementing effectiveness and inclusiveness. When positive examples in the organizational field are lacking (mimetic pressures), and there are no professionals who would be (at least partially) socialized within the governance framework (normative pressures), the new governance mode drifts into the direction of already established patterns of hierarchical or quasi-market patterns of rule. The tendency is illustrated by the arrows in the table (see Table 8.1).

Table 8.1 Institutional pressures, learning, and institutionalization of governance. Cross-case comparison

The first two categories included in the table relate to the value-based promise of public governance to make public management both effective and inclusive (democratic). By using simple indicators combining objective and subjective aspects, I assess the effectiveness and inclusiveness of public dispute resolution and accompanying attempts to solve the social problem that forms the backdrop of the dispute. Effectiveness is assessed on the basis of whether a public service and/or product were eventually delivered and through measuring the overall satisfaction of the parties involved. Low satisfaction means that none of the parties was satisfied with the solution. This was a case of anti-flood facilities, where all stakeholders, including public officials, perceived the process and its outcomes as a complete fiasco. Medium satisfaction means that the degree of satisfaction varies between parties. Examples are the cases of schools and the WWTP, where public officials were satisfied, but some of the stakeholders felt harsh disappointment with the way the process was handled and with its outcomes. High satisfaction means the contentment of all parties involved. This was the case of the dispute over the location of the market. The inclusiveness of the process was assessed on the basis of two dimensions. The first is involvement of stakeholders in decision-making (e.g. inviting them to meetings, seeking their opinion); the second dimension is about openness of the process towards the inclusion of the stakeholders’ perspective and their voice in relation to the definition of the problem and the development of a solution. In all cases, stakeholders took part in the decision-making process. However, in most cases, the stakeholders’ views neither changed the way in which the problem was perceived by public officials, nor did their perspective matter for solutions eventually proposed within the process. The exception was the market case, where merchants and local community representatives were eventually invited to join a collaborative process of working out a solution development.

Learning is the next category of utmost importance for the overall analysis. Only the double-loop learning is taken into account in the table, as it combines critical reflection with experimenting and enables the development and design of governance patterns that are both new and most suitable for a given policy problem. The presence of double-loop learning is hence a precondition of good (reflexive ) governance, which is understood here as the capacity to reflect on, and rebalance the mix among modes in response to changes in terms of challenges and/or opportunities that exist at the interface of market, state, and civil society (Jessop 2011, see Chap. 2). Double-loop learning is considered as present in cases where public officials responsible for managing a specific problem bear the traits of reflexive practitioners, that is, practitioners willing to reflect on one’s action and change its objectives or strategies in order to adopt to (re)emerging governance problems, for example, disagreement of an unexpected form and/or intensiveness.

The latter category explicated in the table is linked to institutionalization, a process within which a structure becomes to be taken for granted by members of a social group as efficacious and necessary; thus it serves as an important causal source of stable patterns of behaviour (Tolbert and Zucker 1996 ; Surachaikulwattana and Philipps 2017) . Each case presents dispute as embedded within a rich history of events taking place before and after the most intense phase of the conflict. The time span covered by the investigation ranges from 6 to 15 years (see the methodological appendix). During this period, collaborative approaches to public management went a long way from relatively new, often contested practices requiring exploration and experimenting to relatively well-structured and obvious practices.

Yet these cases differ significantly in terms of the ‘final’³ form of governance approaches adopted. In the rural municipality (first case), collaborative policy making practices are reduced to something resembling an opinion survey. In the two subsequent cases, that is, disputes over the location of a market and that of a WWTP (both cases from the same large city), the idea of participatory approaches to public management became relatively well established. Public administration is equipped with regulatory and organizational procedures enabling top-down and bottom-up processes of spatial planning, including consultation about diverse investments, especially those with a high environmental impact, budget planning and conflict resolution. Yet the horizontal coordination, linking different departments and agencies, still relies more on individual social skills of public managers responsible for a given issue than on ‘ready-to-wear’ organizational solutions, for example, established patterns of communication, explicit inclusion of the horizontal cooperation requirement into the scope of employee duties, and incentives for such cooperation. In the latter case—the dispute over the anti-flood facility—horizontal coordination was a necessity, yet the agency responsible for the investment failed to establish it. Participatory practices were framed, and subsequently institutionalized, as a top-down ruling mode, that is, education about already adopted solutions. Public officials’ persistent inability to respond to the expectations and anxieties of the public, as well as their failure to manage the consultation process, led to the rejection and degradation of the idea of collaborative approaches. Public officials perceive these approaches as exposing public management to the risk of demagogy, while in the eyes of the local community they are a smoke screen enabling public agencies to pretend they care about citizens’ needs and opinions.

The analysis below focuses on the relationship between institutional pressures and learning. Patterns of learning and institutionalization are subject of analysis in subsequent parts of the chapter.

8.1.1 How Governance Learning Needs Both Force and Enhancement

To a varying degree, coercive pressure , originating from newly introduced legal regulations, was detectable in cases under consideration. In all of them public officials were required to inform, consult, or involve external stakeholders in the decision-making processes or in the implementation of policies. In each of these cases, officials responsible for the process were taking these regulations into account. Yet in each case, they were accused of violating some regulation, of advancing false interpretations, and more generally of ignoring the perspective, values, and interests of the public. This kind of accusation became part of public disputes. Under the pressure of protests regarding goals and/or means of public management, public administration officials experienced uncertainty. They realized that something could impede or even prevent resolution based on old ways and procedures. Bureaucrats and elected officials were facing the challenge of proposing and learning new ways of organizing decision-making processes.

In two cases no double-loop learning processes were observed, that is, cases were solutions proposed by public officials were not able to break through established, mainly hierarchical, approaches and to propose substantially new ways of developing solutions, based on collaborative and participatory approaches. In the case of schools, the Mayor turned regulations potentially promoting collaborative forms of governance into a ‘business-as-usual’, top-down management process focused around the efficiency of services provided and consistency with legal regulations. Strong emphasis on the latter translated the idea of consultation into the ritual practice of ticking the boxes (meeting with trade unions—check. With school principals—check, getting the approval of the educational bureau approval—check, parents, community—not mentioned in the regulation, they can therefore be informed about the solution later on). In other words, the original institutional context regulating daily operations of public officials penetrated the learning process by delivering cognitive and normative guidelines on how to adopt a new governance mode. In the case of schools, pressures for inclusion were not intense. Formal institutional structures of local government promoting a strong executive, to the detriment of the legislative, reinforced the dominant position of the Mayor and limited other actors’ possibilities to mobilize around the development of alternative interpretations. This configuration of relatively weak pressures for alternatives and institutional structures maintaining power asymmetries enabled a dominant actor to successfully push out dissenting voices and turn the process into a linear procedure where everything is known and planned beforehand. Within such a process, opportunities for learning are scarce, because there is no place for questioning assumptions, bringing alternative definitions, and diverse types of knowledge.

One could assume that learning opportunities would be better if only pressures for participation were greater and power asymmetry less striking. However, the second case, where the double-loop learning was also lacking, forces us to question this hypothesis. Protests over flood-prevention infrastructure were intense, took years and ended up in a spectacular fiasco, namely public agencies’ withdrawal from investments and leaving the entire area at a high risk of inundation. Yet, throughout the long-lasting conflict, public agency was unable to design new approaches to governance with a view to effectively respond to social discontent and mistrust. Public managers’ attempts to overcome resistance towards flood barriers were based on a single-loop learning mechanism. The goal they pursued through their actions was neither changed nor questioned from the very start and is aptly expressed in the following statement made by the director responsible for the process: “the goal was to show that what we proposed was right” (interview, October 2013). Attempts to improve were therefore focused on manners in which to communicate more effectively and educate people about plans to build anti-flood barriers. First, the community was informed about the project in a letter. Information was detailed, technical and—according to officials—transparent. It was not received well by the local community. In response to the letter, several public meetings with officials were organized. When face-to-face communication failed, officials decided to invite respected individuals equipped with diverse forms of authority: a university professor as a representative of the academic community, the charismatic and respected marshal of the region, and experts, who had prepared the project—officials believed they were compelling due to their professionalism and academic degrees held. None of these improvements brought expected results—quite the opposite. The level of anger and frustration among the protesters was rising. At this point, officials were clueless and eventually decided to withdraw from the process, leaving both the dispute and the problem of flood prevention unsolved. Moreover, those responsible for the investment process still believe that they either failed in their attempt to convince the public that barriers were needed, or that the public is immune to rational arguments and the process was doomed from the start. Yet, as I argue in the case analysis, for those opposing the investment, the discussion should not have related only to barriers, but also the entire flood-prevention system and public agencies’ unreliable performance in the past. These perspectives could not converge in a constructive discussion without redefining the problem, and therefore redefining the goals of public management in this particular case. All those involved in the process knew something should have been done differently. But how? Not knowing what to do, those responsible for the investment where first running in circles of single-loop learning, and eventually stepped back.

It is not enough to prescribe legally what needs to be done; there need to be people who understand and/or share original values behind legal guidelines, as well as practical examples of such new practices within the organizational field. These two conditions correspond to two remaining types of institutional pressures, that is, normative and mimetic. Their meaning and role in the learning processes are explained through the analysis of cases where double-loop learning eventually took place.

Initial stages of dispute in the case of market location and WWTP followed a similar single-loop learning path. In the case of the market, city authorities tried, over and over again, to win over the communities and municipalities by inviting them to discuss the project whose basic assumptions were non-negotiable. Eventually, the process slowed down and ground to a halt. For a period of almost two years, city administration took no action. In the case of WWTP , authorities strove to inform the public about the project, just as in the case of flood-prevention facilities. While learning new practices of participatory governing, they were still using old hierarchical reasoning. Stakeholders were invited to participate so that they could be convinced about the validity of the proposed solutions, not to contribute to them.

In both cases (market and WWTP) findings suggest that the goal of learning did not need to be attained in order to ensure excellence in the delivery of public goods and services; rather, it was necessary to gain new tools and arguments in the struggle over regaining legitimacy and over control in a policy domain. Gilardi and Radaelli (2012) call learning for legitimacy, which does not improve policy performance, ‘symbolic learning’, while learning for maintaining control is referred to as ‘political learning’. In these two cases, both forms of learning were observed.

The breakthrough was possible, because normative and mimetic pressures complemented coercive pressures. Crucially, in both cases in which double-loop learning took place, professionals whose approaches and orientations were, at least partially, formed by collaborative and participatory governance ideas. They were both employees of public administration and external experts. In a dispute over the market location, a central role was played by the new Vice President, whose appearance brought an end to a two-year impasse and initiated a new, collaborative process of market planning. In addition, the presence of a number of other important figures brought to light the participatory logic in the dispute. One of them was the municipality Mayor, a tireless, though sometimes counterproductive, supporter of public participation.⁴ There were researchers from the Polish Sociological Association, who experimented with mediation in order to solve the dispute. In the case of WWTP , the most intense and long dispute among those under analysis, there were two directors of city administration departments and the director of a regional agency lobbying for participatory approaches, as well as a professional mediator eventually employed by the investor. Importantly, all individuals mentioned, with the exception of the Vice President in the market case, failed in their attempts to solve the dispute. Their strategic actions were however significant in the long run, because they paved the way for the institutionalization of participatory approaches to governance (the issue explored in the following sections).

