We combine a dual perspective to analyze government action in administrative reform: the organizational approach (Friedberg, 1997; Garfinkel, 1967; Gibert, 2009; Hatchuel, 1990, 1994; Pichault, 2008, 2009; Vas, 2015) and the cognitive public policy analysis approach that relies on the concept of the global frame of reference (Muller, 1985, 2000, 2005, 2015). These perspectives allow for the consideration of actors in terms of their strategy (organizational approach) and in terms of their ideas and interests (cognitive approach to public policy). Furthermore, they adopt the systemic approach and therefore integrate all its principles, including those presented above.

From the point of view of the management of administrative reforms, the organizational approach deals with “the construction and maintenance of local orders that ensure the regulation of behaviors and the integration of the divergent, if not conflicting, strategies of the actors concerned” (Friedberg, 1997: p. 187). In the role of third-party helper, the intervener (the consultant) focuses on the methods and instruments (or tools) used by the “actors in situation (along the workflow of the reforms conducted)”, i.e., on the relations that are established between the actors concerned, rather than on the individuals (Friedberg, 1997: p. 227, 304).

Following Muller’s cognitive approach, the global framework of reference designates “an overall vision of the economy, citizenship and the State from the point of view of the conditions of “good government”; “a vision of what society must “do”(from the State)in order to “be”what it “is”and “must be” (Muller, 2015: p. 413). It constitutes a framework of cognitive interpretation and normative prescription constructed by identifiable actors for each cycle of public action and corresponding to a certain state of the world that is both “real” and “constructed” by their self-referential thinking (production of thought about oneself). The fundamental characteristic of the global framework of reference is to be contradictory, because social reproduction is itself a contradictory phenomenon in that it defines a field within which the conflicts and confrontations of society are organized (Muller, 1985).

To process the data collected after the literature review and drawing on various field experiences, we used three grids. The first allowed us to answer four questions: 1) the origin of change, 2) patterns of change, 3) resistance to change, and 4) change management. In the form of “matrices of the effects of the changes introduced” (Miles & Huberman, 2003; Ngouo, 2018), the second focused on deciphering, in the documents, the descriptive elements of the real workflow that reflected the actual progress of the projects/activities (discrepancies between planned and realized, discrepancies between expected and observed effects, monitoring of performance drivers...); the third sought to identify the origin and typology of the paradoxes, as well as the appropriate strategies for managing them according to a recognition logic of their creative potential.

With regard to the scientific work selected for this paragraph, we would like to echo the advice of Rousseau et al. (2008: p. 506) who point out that “all studies have limitations. Only in their combination does evidence emerge”. Le Moigne (1990), as quoted above, teaches us that a phenomenon considered complex cannot be locked into a single model of representation, no matter how sophisticated that model may be.

We define the concept of paradox, followed by an analysis of the epistemological issues involved in managing paradoxes, and conclude by exploring the contribution of systems analysis principles to paradox management.

The concept of paradox

By definition, a paradox is an expression of “enduring, even permanent, contradictions between elements that are apparently mutually exclusive but coexist nonetheless” (Cameron & Quinn, 1988: p. 2; Bollecker & Nobre, 2016: p. 45; Lindgreen & Maon, 2019: p. 139; Niesten & Stefan, 2019: p.233). Tensions arise from these contradictions between interdependent elements. By tension, we mean “the stress, anxiety, discomfort, or oppressions of making choices, responding, and moving forward in organizational or [social] situations” (Niesten & Stefan, 2019: p. 233; Putnam, Fairhurst, & Banghart, 2016: p. 68). These effects are sources of cognitive and behavioral paralysis within organizations (Dietrich, 2000; Gaulejac (de), 2010; Ngouo, 2017; Ragaigne et al., 2018; Watzlawick et al., 1972).

Some examples of paradoxes. Organizations need to be both global and local, to be large in some ways and small in others, to be centralized at times, and decentralized most of the time (Handy, 1995: p.41). From a Communication Constitutes Organization (CCO) perspective, Schoeneborn, Kuhn, and Kärreman (2019: pp. 480-489) identify and describe three types of tension, namely 1) the tension between “communication” as a verb or process and “organization” as a social entity or actor (verb-noun tension); 2) the tension between “communicate” as a verb or process and “organize” as a verb or process (verb-verb tension); and 3) the tension between “communication” as a verb or process and “organization” as an adjective or attribute (verb-adjective tension). Still in the communicative perspective, the pragmatic paradox is defined as a construction of the actors that locates the contradiction in the effect that is produced by what is said in a communication-type interaction. This is the case, for example, of injunctions such as: “don’t treat me as if I were in an inferior position!” (Conflictual interaction situation), “I would like you to do that!”(Indirect exercise of authority), “Your defense proves your fault” (Interaction in groups where suspicions of jealousy float) (Keller, 2004), “Be efficient” or “Be creative and proactive” which the manager activates as a support of the personality of the organization (Gramaccia, 2011).

The literature (Aucoin, 1990; Bollecker & Nobre, 2016; Emery & Giauque, 2005; Grimand et al., 2014, 2018a, 2018b; Hudon & Mazouz, 2014; Mazouz (dir), 2017; Mazouz & Gagnon (dir), 2019; Lewis, 2000; Lindgreen & Maon, 2019; Niesten & Stefan, 2019; Poole & Van de Ven, 1989; Trosa, 2017) offers a varied typology of paradoxes and strategies for managing them that we will mobilize throughout the article.