All of these individuals were ‘carriers’ of normative pressures towards a new participatory governance mode percolating into various institutions training public administration employees and those in related professions. Some of them participated in externally funded workshops organized at the local, national, and international level devoted to public dispute resolution, participatory planning, and public consultation. Others were relatively fresh graduates of social sciences (e.g. political sciences, sociology, social policy), or had done post-graduate studies in public management, where they became familiar with various concepts of participatory democracy.

Membership in various, often cross-sectoral professional networks exposed them, as well as other public officials, to mimetic pressures. When talking about the gradual dissemination of participatory practice, a director within the city administration said

There is a competition between local governments. We have this association called the Union of Polish Cities. Representatives of local administration meet there and exchange their experiences. Then people come back and say “What an interesting thing they have done there! We are going to do the same, only better”. (Social Communication Department, interview, June 2013)

Being embedded in communities of practitioners, permeated at least to certain degree with ideas of participatory democracy, alternative dispute resolutions, collaborative planning and the like, nurtures individuals with practical ideas about ways in which to implement regulations and solve emerging problems. Finally yet importantly, it incentivises and legitimizes experimentation.

While coercive pressures derive from legal sources, normative and mimetic pressures are rooted in diverse educational institutions (e.g. university curricula), research and policy programmes (e.g. financed by the EU or the European Economic Area grants), professional networks, and communities (e.g. local governments, partnerships, associations). It is an institutional infrastructure supporting the creation of leaders skilful in managing multi-stakeholder collaborations for public goals (Sørensen and Torfing 2015), educating new professionals or preparing those who have exerted the profession of a mediator or facilitator to support the transformation of governance.

Taking the above into account, how do the three institutional pressures influence governance learning processes? Coercive pressures legally bind public officials to include participatory tools into certain types of decision-making processes. New regulations challenge them to do things differently and potentially trigger learning process. The process of learning is accelerated through indirect coercive pressures, namely: regulations empower stakeholders to legitimately claim their right to be heard and involved in governance processes. The indirect pressure exerted by empowered stakeholders accelerates the learning process, because it helps public officials to identify a given situation as potentially requiring the adoption of a new governance approach. Protests and claims come across as a surprise (Schön 1983)—a routine-breaking phenomenon.

The following figure illustrates the relationship between institutional pressures and governance learning. Figure’s nested, Russian doll-like structure suggests that public organization is embedded in a number of larger fields; here, within a local field where other public and non-public organizations interact following institutionalized rules of the game, and larger, national and international field with its own actors, for example, governments, international organizations and the like. The figure illustrates a situation where all pressures making learning possible are present. In the following sections, the nested structure will be explored: we will open up/dissect an organization and look at it as a ‘nest’ and an immediate context for collective action, as the subsequent services and departments provide a ‘nest’ for individuals (Fig. 8.1).

Fig. 8.1

Institutional pressures and governance learning

To sum up, legal regulations introducing collaborative patterns of governance in public administration create direct and indirect pressure on local practices. Direct pressures mean that public officials are legally bound to include certain steps into decision-making processes. In terms of learning, indirect pressures seem to be far more important. Legal changes towards governance create legitimate pressure groups, who can claim their right to be involved in governance. The more powerful the pressure groups are, the greater the probability that their alternative perspectives and knowledge will open up space for critical reflection on governance. Yet a coercive pressure, even if accelerated by a dispute potentially triggering critical reflection, does not suffice to push the learning process forward. Therefore, I propose:

Proposition 1

Coercive pressures trigger governance learning processes and provide reasons for governance learning, but as such do not guarantee double-loop learning.

In all analysed cases, initial stages of public dispute were characterized by single-loop learning mechanisms at best. Yet, the change in practices was accompanied by old ways of thinking and/or directed rather at maintaining control over policy issues and not necessarily at improving performance. A qualitative change in approach was possible, because coercive pressures were accompanied by mimetic and normative pressures that delivered instruction defining how things could be done. It is through these pressures that officials learn to think differently.

Proposition 2

Mimetic and normative pressures push learning processes forward through delivering examples of new practices in the organizational field and by shaping professional orientations and attitudes consistent with the new governance mode.

Proposition 3

Without the support of normative and mimetic pressures, governance learning tends to be performed as ‘fake learning’ (political, symbolic learning) that is, learning oriented not towards gaining excellence in the delivery of public goods and services, but towards regaining legitimacy and maintaining control over a given policy field.

8.2 Between Structure and Agency: On the Significance of the Institutional Context for Individual Responses to Disruption

As argued in Chap. 3, the explanation of the governance turn and learning based on Powell and DiMaggio’s framework of institutional pressures is relevant mainly on the macro and meso levels of analysis. However, it fails to take into account the micro level of individual practices and perceptions. It also ignores the process of transforming the micro-level answers into meso- and macro-level structures, which is of key importance for the analysis of organizational learning and institutionalization processes. This is why the top-down approach of new institutionalism is complemented here with the bottom-up approach focusing on the role of actors in creating, maintaining, transforming, and disrupting institutions.

First, I propose to interpret the phenomenon of public dispute from the perspective of professional and reflexive practice (Yanow and Tsoukas 2009; Sandberg and Tsoukas 2011) . Disputes have the potential of delivering a breakdown of a default activity, thus far taken for granted, and illuminate the logic behind it . Disagreement forces professionals to reflect on their own practices and to improvise responses (Yanow and Tsoukas 2009; Laws and Forester 2015) . I am interested in extending the concept of reflexive practice into a collective setting by investigating how original patterns of action and responses to a disruption can be linked to normative and cognitive pillars of institutions. What happens when responses enter into interaction with the environment?

Yanow and Tsoukas distinguish different types of responses to disturbances interrupting routine practice (2009). Routine practice—marked by an absence of disturbances (surprises)—is characterized by absorbed coping, that is, performing a set of established activities based on tacit, experiential knowledge. Routine is not mindless, yet it is too close and too familiar to compel attention. Developing Donald Schöns’ concept of surprise (Schön 1983, 1987) and Dreyfus’ model of knowledge-based work (Dreyfus 1991), Yanow and Tsoukas argue that there are different kinds of disturbances (surprises), ranging from ‘malfunction’, to ‘temporary breakdown’, to ‘total breakdown’. Each one elicits a different type of improvisational response, ranging from ‘non-deliberate’ (spontaneous readjustments), to ‘deliberate’, to ‘thematic’ (explicitly intentional). When routine is disrupted by malfunction (first type of disturbance), the practitioner shifts her attention to what she is doing, corrects the action and returns to routine practice. This kind of response is called reconstituted absorbed copying. The second type of disturbance, that is, a temporary breakdown results in deliberate coping: the practitioner is now paying careful attention to the task at hand. An involved deliberation is an alternative response to temporary breakdown and entails a more focused consideration of what is being done. Practitioner “stops and considers what is going on and plans what to do, all in a context of involved activity” (Dreyfus 1991, p. 72). Difference between deliberate coping and involved deliberation is a matter of degree, not of quality. Both responses still take place against the background of absorption in the world and, in this sense, are not a detached cognitive reflection (Yanow and Tsoukas 2009, p. 1352). At this point, however, the availability of tools enabling practice becomes questionable and a subject of reflection. Total breakdown is the third type of surprise. It results in a response called thematic intentionality, a more analytic or theoretical reflection. This kind of response enters the scene when “involved deliberation is no longer effective or operative, we need to move to a detached analytic (cognitive, theoretical) stance on the problem as we try to comprehend the underlying mechanisms involved” (2009: 1352).

The first three types of responses (reconstituted absorbed copying, involved deliberation, detached intentionality) correspond to single-loop learning. The last type (thematic intentionality) resonates with the double-loop learning concept, because analytical reflection is needed in order to uncover hidden assumption of own practice, question them and design new goals and means of practice. How does this theoretical framing help us to understand responses to public protests, institutional change, and learning processes accompanying the ‘governance turn’? And, how can this framing be modified in order to capture the structuralist moments of responses, that is, moments where improvised responses are infiltrated by institutionalized patterns of thinking and acting?

8.2.1 Entering Single-Loop Learning

Let us try to classify public officials’ responses to protests according to Yanow and Tsoukas’ typology ranging from spontaneous readjustments to detached, analytical reflection. Already at the very beginning of the analysis, it turns out that a collective setting emphasizes an important quality of surprise, that is, its openness to social construction processes. What might be surprising for some is not a surprise for others. The observation is not new; Yanow and Tsoukas admit that surprise “requires a degree of mindful openness or ‘permeability’ that enables perceiving an ‘event’ as surprising” (Yanow and Tsoukas 2009). Here I argue that mindful openness can be influenced by institutions.

In cases under consideration, a continuum of responses is opened by a situation in which a protest—a phenomenon perceived in the research as the incarnation of disruption—was not perceived as a disruption by actors in the field/the stakeholders. When deciding about the delegation of schools, the Mayor and her supporters were not surprised by the protests and they proceeded without taking the dissenting voices into account. The use of institutionalized (routine) patterns of rules grounded in the concepts of representative democracy, New Public Management, and bureaucratic rule of law enabled them to ‘naturalize’ protests as a normal reaction to an unpopular policy decision. One of the councillors, supporting the Mayors’ strategy, stated in an interview:

We were able to work out an optimal solution. Currently, the general mood here is not as bad as you could expect. The municipality had anticipated a far worse situation, including calls for a referendum or for the dismissal of the local government. (June 2013)

The dispute forced the Mayor to articulate the logic behind the practice, but not to change the practice itself. Actors questioning goals and means of governance were able to neither mobilize wider support nor refer to an alternative framing of the decision-making process. In fact, when questioning the adopted patterns of management, dissidents tried to use the logic imposed by the Mayor, and hence their attempts were doomed to failure.

Proposition 4

Since old (original) institutions penetrate practitioners’ thinking and acting, they can also prevent practitioners from recognizing a phenomenon as a surprise, and hence prevent the learning of a new governance mode.

In the remaining cases (market, WWTP , barrier), at first, practitioners regarded protests as a malfunction. Initial reactions of authorities followed the strategy of ‘more of the same’ (single-loop learning and reconstituted copying) or, in other words, the implementation of minor changes. They used the bureaucratic logic based on the rule of law to legitimize the exclusion of stakeholders’ from the decision-making process and to prove that all measures required by the law had been taken. That is, all informing procedures were conducted and the comments of those entitled to be party in the administrative process were answered. Despite these attempts, protests endured. Importantly, in all four cases protesters used a framework of different institutional logics to interpret the same legal and procedural rules. While officials focused on following the letter of law, protesters reasoned in terms of collaborative and participatory logic.

Gradually, disruptions started to present themselves to practitioners as a temporary breakdown. Officials enacted involved deliberation by paying more and more attention to their actions, which they modified in order to accommodate at least some of the protesters’ voices. Still, new practices were permeated with old institutional logics, for example, in the cases of WWTP and barriers, where decision-makers perceived consultation as education. Involved deliberation also included an instrumental use of participatory practices to restore legitimacy, for instance in the WWTP case, where the President of the city first ordered an administrator trial, opening a public discussion about the project and then issued a building permission an unchanged form, or in the market case, where officials enabled the consultation of the project whose very existence was contested.

Proposition 5

Initial responses to a surprise tend to follow adjustments prompted by the most accessible, dominating institutional logic regulating the practice of governance. Modified and new practices are pervaded by original patterns of thinking and acting.

Over time, enduring exposure to conflict and exposure to the alternative definitions of the problem made practitioners aware of the fact that tools they had at their disposal might be inadequate and would not let them solve the problem. Detached cognitive reflection and a total breakdown seem to be at arm’s length. Since an involved deliberation is no longer operative, the only remaining option is experimenting with new approaches. Or is it? Analysis shows that the logic of a logician is not the logic of a practitioner (Bourdieu 1990; Sandberg and Tsoukas 2011).

The case of barriers and—to a certain extent—also the case of market location, are examples of a situation where, faced with the inability to solve the problem, practitioners decide to step back and redirect their attention towards other problems. The complexity of the context in which they function (Weick 2003) makes this option feasible. In the case of the market, the first Vice President did not risk any significant losses in terms of legitimacy when he decided to abstain from action for almost two years. Merchants and local community representatives were just one of many groups involved in the many decision-making processes he had supervised over the years. To a lesser extent, the same holds true in the case of officials involved in the barrier-planning process, since their accountability for this particular exercise was clearer. Yet they had been quite successful in playing the “blame game” (Hood 2011). Through attributing the blame for failure to protesters, and by insisting that dredging works demanded by the inhabitants remained outside of agencies’ competence, they could ‘save the face’ while withdrawing from the investment.

A closer look at withdrawal reactions suggests that they can be interpreted as reactions to uncertainty, which is linked to the phenomenon of surprise. Uncertainty is caused by the conviction that all ready-made procedures (patterns) that may be conducive to solving the problem, as well as their modified variants, have already been exhausted. A practitioner cannot think of any other ways in which the problem could be solved under given circumstances, or they feel they lack the necessary competence, understood both formally and individually, to further engage in the process. Conflict itself further contributes to this kind of uncertainty, as it involves diverse stakeholders with divergent interests and values, whose actions are, to a large extent, unpredictable and call for improvised answers (Laws and Forester 2015) . It is a situation radically different from the relatively high predictability guaranteed by designing the process pursuant to a legalist approach. At this point, emotional responses come to the surface, including reactions of anxiety and fear. It is particularly evident in the case of market location, where fear became a significant category organizing the narrative about old and new public management strategies. Describing the governance practices enacted by city authorities in the initial phases of the conflict, the Mayor of the district states: “City authorities do not talk to people. They are afraid of people (…) I can see fear in their eyes when something like this happens: protests, openly expressed dissatisfaction of certain groups” (March 2013). The new Vice President, who eventually solved the conflict, reflects “there is a moment where you just need to accept the challenge and not be afraid to engage in a dialogue” (June 2013). A mediator involved in managing the dispute remarks: “many of these meetings are very unpleasant or even endangering. This experience accumulates and creates fear among the representatives of authorities. They fear direct interactions with inhabitants and avoid them” (January 2013). The street-level bureaucrat who was responsible for organizing the temporary relocation of the market mentions: “The new Vice President comes to the market himself! He is not afraid. You would not see the former Vice President there! (laughs)” (September 2013).

The exploration of interlinkages between emotional and cognitive dimensions of institutions is important for understanding the reaction of withdrawal. Empirical data delivers an illustration of how efforts to challenge established expectations of governance practices “threaten the individual’s sense of security” (Powell 1991, p. 194). The above considerations are summarized in the following theoretical proposition:

Proposition 6

When gradually exhausting responses based on deliberate copying and involved deliberation, practitioners (public officials) face uncertainty of the decision-making context and, consequently, display emotional reactions of anxiety. A possible reaction is withdrawal instead of engagement into thematic intentionality.

Further on, I will argue that emotional reactions to uncertainty, including reactions of anxiety and fear are the function of an institutional void. After trying out all possibilities that appeared feasible and legitimate to them, public officials find themselves in an almost physically experienced emptiness. There are no other ready tools or procedures coordinating multilevel or multijurisdictional work, or tools coordinating public engagement in decision-making processes. In cases under analysis, which refers to the Polish institutional environment, a collaborative, network-based logic, which could inspire and justify further modifications of tools, is far from dominating in any arena of public life, be it a workplace, a local community, or the political sphere (see Chap. 3). Hence, this logic is not at hand, it is not built into the “habits of the heart” (Bellah et al. 1985) or into long-lasting experiences of social actors (Putnam et al. 1993) and, as a result, does not provide much inspiration. In order to practice collaborative forms of governance in these cases, public officials would have to take a leap of faith, start to construct new tools or implement those already in use, yet according to a different logic. Initiating new practices or redefining old ones would mean improvise, performing “bricolage” or “making do” with whatever is at hand (Mair and Marti 2009; Mair et al. 2012; Forester 2009, 2013; Laws and Forester 2015) . It would involve critical reflection and double-loop learning. Practitioners engaging in this kind of activities would work in and around governance voids in order to eventually close them. Not all practitioners are ready to do this. When faced with a void, many of them step back. In Yanow and Tsoukas’ terms, the key question is: what makes practitioners willing to perceive a disturbance as a total breakdown and, subsequently, what makes them prone to respond with detached reflection upon mechanisms of the situation at hand, and what entices them to explore new possibilities? Before we embark upon the analysis of the ‘bricoleurs of governance’, let us take a closer look at the void that is the object of their engagement.

8.2.2 Governance Void

First of all, I argue that the void that practitioners are facing is of institutional nature. Mair , Marti, and Ventresca suggest that institutional voids occur “amidst institutional plurality and are the intermediate outcome of conflict and contradictions among local political, community, and religious spheres” (Mair et al. 2012, p. 820). This perspective seems to be more suitable then perceiving voids only as spaces where certain institutions are absent or weak. It is because governance, as a certain idealized concept and approach to public management, is introduced into a field that is already governed by diverse political and community institutions. Even when certain institutions of governance are non-existent or weak (e.g. patterns for multijurisdictional coordination and inclusion of the public into decision-making processes), a given field (e.g. anti-flood facilities) is being regulated within complex local arrangements and interdependences, with formal and informal institutions influencing individual practices, as it is already shown in the analysis above.

Definition

Governance void is a space where direct and/or indirect institutional infrastructure supporting collaborative forms of governance is absent or weak (underdeveloped), and where the contradiction of institutional logics provides inspiration on how to manage a given public issue.

Two examples of the governance void are outlined below. The first example relates to an explicit lack of institutionalized tools that officials could use in order to enact a new governance mode. The second example illustrates the existence of the void within a larger institutional infrastructure regulating collective action within the Polish context.

Governance understood in a narrow sense (as a governance mode based on mainly networks, see Chap. 1) by definition involves the coordination of diverse stakeholders: across levels of administration, across diverse departments of the same public organization, as well as across diverse organizations. To a large extent it requires an approach that is problem-centred and not task-centred (Ansell 2011). Yet , those formally responsible for the planning process in the case of the market and in the case of anti-flood facilities had neither legal and organizational tools nor enough power to implement this approach. In the first case, the planning process should have been coordinated by the Mayor. According to formal regulations, it was the Mayor’s task to coordinate the four entities across different government levels and jurisdictions, with the actions of external stakeholders. These were real estate departments at municipality and city level, the bureau of architecture (city), the investor (city-owned construction company), and the merchants and local community representatives. The new Vice President, who eventually took over the responsibility and successfully finalized the process, admits that despite of the fact that the Mayor was formally accountable for the process, he did not have the necessary tools (e.g. procedures) to secure and enforce cooperation of the four public administration entities listed above. He mentions that department directors within the city and municipality administration are neither used to coordinating their actions nor are they held accountable for that (see Chap. 5 for details), and that the competence of coordinating the processes rests with the highest levels of organizational structure occupied by the President or Vice President. After stating this generalized observation, he continues with an example from his own professional experience.

Before I became the Vice President, I was the head of a department whose operation depended entirely on cooperation: it was responsible for acquiring European funds. I can tell you that we went through hell. It was OK if I spoke to someone and the person understood that I needed something from them, and they needed something from me. But if the head of the department was… reluctant… a phone call to the President was necessary. The President had to persuade the person to work with me.⁵ (interview, June 2013)

In other words organizational structures (procedures, motivation systems) enabling horizontal and vertical coordination were lacking and the successful coordination was dependent on individual skills and the position of power, observation that will be further explored later on in the analysis (see considerations on institutional entrepreneurship, Sect. 8.3).

In the case of anti-flood facilities, problem-centred approach would require ‘the owner of the process’ (regional agency) to coordinate action with municipalities , a national-level agency and external stakeholders, such as local community organizations and individual citizens. In both cases (the market and anti-flood facilities), further action would require from those formally responsible to go beyond their competences—an act involving high individual and organizational risks. In these kinds of situations, practitioners face a paradox described by Chris Ansell as a large gap between low discretion, that is, what public officials are legally allowed to do, and high responsibility that is, being responsible for effectively solving public problems (Ansell 2011). This gap is one of the manifestations of the governance void.

Even if network-based coordination of relationships between different actors within and outside of public administration is not explicitly built into its organizational structures, the institutional environment may deliver a more or less productive ground for establishing such ways of coordination. As already mentioned (see Chap. 3), traditionally speaking, horizontal ways of governing are not typical of the Polish institutional environment. As a result, new collaborative and participatory approaches to governance are interpreted according to the traditionally adopted hierarchical logic, or to relatively new, but firmly embedded quasi-market logic of New Public Management. Lack of examples proving that horizontal coordination and partnerships may bring desired outcomes not only confines imagination about the realm of the possible, but also reduces resources necessary to initiate cooperation, most prominently social trust (Ostrom 1990; Putnam et al. 1993). The latter is built in the course of collective actions, for instance within civil society initiatives aimed at solving local problems, or ‘only’ at organizing collective leisure, or within citizens-government interactions taking place during gatherings in local communities. Practising cooperation turns it into a value worth pursuing. In this sense, each attempt to solve local problems is path-dependent, influenced by the history of mutual relationships and past decisions (Forester 2017; Schreyögg et al. 2011), which determines the scope of choices accessible to actors involved in the governance process.

The path-dependency of the governance process and the significance of social trust as a factor affecting its success are most clearly visible in the case of the dispute around anti-flood facilities. Officials did not have any good experiences with participatory processes (rather the opposite), they did not value cooperation with stakeholders per se, and they did not expect to learn anything from them. As a result, they were not motivated to experiment further with consultation. Once they had come to the conclusion that it was impossible to convince the local community, they stepped back. Regional agency officials and experts involved in the planning process had a solution (barrier) to the problem (flood), and what they thought they needed to proceed was an informed consent of the stakeholders. What they felt they needed to learn was how to convince the stakeholders more effectively that the barrier was the right solution. The contested barrier was an important part of a wider system. Several new barriers were to be constructed in the following years. The regional agency was supposed to perform additional services, namely dredging of the bottom of the Vistula River.

For the local community the problem was not the flood or even the barrier itself, but the unreliability of public agencies. As one of local activists put it: “We did not believe them. We will be left to our own devices. If there was a crisis and no money was available, no other barriers would be built (except for this one)”⁶ (interview, March 2014). Over the years, local stakeholders learnt that plans and promises are made, but they are often not implemented in their entirety; sometimes they are abandoned altogether. Dredging was supposed to be conducted on a regular basis, yet it was discontinued for 20 years because of budgetary cuts. After the disaster of 2010, national authorities promised compensation to those affected by the flood, but no money was paid and claims for damages were challenged. The state of existent flood-prevention facilities, such as embankments and drilling ditches, was unsatisfactory.⁷ Critically, barriers would require an immediate and efficient evacuation in case of a flood. A statement of the Mayor who opposed the barrier deserves reiteration: “if someone promises to do this or that, but does nothing, we don’t believe him. Would you?” (interview, March 2013).

In the analysis of the case (see Chap. 7), I argue that the perspective of these two parties, that is, the local community and public officials could not converge into a constructive discussion of the problem and, therefore, the goals of governance in this particular case are not redefined. In other words and according to the theoretical frameworks of good governance, metagovernance, and learning, critical reflection was needed in order to question the methods and goals of the planning process. No reflection took place. The problem remained unsolved. Yet, even if reflection had occurred, restoring the trust among the stakeholders of the process would require great facilitation and mediation efforts.

Paraphrasing Kenneth Arrow, one can say that social trust is an important lubricant of the governance system. Its absence, as well as the absence of institutions incentivising network-based cooperation in diverse areas of social life, constitutes deeper levels of the governance void. Here we should emphasize that it is not only officials who need to learn collaborative approaches. Since the idea of public governance is the inclusion of external actors into governance processes, these external actors also need to learn how to function within this new context and to perceive participation not only as a right, but also as responsibility. In the process of participation, external stakeholders can learn about both possibilities and limitations of designing public policy and public services. If their participation is not treated seriously, like in the case of barriers, they are only learning that if they protest intensively enough, they can stop any government initiative. Blocking an initiative does not, however, solve the problem. This kind of result is suboptimal, since the goal behind collaborative approaches to governance, and behind participation in decision-making in particular, is to design possible solutions to a problem using diverse sources of knowledge and exploring different needs and perspectives. The aim should be to expand the scope of choices for those participating in the decision-making process, not to reduce the choice to accepting or rejecting a pre-given solution.

8.2.3 Entering Double-Loop Learning: From a Reflexive Practitioner to an Institutional Entrepreneur

After having delineated the institutional background of practitioners’ withdrawal from problem-solving, I now come back to the question: what makes practitioners willing to perceive a disturbance as a total breakdown and, subsequently, what induces them to engage in detached, critical reflection upon the mechanisms governing the given situation, and in an active exploration of new possibilities?

Analysis of the cases suggests that critical reflection and experimentation emerged in contexts where practitioners responsible for a given governance process were exposed to institutional ambiguity, that is, the existence of conflicting frames of interpretation suggesting the manner in which to manage the problem and the goals of the management process. Importantly, sources of ambiguity triggering double-loop learning were located within public organization, or at least within the epistemic community of professional and academic experts, sharing normative orientations and knowledge on possible actions and desired outcomes (Haas 1992). As literature on institutional change suggests, shared understandings of institutional assumptions may exist to differing degrees and may shift over time (Mahoney and Thelen 2010) . In the situation of a protracted dispute and inability to solve the problem on the basis of old approaches to governing processes, the responsibility for the process will more probably be delegated to alternatively oriented actors. For the authorities, the existence of alternatively oriented actors becomes an opportunity for regaining the effectiveness of action, restoring the legitimacy of public agency, and getting things done.

At this point of the analysis, the use of the reflexive practice model as describing an individual experience becomes even more challenging than before. It is because the most striking observation is that together with a change of logic, a constellation of actors involved in the governance learning process is shifting. On the operational level, those who have failed to solve the problem with the use of a specific logic are not those who have introduced the practice rooted in the new logic. Top decision-makers implicitly commission the change of logic by allowing new actors to enter the stage. In the cases under consideration, they commission the change not because they think or believe that this problem should be dealt with in a collaborative way, but because all accessible alternatives have been exhausted.

The process of perceiving a disturbance (public dispute) as a total breakdown and responding with detached reflection is extended in time and takes place rather on the meso level of social interactions than on the micro (individual) level. It is not so much about individuals experiencing an “aha moment” and altering their own practices, as it is about individuals discovering their own limitations and realizing that it is necessary to engage actors who think differently. The observation of double-loop learning—to a greater extent than single-loop learning—allows us to understand that learning is a social process. Actors learn new approaches through interaction. Within interaction, they challenge each other’s views instead of supporting and reinforcing them, as it is the case in single-loop learning.

The two cases in which double-loop learning took place illustrate the social nature of the learning process and, in particular, how double-loop learning is facilitated by the heterogeneity of actors’ orientations within public organization and, more generally, within networks of practitioners and academic experts. In the case of market location, there were two moments in which traditional, hierarchical logic was replaced with the new collaborative logic and the process of double-loop learning began. The first was the moment when the first Vice President allowed academic researchers from the PSA to take control of the process. The second was the appointment of the new Vice President, whose professional orientation, just as the orientation of academics from the PSA, was aligned with the ideological orientation of the concept of collaborative and participatory approaches to governance.

Importantly, the first attempt to introduce alternative logic into the governance system proved a failure. The first Vice President allowed the PSA to take control of the process, assigned an official representing the city in the mediation, and stepped back. The disappearance of a major decision-maker impeded the learning process, because it was a signal that alternative governance practices were not actively supported by the top manager who, in this case, was also a decision-maker. Public officials, including the person assigned to take part in the mediation as a representative of the city, felt insecure and irritated by the new situation and persistently reverted to the original logic by arguing that “the procedure described by mediators was not defined in regulations governing the work of city authorities” (an open report prepared by the researchers, 2012). This particular situation helps us to pinpoint the issue considered central to the “governance turn” (Gilardi and Radaelli 2012). It involves experimenting with the roles of diverse actors in decision-making. Yet there are important limitations when the role of the government is concerned. Even if participative decision-making takes place, and even if non-governmental actors implement strategic public objectives, the government still should have a steering role (Pierre and Peters 2000; McLaverty 2011 ; Osborne and Gaebler 1992). The director of the Social Communication Department, who was implicitly involved in the process,⁸ summarizes the lesson from this experience by saying that city authorities should not have withdrawn and taken the role of a party in the mediation process, “instead, they should have played the role of the host” (interview, June 2013) who invites the public to participate and listens to its opinions, and eventually makes a decision.

The new Vice President entered the scene when the experiment with mediation was still ongoing. He refused, however, to step into the former Vice President’s shoes, and took back control of the process. He constructed a new temporary structure (a working group) on the interface of city administration, municipality administration, and external stakeholders’ organizations. He established communication channels between all relevant officials, councillors, politicians, and the working group. The process of designing solutions for the market was personally coordinated by the Vice President. Importantly, although the new Vice President shared the ideological orientation of collaborative and participatory approaches to governance, at that time, he had had no experience in governing multi-stakeholder processes. This is how he describes the encounter with protesters:

There were eight meetings and, to be honest, the first two were totally wasted in terms of any work on the design. These meetings ended very emotionally. I must admit, I have learnt that it makes no sense to expect that in a conflict situation parties just meet, talk and solve the problem. It does not work this way. These are strenuous processes; you really need to bleed and sweat in order to arrive at something. For me it really is a personal success that we were able to peacefully move the merchants to a temporary location and start modernization works. (interview, June 2013)

What makes the new Vice President prone to learn new governance mode and experiment are social skills⁹ (Fligstein 2001; Fligstein and McAdam 2012) and a strong conviction that involving external stakeholders is the right thing to do (individual conditions). His formal position of power accelerates his agency, because it makes other people follow him (structural condition). The institutional and organizational context of his actions makes him something more than a reflexive practitioner, he becomes an institutional entrepreneur who fills the governance void by: (1) initiating divergent changes (here: changes following an alternative governance logic); and (2) actively participating in the implementation of these changes (Battilana et al. 2009).

The WWTP case provides yet another example, illustrating how ambiguity of new governance institutions and the presence of actors delivering alternative framings open up a space for double-loop learning (and subsequent institutional change, as will be argued in the next section). Governance mode enacted by public officials responsible for the planning process did not serve ideas, needs, and interests of individual citizens , local communities, and civil organizations and—most importantly—some of the individuals and departments within the city administration.

In a response to coercive pressures (changes in law) taking place in the late 1990s and early 2000s, some organizational structures were created, whose role was to maintain communication channels with general public and non-governmental organizations. These structures involved a small department for public dialogue and certain new responsibilities regarding information and consultation were delegated to existing departments. The conflict around the WWTP , one of the longest and most intensive public disputes in the city’s recent history, infused these structures with new energy. The ‘inhabitants’ of these structures, that is, public officials employed within new departments and officials equipped with new responsibilities had an opportunity to justify their own existence and increase resources at their disposal. As introduced in the case description (Chap. 6), the investor, that is, a city-owned company supported by some of the officials, tried out many novel participatory tools and mechanisms, such as a special position on the company’s executive board (i.e. the board’s plenipotentiary for social obligations), the organization of meetings with the parties in the administrative trial, the creation of the Social Council with the purpose of investment monitoring and, last but not least, the employment of a mediator. Public officials and professionals involved in the process were learning how to use new, participatory governance tools, yet they were not successful in resolving the dispute. Crucial decisions were made without public participation, and despite being contested by the public, they proved irreversible. In this sense, the process was locked in through past decisions, and even participatory practices conducted along with participatory values and intentions failed to substantially improve the situation. Yet, failure to solve a problem is not tantamount to one’s failure to learn or to institutionalize (see the subsequent section, the second institutionalization pattern).

Cases in which no double-loop learning emerged deliver an important counterpoint to the argument, highlighting the importance of institutional ambiguity and actors’ heterogeneity. In the discussion on the delegation of schools, opposing voices coming from the local community, trade unions, and certain councillors were systematically set aside by decision-makers, who proceeded in accordance with hierarchical and quasi-market logic. Their dominating position enabled them to either dismiss or ignore alternative approaches proposed by external actors. The homogeneity of perspectives among those in control over the process reinforced the conviction that the right way to handle disagreement had been chosen. The Mayor’s conviction that people protest because they are excessively demanding or manipulated tallies with councillors’ belief that the public do not understand internal governance processes and, therefore, should delegate decision to democratically elected representatives. The emphasis put by the legal advisor of the municipality on the necessity to exactly follow the rule of law makes experiments with participation redundant and reduces them to documenting the fact that public consultation has indeed taken place. Although different, these perspectives have at least one important thing in common: they do not regard external actors as either capable or legitimate partners in the governance processes. In an interaction, these views reinforce each other and minimize chances for any critical inquiry and entering the double-loop learning process.

The barrier case represents a similar pattern. Here, the mutually reinforcing perspectives are those of public officials, and academic and non-academic experts involved in the process. They all agree on one thing: people need to be convinced that “what we are proposing is right” (regional agency representative, interview, March 2014). They all perceive participatory approaches to governance through the paternalistic lens of education. Within their community of practitioners there is no one who could challenge their view and propose an alternative framing, suggesting that there is some important, context-based and local knowledge to be gained from those opposing the investment. Despite their emotional engagement in the process, they were not able to think outside the hierarchical logics. Here, officials did not make an instrumental use of participatory tools, as it was the case in the dispute over the WWTP . They believed that the aim of the process was to educate people on the benefits of the solution and to gain acceptance for the project. Inability to reach this goal despite multiple attempts and modifications in the form of meetings exposed officials and experts to a great deal of stress. The latter is exemplified in an expression used in a conversation between a professor invited to a meeting and a representative of the regional agency: “it’s time to retire (…) it is not worth trying to convince them, it will only make you sick.” (academic expert, interview, November 2013). Homogenized and technocratic orientations regarding participation and governance modes directed all their efforts and resources towards single-loop learning and prevented the questioning of the goal they were trying to reach.

The above considerations focused on governance double-loop learning are summarized in the following proposition:

Proposition 7

Critical reflection on governance and double-loop governance learning is more likely when public administration practitioners are exposed to conflicting interpretations of governance institutions (institutional ambiguity) within their own organization or, more broadly, within their own epistemic community.¹⁰

8.2.4 Practitioners’ Responses to Disruptions in the Institutional Perspective

Observing decision-making processes in a collective setting draws attention to how institutional and political context influences practitioners’ responses to disruption. Sociological accounts of institutions focus on non-codified, informal conventions and collective scripts that regulate human behaviour (Mahoney and Thelen 2010) . This perspective enables capturing the cognitive significance of institutional structures and recognizing the fact that actors may reproduce institutional logic without any conscious scrutiny—a conclusion of central importance for the governance learning phenomenon. Through the analysis of public dispute resolution, the working of cognitive pillar of institutions could be identified. I have developed propositions supplementing the typology of responses to surprise (disruption), which indicate how institutions penetrate thinking and acting, and how they prevent practitioners from being surprised, induce them to perpetuate a given mode of governance despite the occurrence of disturbances, even when doing so is not “efficient” (Mahoney and Thelen 2010; Powell 1991). In this sense, an “institutionally sensitive” definition of surprise could read as follows¹¹:

Definition

Surprise is a cognitive state caused by a disruption of institutionalized patterns of thinking and behaviour deployed by an (public) organization to deal with a specific (social) problem.

Proposition 8

The experience of surprise is crucial for governance learning because it triggers reflection in action and on action.

Importantly, in cases under consideration, detached reflection and double-loop learning were not the result of a planned organizational effort. Rather they occurred thanks to the favouring configuration of specific structural and individual conditions, in particular the diversity of perspectives on goals and means of governance within public agency itself. It should be emphasized that this diversity is, in fact, a manifestation of coercive, normative, and mimetic pressures. If legal regulations had not been amended, protesters’ claims would not have been legitimate; if professional education of diverse kinds had not transmitted certain values and concepts of participatory approaches, officials willing to engage the public would have been hard to find, and it would have been even harder to find skilled mediators and facilitators. Last but not least, if examples of collaborative approaches had not been accessible to practitioners, it would have proven extremely challenging to enhance and justify experimentation.

The table below illustrates linking the perceptions of surprise with responses to surprise, as well as institutional and political influences. The latter refers to the existence of parties whose needs and perspectives, including value-based perspectives, are not well served by the dominant institutional arrangement. Here, parties are divided into two groups: (a) a group of external stakeholders of the decision-making process, who actively oppose the means and goals of governance performed by public administration representatives; and (b) a group of internal stakeholders who—albeit to a varying degree—share understanding of goals and means of governance (Table 8.2).

Table 8.2 Implementation of newly introduced governance institutions. Practitioners’ responses to disruption in the institutional perspective

The table summarizes considerations on how practitioners’ experiences are grounded in institutions, and how institutions are inhabited and enacted by individuals. These conclusions should be perceived through the lens of considerations on metagovernance and learning presented in Chap. 1. If good governance means the ability to reflect and rebalance diverse governance modes, it would require public agencies to institutionalize reflection and learning. What is described here on the basis of case studies was an adaptive and unplanned process of learning of a new, network-based governance mode. It took place through incremental changes in hierarchical and/or quasi-market-based logics of governance (single-loop) and then through the questioning of the last two.

Even if the new governance mode based on participation had been learnt by practitioners and institutionalized within public administration, it does not mean that public administration is ready to perform good governance. Ideally speaking, public agencies should be able to gain excellence in using all three logics of governing and be able to detect when a different logic is needed as a supplement, or when a diametrical change of the logic proves necessary. In this sense, the most important challenge for public agencies is to find an answer to the following question: how to institutionalize double-loop learning?

8.3 Three Patterns of Governance Institutionalization

By highlighting the importance of (a) institutional ambiguity; (b) the agency of actors who, while enacting institutions, can maintain or transform them; and (c) drive the influencing actors’ ability to challenge existent institutions, the previous section has set the stage for a description of how new governance mode is not only learned, but also institutionalized within public agencies.¹² As the analysis of empirical data suggests, institutionalization of new governance mode follows diverse patterns and results in hybrids containing original and new institutional arrangements in various proportions. As illustrated in Table 8.1 at the beginning of the chapter, three results are observed in cases under scrutiny. Within two of them (the case of schools and anti-flood facilities), a new governance mode was institutionalized as a top-down approach. Within another two (market location and WWTP ), the institutionalization of coordination between bottom-up and top-down governing mode was advanced. In all cases, the institutionalization of horizontal coordination lagged behind. As a result of cross-case analysis, three institutionalization patterns can be distinguished. The first pattern is marked by the reinforcement of the original institutional logic into new patterns of rule and is marked by the absence of challengers who would question the dominant logic and initiate changes. The second pattern develops through interactions between actors representing the dominant logic and challengers occupying peripheral positions within the organization or the epistemic community (Zietsma and Lawrence 2010; Battilana et al. 2009). The third pattern evolves through the encounter of actors representing the dominant logic with challengers occupying a central position in the organization or the epistemic community.

In each of these three patterns, exogenous shock displaying within a larger field (political and legal changes described as the ‘governance turn’) are translated into field-level changes through the interplay of public and non-public actors. In each of them ambiguity, agency, and power play an important role. Below, I describe these patterns and reflect on how learning is connected to, and yet decoupled from, the institutionalization process.

As already argued at the level of practice, the governance turn necessarily takes place within a governance void, a relative weakness or lack of institutional infrastructure enabling an effective implementation of new modes of governance. The new institutional logic—and a new set of institutions—need to be operationalized on the ground. New practices have to be introduced and, within these practices, new roles assigned to public officials, politicians, citizens, NGOs, and businesses. When the process starts, new institutions attached to the governance turn are ambiguous—they have a high degree of openness in interpretation and implementation. Because they are new, they disrupt routines and are enacted in a more reflexive, intentional way.

In the previous section, I argue that the degree and quality of reflection upon disruptions varies depending on how original institutions impact professional orientations, and on the constellation of different groups of actors: those reproducing and those challenging the status quo. Here, I shall extend this line of argumentation by indicating how these phenomena (original institutions and constellation of actors) continue to affect the process of institutionalization of the new governance mode, that is, the process within which network-based and participatory approaches become valued, formalized, and established within public administration. Specifically, I suggest that the reflexive and intentional way of acting in the process of interpreting and implementing new regulations means that actors on the ground perform institutional work, practice that is to intentionally affect institutions (Lawrence and Suddaby 2006; Lawrence et al. 2009; Zietsma and Lawrence 2010). In this case, it is about affecting governance institutions enabling public administration to coordinate multi-stakeholder, multijurisdictional, and multilevel processes. The work of actors who intend to affect institutions may involve projective, future-oriented agency, as well as habitual agency, selecting among sets of established routines (DiMaggio 1988; Maguire et al. 2004 ; Zietsma and Lawrence 2010) . While the latter replicates existing patterns of thinking and acting, the first requires institutional innovation and entrepreneurship (DiMaggio 1988; Hargrave and Van de Ven 2006; Mair and Marti 2009).

For at least two decades, institution studies have focused on untangling the paradox of embedded agency: how those subject to institutions in a field can effect changes within them. Such a frame put the emphasis on the active creation of new practices. Here, I am interested in both: how those subject to institutions in the field can both effect changes and resist them. The balancing of these two angles brings me closer to approaches that put power struggles and institutional inertia at a prominent place in the analysis (Becker 1995; Hallet and Ventresca 2006; Mahoney and Thelen 2010) . In this vein and in the course of empirical analysis, I have combined two types of institutional work: practice work, understood as developing and legitimizing practices, and identity work¹³ understood as developing and legitimizing roles and identities (Svenningsson and Alvesson 2003; Zietsma and Lawrence 2010; Gawer and Phillips 2013) with habitual and projective agency. This analytical exercise enables capturing actions oriented to the creation of qualitatively new solutions (following a new logic) from actions oriented to proposing “the same but in a new wrapping” (following the original, old logic) (Table 8.3).

Table 8.3 Institutional work around logic shifts

The governance turn exemplifies the logic shift in organizational fields of public administration agencies. These pressures generate tensions between the established and new practices and identities. In such instances, organization members couple environmental pressures for change with their daily practices (Binder 2007). Following Scott’s insight that organizations are opportunistic collections of divergent interests (Scott 1967, p. 23) , we must bear in mind that practitioners might respond to environmental pressures differently, with disruptive and incremental changes to be introduced to existing practices, or with an active resistance to change. The latter might take the form of persistent reference to the established logics, reiterating its assumptions and legitimizing attempts to preserve the status quo. As demonstrated by the analysis of cases, these reactions represent a combination of strategic and self-activating responses.

A major feature of the first institutionalization pattern observed in the empirical data is the (re)enforcement of original institutional logic into new patterns of rule, and hence a process evolving through habitual agency. The case of schools and the case of the anti-flood facility are exemplifications thereof. Actors resorted to habitual agency, reaching to cognitive and behavioural scripts encoded in old institutions in order to make sense of new regulations. Importantly, in both cases their actions were not purely strategic in the sense that decision-makers did not fight to maintain original power arrangements and protect their own interests (at least this was not the driving force behind their actions). This remark makes the label ‘habitual’ even more accurate. Following the collaborative logic when designing and implementing solutions for the problem of schools and anti-flood facilities would not diminish resources at the disposal of dominating actors (the Mayor and ‘her’ administration in the case of schools and the Regional agency in the case of anti-flood facilities). The major obstacle in practising new rules was that its logic was contradictory to their belief system, in particular convictions about professional identities, and about the kind of relationship between professionals (public officials and experts) and the field in which they operate.

Importantly, officials and experts seemed oblivious to this contradiction, as they looked at the new rules through an old frame. The institutional work they performed was, in principle, an intentional readjusting of new rules to fit old ways of thinking and acting. The more visible the disagreement of challengers outside an organization would be, the more intentional the readjustment would become. The identities of parties involved in public disputes were reconstructed according to well-established institutional logics: citizens and non-public organizations as those who listen and vote, or possibly deliver feedback on their level of satisfaction with services and products they have been provided with, and identities of public officials (especially those who have been elected) as solely responsible for the development of ideas and their implementation. Institutional work directed at developing and legitimizing new roles within an old logic reinforced and partially overlapped with the work directed at developing and legitimizing new practices: the Mayor could not imagine that citizens might contribute to the development of ideas. This is why she advanced the solution herself, and then argued for the need to vote over the proposed solution at Council meetings; and eventually, having a ready resolution accepted by the Council, she considered it ready for public consultation. The particular perception of the role played by citizens and officials also contributed to the creation of a bulletin that was to educate citizens on how public management processes work and why certain decisions, even if unpopular, need to be made.

A similar mechanism of mutual reinforcement of practice and identity work was observed in the case of barriers, where the perception of identities influenced practice work, for example, the manner in which meetings were organized and presented to people: rather as lectures than as spaces for an open exchange of information and discussion.

Institutional work performed by public officials precipitated responses from the field. In the case of schools, local protesters, as well as some insiders, for example, councillors lacked adequate social and organizational skills to perform institutional work effectively, or to challenge the dominant patterns of interpretation and implementation of a new governance mode. Attempts to challenge the dominant actor took place within the dominant frame, for example, councillors were trying to enter the discussion by questioning the legality of the proposed resolutions and not the logic of the operation itself. External actors who were not invited to take part in discussions did not have a chance to learn to participate and their passive attitude was further reinforced by the paternalistic attitudes of municipality authorities (for more a detailed description, see Chap. 4). As a result, the original, dominant pattern infiltrated new institutions, shaping mutual expectations and relationships. Over the following years, collaborative ways of governance were practised and refined, becoming quasi-surveys.

In the case of the barrier (anti-flood system), the response from the field was far more active and focused on resistance. Yet again, although more active, it remained resistance within an imposed frame. The local community participated in the process presented to them as participatory and as seeking their opinion and agreement. Within this process, officials were educating the local community on the advantages of the barrier, while the inhabitants continued to oppose the barrier and demanded a different solution. Data from interviews with professionals and experts involved in the process allow to speculate that despite this failure, their way of thinking about the participatory governance process remained unchanged. If we stick to Selznick’s definition of institutionalization, according to which “to institutionalize” means to “infuse the task with value” (Selznick 1957, p. 17), we are forced to assert that, in this particular field, participatory governance was not institutionalized at all, and that the very idea of governance and participation was degraded. None of the actors involved in the process could see any value in participation. As already mentioned, public officials perceive these approaches as exposing public management to the risk of demagogy, while local community sees them as a smoke screen enabling public agencies to pretend they care about citizens’ needs and opinions.

Proposition 9

In the absence of internal actors displaying projective agency, the process of interactions between actors within and outside the organization reinforces the well-established (old) institutional logic in practices and identities proposed by a new governance mode.

The second pattern of institutionalizing a new governance mode develops through interactions between actors representing the dominant logic and challengers occupying peripheral positions in the organization or epistemic community (Zietsma and Lawrence 2010; Battilana et al. 2009) . This pattern is best exemplified by the case of WWTP . The long-lasting conflict over the investment opened a space for alternatively oriented actors within an organization and, more broadly, within an epistemic community. In their attempts to maintain control and regain legitimacy questioned by the protesters, top-level public officials commissioned the use of participatory practices. Empirical data indicate that this decision was based on an instrumental approach to these practices and that new, participatory practices were accompanied by old ways of thinking (see case description for details). In this case, however, political and symbolic governance learning—as an intentional activity aimed at controlling policy spaces, not at improving governance itself—produced unintended meso-level outcomes. Despite their initially ritual character, participatory modes of governance started to penetrate organizational spaces, because certain actors found them right and/or beneficial. These actors performed both identity and practice work by developing and legitimizing new practices and identities within a new, collaborative and participatory logic. In the case of WWTP , several actors advocated the adoption of a new logic. Here, I shall focus on the Social Communication Department, which over the course of the conflict grew from a small unit at the bottom of the city administration pyramid to become an important department reporting directly to the President and be responsible for public consultation throughout the city (for a description of other actors who built their position by ‘riding the participatory wave’, see a more detailed description in Chap. 6).

In 2008, when the conflict escalated, the symbolic importance of the Department was strengthened by the President, who elevated its position within the organizational hierarchy. Importantly, this symbolic gesture was not followed by any decisions about providing the Department with additional resources. Nevertheless, it proved to be enough to trigger an institutionally entrepreneurial action. The director, a former NGO employee, used his knowledge and embeddedness in the non-governmental sector to mobilize external resources. The Department developed a project of public consultations in the municipalities of the city and received financing from EEA Grants. Having secured the necessary resources, the director employed new professionals with a background and orientations similar to his. All of them had had experience in the field of conflict resolution and facilitation of public meetings.

Over the years, the Department, its director and employees were a driving force behind the development and experimentation with participation, as well as the institutionalizing participatory logic in organizational structures and procedures of the City Hall. It does not mean, however, that the Department changed the way of thinking about and practising public management throughout the city administration structure. Instead, they accustomed the administration to the new participatory logic of governance through building a narrative justifying the need for a new approach, while at the same time allowing other organization members to keep some (or even most) of their original identity. This is an example of such an ‘inclusive’ frame, reducing internal tensions between a hierarchically oriented organizational identity and new practices associated with the new logic:

We need to combine the knowledge of experts representing different parties: those who use the city, those who manage it (officials), and experts in a given domain (for example architects). Each of them thinks she knows best, but they know different things. Officials have a vast knowledge of procedures, they know what is allowed, and happens elsewhere. The experts know what can be done from the technical point of view. Citizens know, because they use the city every day. We facilitate the coordination of these perspectives, we make them respect each other to see the complementarity of their knowledge. (interview, June 2013)

The strategy adopted here is based on avoiding any open confrontation and the direct questioning of original identities and practices. The staff of SCD do not push organization members to change, rather they try to exploit a void (governance void), that is, a space to which no one in the organization has any claims—on the contrary, it is a space that inspires feelings of uncertainty and fear.

SCD employees coupled this inclusive and non-confrontational identity work with similar practice work. This is how the director recalls the first attempts to implement public consultation practices within the city’s municipalities:

when we approached the mayors of municipalities with a ready project and the money to implement it (EEA Grants), the dialogue was different and much easier. Only one mayor was reluctant, he did not want any consultation, he told us it was a total waste of money. We were unable to convince him. Eventually, I told him: ‘come on, if you don’t agree, we won’t be able to settle the project’; it was only then that he finally agreed. In the end, he liked it so much that at the moment, he is constantly looking around for consultation opportunities. These (consultation) processes really show you a different perspective. It triggered learning processes … (interview, June 2013)

The core features of a strategy employed by an institutional entrepreneur with a peripheral position are: avoiding confrontation by showing the space in which new practices can be implemented along with old practices, by increasing the amount of resources to be shared, hence creating a situation of ‘a cheap try’,¹⁴ where all actors get a chance to experiment with the new logic, and learning in action, without risking too much. SCD employees skilfully built the Department’s identity as complementary to more established identities. They built their image as those capable of reducing tensions internally, between the organizational identity and new practices associated with the new logic (Svenningsson and Alvesson 2003). The following quote is a good illustration of this tension:

they [officials from different departments] come to us when they face a conflict, because they are afraid and do not know how to deal with it. We are an entity that stands out from the rest. Officials do not perceive us as part of the city administration, but rather as an extension of NGOs. To the outsiders, though, we are perceived as part of the administration. (SCD director, interview, July 2013)

SCD employees invested significant amounts of effort into building an institutional infrastructure within the governance void. They pushed for creating procedures, organizational positions and spaces designated for facilitating participatory processes. Thanks to their activities, district offices were equipped with a position formally responsible for organizing dialogue with the public. They established the Public Consultation Platform, a register of all past (since 2008) and current consultation processes freely accessible to the public. Eventually, with their help and advice, the city President issued an act defining public consultation and making it compulsory to announce each consultation process held by the city administration authorities on the Public Consultation Platform.

An important condition for this successful institutional work was the position occupied by the SCD in relation to diverse actors involved in the process of institutionalization of the new governance mode. The position bridged multiple fields representing different logics (Greenwood and Suddaby 2006 ; Kostova et al. 2008). This position of ‘in-betweenness’ increased the resources at their disposal—they were able to combine the support of social movements and local organizations (Hargraves and Van De Ven 2006) with an internal (yet instrumental) support of top-level management. Both kinds of support build their legitimacy inside an organization, yet if only the latter existed, their agency would be significantly weaker. The last quote from the interview with the SCD director:

There is no doubt that external financial resources leveraged change. You can write it explicitly: if it had not been for EEA Grants, we would have never accomplished so much in such a short time. Of course, political will is also necessary and there must be a leader who wants this kind of change. As for myself, I have drawn a lot of my strength from my background, from the NGO environment. It gave me a lot clout here within the administration, I felt supported in what I endeavoured. (interview, June 2013)

The ability to combine both: support from the outside and the support of top-level management seems to be crucial, especially when we compare SCD actions with those of another institutional entrepreneur attempting to affect institutions from a peripheral position. In the market location case, the Polish Sociological Association was such an institutional entrepreneur . Sociologists who implemented the participatory research project and conducted mediation between city officials on the one hand, and merchants and local community representatives on the other hand, were actors from outside city administration, yet closely connected to it through participation in the local epistemic community of academics and experts, who occasionally cooperated with city administration officials when developing local policies, organizing conferences and public meetings. Within this community, they developed particularly close relationships with SCD employees. Some of the researchers participated in abovementioned public consultation projects financed by EEA Grants. Researchers’ entrance into the conflict scene was preceded by the withdrawal of top-level management from the steering role. As described in a previous section (see Sect. 8.2), when not provided with a signal that experimenting is welcome, officials locked in the well-established logics. This observation is crucial since without the leading and supportive role of authorities, the new institution can easily slide into the path of degradation, as it was illustrated in the case of anti-flood facilities. When the mediation process ended, the majority of its participants¹⁵ were disappointed and angry and they started to contest the new mode of governance. Merchants had already been sceptical towards participatory approaches after observing that all collective bodies established by the municipality Mayor had been ignored by the city administration. The fiasco of the mediation process only reinforced their scepticism; they started to perceive mediation as an attempt to deceive them. Confusion about the role of mediators and city administration emerged and undermined their trust towards the institution of mediation and dialogue itself. In the interview, a representative of merchants stated: “I have learnt that before you start a dialogue you need to check who is financing the initiative, and if I hear that the city is paying for it, you will not see me at the table anymore” (interview, July 2014).

To sum up, the crucial characteristic of the second institutionalization pattern is institutional work performed by institutional entrepreneurs occupying a peripheral position within the organization. With their peripheral position, they tend to avoid direct questioning of the original, well-established logics. Disturbance in the organizational environment seems to them as an opportunity to act upon their values and increase resources they have at their disposal. They begin to develop new practices and identities within the governance void—a space that is unfamiliar and strange for other members of the organization. They use the frame of the new participatory logic, which resonates with expectations from the outside, expressed during the public dispute. The ability to mobilize the support of actors external to the organization, functioning according to the alternative, network-based logic provides them with tangible (money) and intangible resources (legitimacy). Hence, what makes their position special is not only peripheral, but also boundary bridging location. As a result they are able to fill the governance void with the institutional infrastructure (like formal and informal patterns of action, the creation of new roles and positions, and building narratives that legitimize them) from the bottom upwards and from the borders towards the centre of the organization.

Proposition 10

Institutional entrepreneurs occupying peripheral positions within an organization are more likely to succeed when developing and legitimizing practices and identities linked to a new governance mode if they are able to mobilize support from outside the organization, and use non-confrontational, inclusive strategies while negotiating the meaning of new identities and practices with organizational members.

The third institutionalization pattern evolves through the encounter of actors representing the dominant logic with challengers occupying a central position within the organization or epistemic community. This pattern was observed in the market location case, where a newly appointed Vice President “switched the cognitive gear” (Louis and Sutton 1991) from hierarchical to collaborative, and actively engaged in collaborative attempts to solve the dispute. His actions have already been described before (see Sect. 8.2 and Chap. 5); here, the analysis shall focus on the qualities of his interactions with other organizational actors and on the manner in which he negotiated the meaning of practices and identities within participatory approaches to public management.

As I argue in the case description, the new Vice President entered into a space that was already prepared for the adoption of the new logic. As a response to the conflict, authorities had resorted to participatory rhetoric in order to regain legitimacy. The previous Vice President went as far as having recourse to the use of ‘we have learnt, mantra (Hood 2000), when he declared in the local newspaper that authorities were in the process of ‘learning’ to organize dialogue (local newspaper interview, September 2009). Although the use of participatory approaches was instrumental and learning ‘fake’, a space for change opened up, as the hierarchical logic was officially questioned, and because it became obvious that old ways of thinking and acting fail in solving the problem. In other words, the governance void revealed itself and waited for someone to fill it.

Contrary to institutional entrepreneurs occupying a peripheral position within the organization, an institutional entrepreneur holding a central position can afford to adopt a confrontational strategy when performing institutional work. Having a formal position of power, the new Vice President was able to design new practices and make his subordinates follow them, even if it meant going against procedural and habitual methods. He initiated new collaborative practices and convinced organizational members across departments and levels of public administration to adopt them. Within these processes, the role of the members of administration was to discuss different solutions with the community and merchants’ representatives, and to provide them, as well as decision-makers, with information about any procedural and material aspects of the process. One of the merchants recalls:

he was always there during meetings. We did not need any mediators anymore. We spoke directly to the Vice President. He was even able to call for someone if some additional information was needed, and the person would come and report to all of us directly, e.g. about what had been done about a certain issue. (interview, April 2013)

Importantly, the employees who tried to work according to the new collaborative logic, received substantial support and could further learn and improve their practice. In this particular case, these were street-level bureaucrats, directly responsible for managing the market area, who were in a touch with the merchants on a daily basis.

Apart from collaborative practices involving external stakeholders, new practices introduced by the new Vice President were linked to multilevel and multijurisdictional coordination practices taking place within public administration (see Sect. 8.2.2 on governance void). In terms of institutionalization, it is important to emphasize that despite significant amounts of institutional work that the Vice President put into bridging silos structures, this kind of coordination remained hand-controlled and was pushed forward by an individual. No formal structures were in place to coordinate processes horizontally, around problems, instead of vertically, that is, around tasks (Ansell 2011). In order to put this coordination in place, the Vice President used his position of power either by ordering coordination directly, or by having a powerful ally—the city President—to convince a reluctant employee (see Sect. 8.2.2—the ‘phone call to the President’ scenario mentioned in an interview).

Along with introducing new practices, the new Vice President performed institutional work by building narratives justifying the new ways of acting, as well as the new roles adopted by the actors involved. The most prominent example is the ‘there-is-no-alternative’ narrative he provides when talking about replacing the hierarchical approach to public management with a participatory approach. First, he claims that such replacement is a fact and, second, that it is a historical necessity, a sign that a public management structure is becoming “civilized” (an original expression used by the Vice President). When describing the hierarchical approach, he directly questions the assumption of representative democracy in which citizens delegate power to politicians, who then solve public problems through public administration. He frames this approach as obsolete and old-fashioned, even shameful.

In 2006, you would hear such opinion everywhere. They would argue in favour of the iron rule of representative democracy and ask: if we should have direct participation and consultation, what are the councillors for? I hear such voices now, but no one dares say it out loud.

The major line of his argumentation is very different from the one used by the director of the Social Communication Department. Both of them believed in participatory approaches and promoted them, but the SCD director used to work with a more inclusive, conciliatory strategy. The Vice President could afford to be confrontational because of his formal position of power.

Yet, it was not only the structural condition of power that made the Vice President’s experiments with new practices so effective. He was equipped with social skills, above all the ability to conduct a respectful conversation and discuss conflicting interests and values in an open manner. During an interview, one of the merchants related the moment when the new Vice President replaced the old one; he began by saying:

It is a different kind of person. You can actually talk to him. (…) the previous Vice President was talking with us and, at the same time, looking at the documents, consulting something with his officials and checking his phone. (interview, April 2013)

These individual conditions (social skills), the ability to talk to people and make alliances, were a significant supplement to the structural condition (formal position of power). An employee of the city administration mentioned that the Vice President liked to say that he worked like a plumber.

It is because when he cannot come to an agreement with one person (usually a department director), and hence ‘the pipe of the decision making-process gets clogged’, he tries to work out a solution by bypassing the person and, thus, restores the flow. (SCD employee, interview, February 2016)

Good social skills is what makes both institutional entrepreneurs —the new Vice President and the CSC director—very similar. Again, juxtaposing them with other highly skilful institutional entrepreneurs , that is, PSA researchers and mediators, indicates that this feature does not suffice to trigger the process of institutional change, or even the process of learning within a (public) organization. The support of top-level organizational leaders remains necessary.

In terms of the institutionalization of the new governance mode within city administration structures, the results are mixed. On the one hand, the Vice President successfully solved the problem and he accomplished it using the practices and logic of the new governance mode. When focusing on the inclusion of external stakeholders, one can say that the institutionalization process proved a success. Nevertheless, issues of multijurisdictional and multilevel coordination lagged behind. As the Vice President himself admits, heads of departments are still neither expected to reach out to different departments and levels of organizational structures, nor held accountable for coordination. In this sense, breaking the silos structure emerges as a greater challenge than, for instance, the inclusion of an external stakeholder into governance processes.

This imbalance in the degree of institutionalization of two governance-related areas can be explained through reference to three issues. First of all, in the area of organizing dialogue and participation, top-down institutionalization efforts of the Vice President complemented bottom-up efforts of the SCD, and these two processes mutually increased their impact. Second, this area was an explicit subject of pressures from outside the organization, and both institutional entrepreneurs could relate to these pressures and draw legitimacy from them. Such external pressures were lacking in the case of multilevel and multijurisdictional coordination. And third, the area of dialogue and participation benefited from the institutional work performed by the SCD, directed at building infrastructure (procedures, positions) for inclusion, while institutional work performed by the Vice President focused on organizing ad hoc solutions. Practices proposed within these ad hoc solutions followed the new institutional logic and questioned the old one. They could have proven inspirational, but failed to provide clear rules that would govern future action. In other words, the new Vice President’s actions were, to a large extent, less sustainable. If he had been replaced with someone representing a different outlook and having lower social skills, practices coordinating multijurisdictional and multilevel collaborations might have been abandoned. Paradoxically, despite the fact that the Vice President’s strategy was far more successful when a particular problem needed to be solved, it is the SCD strategy that seems more effective when it comes to the institutionalization of governance.

Proposition 11

Institutional entrepreneurs’ success in pushing forward institutional change towards a new governance mode is more probable if they combine social skills with attempts directed at building formal structures enabling governance.

Proposition 12

The institutionalization of governance is most likely to succeed when bottom-up and top-down institutional processes are linked to each other and coupled with external institutional pressures.

Proposition 13

The institutionalization of governance is most likely to succeed if institutional work performed by an institutional entrepreneur contains critical elements, that is, directed at confrontation and questioning the dominating logic, as well as constructive, that is, focused on building formal and informal patterns of acting.

The following figure illustrates the three patterns of institutionalization. Each of them is triggered as a response to disruption in public management processes and the ‘discovery’ of the governance void—the absence or weakness of the institutional infrastructure enabling the practice of participatory and collaborative governance approaches. Even though the rules on public access to information and participation are formalized, these institutions are new and open with regard to interpretation and implementation, that is, they are marked by a degree of ambiguity. Whether new approaches to governance will be institutionalized along with the new logic depends on the presence of institutional entrepreneurs who engage in institutional work, developing and legitimizing new practices and identities within the framework of new logic. In the absence of such an innovator, frameworks typical for the old logic permeate new practices and identities imposed by formal rules (i.e. national legislation). Within this process, old ways of thinking and acting are reinforced—the new governance mode which, according to the idealized form, follows network-based, collaborative and participatory approaches, is institutionalized according to the old, well-established top-down hierarchical or quasi-market logic.

If the institutionalization of the new governance mode is infiltrated with a strong opposition from actors outside the public organization whose interests and orientations are not well served by the emerging institutional arrangement, a probable institutional result is a compromise and degradation of collaborative and participatory approaches to governance. Here, compromise means that none of the actors involved in the process believes these kinds of approaches to be useful or valuable. If the first institutionalization pattern is accompanied with a moderate or weak opposition from the outside, the definition of participation and collaboration designed according to the well-established (hierarchical or quasi-market) logic is accepted and further reinforced in interactions in the local public sphere. In both cases the institutionalization does not happen on its own as a result of institutional inertia. It is enacted by social actors, who actively create meaning and interpret new regulations in the process of social interactions (Fig. 8.2).

Fig. 8.2

Three patterns of governance institutionalization

The presence of institutional entrepreneurs within an organization¹⁶ challenges the status quo, it introduces an alternative interpretation of practices and identities. The entrepreneur (intrapreneur) initiates changes and actively participates in the implementation. Depending on the position (central, peripheral), she may: (a) adopt more or less confrontational strategies of institutional work; (b) draw on external rather than internal support for material and non-material resources. Independently from the position within an organization, the effectiveness of intrapreneurs depends on their social skills and support from the top-level management. Social skills enable her to effectively communicate and build coalitions. Support from top-level management provides her actions with legitimacy towards other employees of the organization. As a result, the process of institutionalization advances from the bottom upwards, and from the borders to the centre of the organization (intrapreneur in a peripheral location), and from the top to downwards (intrapreneur in a central position). Regardless of their peripheral or central location, actions of intrapreneurs are facilitated by outside pressures that intensify the sense of urgency and relevance to experiments undertaken by institutional entrepreneurs.

In cases under empirical analysis, the latter two institutionalization patterns were complementary and mutually reinforcing. The area of multilevel and multijurisdictional coordination, where this complementarity was relatively less visible, is also the area where the governance void remains.

8.3.1 Institutionalization and Learning as Different, yet Coupled Processes

Case studies documenting processes evolving over time and focusing on diverse actors and their interactions demonstrate that the process of institutionalization is different and, to a certain extent, separate from the process of learning. I propose to regard learning as an intentional activity undertaken at the individual level, transformed to the meso level of institutional structures through social interactions and the development of shared understanding and practices. Negotiation and power struggle permeate this transition; consequently, the meso-level effect rarely reflects the intentions of any particular individual or interest group in an adequate manner (Rządca and Strumińska-Kutra 2016). In some of the cases under consideration, officials learnt to use participatory tools for strategic reasons. Nevertheless, even such ‘fake’ learning resulted in institutional changes, initially unintended by the subjects of learning. I argue that motivation for learning (e.g. regaining legitimacy) is secondary. It holds particularly true for double-loop learning, whose core characteristic is the ability to question the appropriateness of originally adopted beliefs and practices. Once the established practice and its taken-for-grantedness has been questioned, you cannot “unthink it”. The questioning of the hierarchical or managerial logic that has been taken for granted, or even a superficial introduction of the participatory logic creates a powerful precedent, forms the basis for building new institutions, opens up space for negotiation among all actors involved in the process.

8.4 Conclusions

Patterns of institutionalization delineated above form part of a larger framework developed here within the frame of ‘practice-based theory of governance learning and institutionalization’. Practitioners maintaining and challenging the status quo as well as engaging institutional ambiguities are, to a certain extent, ‘products’ of larger institutional environments and, specifically, institutional pressures (see Sect. 4.2). If diverse ‘products’ of these pressures are present at a certain point defined by space and time, a critical reflection upon action (see Sect. 8.2) as well as the process of double-loop governance learning are more likely. Actors learning new approaches to governance engage in interactions influenced by local and extra local institutions and by power struggles. In the process of interactions, individual attempts to learn are transformed into a meso-level phenomenon, as shared understandings and practices are negotiated and developed through social interaction. New patterns of thinking and acting emerge and gradually accumulate in an organization as a generally accepted option. In other words, new patterns of governing become institutionalized.

The three sections above provide a comprehensive theoretical framework linking structural and constructivist moments of governance practice, explaining how structures are changed or maintained by consciously acting individuals (Rhodes 2012). It might be wrong or inadequate in many ways, but it delivers an epistemologically and practically useful method in which to investigate the phenomenon of learning in the public sector, as it focuses on core questions: Why and how does learning unfold? How does governance practice develop, as opposed to the ideal model? Eventually, it provides information on ‘critical junctures’ and configurations of influences making a difference in terms of how governance is learnt and institutionalized. This knowledge can provide us (academics and practitioners) with skills that allow turning spontaneous and accidental practices based on critical reflection and experimentation into a planned effort to institutionalize learning, above all learning in its explorative, double-loop form. Such institutionalization is necessary in order to ensure good governance, since governance learning is not about transition from traditional approaches, to public administration, to market-based approaches, for example, New Public Management and, eventually, to collaborative and participatory approaches such as New Public Governance. It is about learning how to use and improve each mode and how to switch between these options and balance them in response to problems that constantly evolve and reappear.

As the cases show, the diversity of actors’ orientations is crucial for triggering both learning processes and institutionalization processes. On the individual level, confrontation with the alternative logics of thinking and acting potentially triggers critical reflection and double-loop learning. Yet, in order to trigger institutional work directed at developing and legitimizing new practices and identities linked to the new logic, this diversity needs to be present within an organization (or at least within the community of practitioners).

Cases in which double-loop learning and institutionalization of the new governance mode took place deliver an illustration of how public organizations are able to accommodate contradictory institutional logics coupled with different governance modes. The diversity of institutional arrangements within an organization builds up in time in response to shifts within a wider institutional context . In literature, these shifts are referred to as changes in paradigms of public administration: from Weberian bureaucracy, institutional logic dominated by hierarchy, to quasi-market approaches, such as New Public Management, institutional logic seeking market-based solutions and, most recently, the ‘governance turn’ manifest in praising such concepts as New Public Governance, guided by institutional logic that refers to networks (Barzley 1992; Osborne and Gaebler 1992; Lynn 2001; Ansell 2011) . The process in which shifts in the institutional context alter the interpretation of organizational structures represents sedimentation rather than replacement (Cooper et al. 1996; Seo and Creed 2002) . Subsequent layers of institutional logic add up, and each of them produces its own structures and ideologies that provide resources for competing interests and values in institutionalization processes (reproduction and reconstruction).

Public administration deals with problems characterized by different degrees of complexity, and therefore needs to use different governance modes and institutional logics to respond to them (see Chap. 2). This is why all three modes of governance are needed and institutional contradictions mentioned above can be productive and create an opportunity for—and not a threat to—good governance. Internal inconsistencies that are the result of spontaneous, evolutionary process of adaptation to external pressures should be cultivated and institutionalized. In this sense, a linear narrative suggesting evolution from bureaucracy through markets to networks ought to be abandoned. Keeping the minimum level of variety (Jessop 2011) or, in other words, a minimum level of inconsistency, is what enables metagovernance—improving a specific governance mode and recomposing proportions of governance modes used in a given policy field. Having the diversity of governance modes institutionalized within public organization enables leaders to choose which governance mode needs to be activated, or decide when to engage in a collective, critical, and creative inquiry about handling a given problem. Embedding opportunities for such a collective critical reflection into organizational structures would be conducive to gaining excellence in practising a given governance mode, as well as mastering the art of switching between the modes and experimenting with new solutions. Within these kinds of structures , single-loop learning and double-loop learning are institutionalized, and public administration officials are turned into reflexive practitioners.

Along with institutionalizing the diversity of governance modes, the institutionalization of multilevel coordination is necessary. Top-level hierarchy leaders should take decisions regarding governance modes through close engagement in concrete problem-solving dilemmas of street-level bureaucrats. Within this model, top executives are freed from thorough operational control and can engage in strategic planning and provide the organization with a clear policy direction. They form a constitutional level focused on the overall policy direction and defining the goals of the organization (Ansell 2011, p. 119; Sabel 1999). Importantly, within this kind of hierarchy (called pragmatist hierarchy, Ansell 2011, p. 119), organizational levels are loosely connected and semiautonomous. Autonomy allows officials to engage in problem-solving, while the coupling of different levels prevents planning from being confined to a lofty and isolated position.

Ruminations on metagovernance and learning have consequences for the discussion on leadership in public organizations. Denis et al. (2009) argue that in order to enrich our thinking about leadership in public administration, we need to recognize the pluralistic nature of the organizational context within which leaders of public sector organizations operate, as well as the dynamic and collective nature of leadership processes in these settings. The analysis presented above strongly supports this conclusion and puts it forward by indicating how this plurality should be strategically used in order to enhance critical reflection and pursue governance learning. Among others, it calls for the reconceptualization of institutional leadership (Selznick 1957), which has thus far focused on managing the internal consistency of an organization (commitment to the values and mission of the organization), and hence has been perceived as leadership based on a backward-leaning vision, which means that the vision is there to remind the organization of its core values (Washington et al. 2008) . Institutional leadership is perceived by Selznick as a process in which values are promoted and protected. An institutional leader is a major actor of institutionalization process within which organizations are infused “with value beyond the technical requirements of the task at hand” (1957, p. 17). He suggests that the process of institutionalization takes place as leaders respond to internal and external forces that exert pressure on organizations. Yet the existence of multiple governance logics, and hence multiple values represented by each mode split the institutionalization process, as in the process of governance, the number of values that need promotion and protection multiplies.

As illustrated in the analysed cases, the value of social justice and democracy pursued by public administration can be understood differently depending on the governance mode and institutional logic that accompany it. For instance, within the hierarchical logic, public agencies are the final link in a chain of democratic representation that begins with the electorate, moves on to the legislature, then to appointed agency officials, and finally reaches street-level bureaucrats. Network-based logic proposes building up societal and democratic consent through collaborative problem-solving (Ansell 2011, pp. 17–18). Market-based logic would build consent around the quality and costs of public services. Considerations on metagovernance presented here and in the Chap. 2 of this book suggest that the concept of institutional leadership within public administration should be extended to include reflexivity, the leaders’ ability to work with different governance modes (or ways of organizing , Alvesson et al. 2017), critical assessment of the act of problem-solving and, when necessary, the ability to switch between the logics of governance and interpretations of values and to blend them depending on the situation. As the empirical analysis indicates and theoretical insights suggest, reflexivity is practised above all collectively and remains closely connected to acting upon the problem. In decision-making settings characterized by high uncertainty and a lack of consensus on values, institutional leadership requires balancing the backward-leaning position (Washington et al. 2008), promoting and protecting values such as social justice and democracy, with a forward-leaning vision of an institutional entrepreneur , focused on changing practices within public organizations.