Managing paradoxes: the need for a prior epistemological framework

Designating phenomena as paradoxes do not necessarily contribute to a deeper understanding of them and the requirement to take them into account in managing organizations (Mazouz & Gagnon, 2019; Lewis, 2000; Niesten & Stefan, 2019). Denial in all its forms (repression, separation, prioritization) are problem avoidance strategies that have the effect of permanently compromising the quality of organizational performance; therefore, it is more responsible to confront reality as it presents itself in its ontological complexity by seeking cumulative solutions. This means, for example, confronting the paradox by integrating it into a new frame of reference that erases the previous contradiction and places the organization in a learning dynamic. Lewis (2000: p. 764) speaks of “adopting a paradoxical way of thinking” to describe this strategy. In the specific case of administrative reforms, Drumaux (2011: p. 8) concludes that a paradox is not a pathology, but a singular unintended and unanticipated outcome in the spirit of reform, which is not unwelcome in itself, and should therefore be managed appropriately.

It is therefore, more of an epistemological problem to construct knowledge related to the concept of paradox considered as a scientific object. Researchers, organizational development consultants, and anyone involved in administrative reform must be aware of the importance of this issue to avoid the kind of confusion referred to by Lindgreen and Maon (2019: p. 139) cited above. Indeed, as Boudon (1986: p. 129) points out, the researcher, the consultant, the analyst-auditor must specify beforehand the epistemological framework he or she adopts. It is a matter of clarifying the intellectual premises and methodological principles that will guide the description and analysis of the “reality” that he or she wishes to define while striving to achieve objectivity that depends solely on the theoretical, epistemological, methodological and axiological choices made (Allard-Poesi & Perret, 2014; Avenier & Gavard-Perret, 2012; Avenier & Thomas, 2012; Johnson, 2012; Le Moigne, 1987, 1990; Pichault, 2009; Rondeaux, 2008; Rousseau, Manning, & Denyer, 2008). The clarification of the epistemological framework constitutes crucial information to enable other stakeholders in the dynamics of the reform workflow (learning, innovation, etc.) to appropriate the explanations provided and then use them to address the development challenges of their organizations. Furthermore, some of the models adopted in the epistemological framework often include unexplained assumptions that are outside the framework adopted and that can have considerable effects on the definition and interpretation of paradoxes. Boudon (1986) calls these “E-effects” (for “epistemological effects”).

Contribution of systemic principles to the management of paradoxes

The problems generated by paradoxes can be analyzed and managed with the help of some systemic principles deemed appropriate for the study of complexity, which itself poses the paradox of one and multiple (Bovais, 2014; Burnes, 2004; Fortin, 2000; Le Moigne, 1987, 1990; Morgan, 1989; Morin, 1977, 1999, 2011; Muller, 1985; Renier, 2016; Vandangeon-Dumez, Grimand, & Shafer, 2018). Burnes’ (2004: p. 314-315) analysis of complexity theories as applied to organizations concludes that three central concepts are at the heart of complexity theories: the nature of chaos and order (chaos is considered as an order that describes a complex, unpredictable, ordered disorder in which patterns of behavior unfold in irregular but similar forms through a process of self-organization); the “edge of chaos” (an intermediate zone that exists between order and disorder. Here, complex systems exhibit the most prolific, complex and continuous change. It is argued that creativity and growth are optimal when a complex system operates at the edge of chaos); and order-generating rules (rules by which complex, nonlinear, self-organizing systems manage to maintain themselves at the edge of chaos even under changing environmental conditions...).

Morin (1977, 1999) proposes the dialogical principle of “order-chaos organization”, which acts at the “edge of chaos”, as an order-generating rule by allowing two non-reducible and contradictory instances to be linked together. Considered as one of the pillars of the complexity paradigm, this principle stipulates that the knowledge of the same domain or organizational phenomenon is based on the superposition of several logics or several types of discursivity (see the example of the attributes of public organizations). It is necessary to ensure both reproduction (in time) and existence (the present), order and disorder, permanence and transformation. In this article, we refer to this general principle. At the operational level, in a process of reconciling divergent logics, dialogical processes consist of bringing together citizens, decision-makers, managers, and experts in the field, informing them, listening to what they have to say, and encouraging exchanges around issues and values according to a problem-solving approach in order to arrive at concrete solutions that will obtain the support of the greatest number of people (Maxwell, 2002; Yankelovich, 2001).

Three principles seem to us to be particularly relevant to the analysis of this question of reconciling divergent logics by emphasizing the capacity for adaptation and self-organization of complex systems. These are the principles: 1) of the required variety which correlates with the principle of subsidiarity (Morgan, 1989; Renier, 2016); 2) of circularity,recursivity and autopoiesis which allow us to appreciate the way in which constructed reality unfolds and transforms itself in a continuous manner (Morgan, 1989; Morin, 1999); 3) of equifinality (von Bertalanffy, 1973; Boris, 2002), one of the strengths of which is to emphasize that there are different ways of arriving at one end of the continuum formed by two antagonistic or contradictory terms of a paradox and thus offers, through the combined mobilization of several managerial responses, a way out of the prevalence of defensive modes of managing paradoxes (see the pendulum movements between the poles of the continuum defined by Aucoin (1990); see also the oxymoron “a wonderful misfortune” used by Boris (2002), to analyze the resilience process).

The mobilization of these principles participates in the reinforcement of attitudes favorable to the management of paradoxes, including: the valorization of the virtues of paradoxes, the stimulation of contradictory perspectives, the climate of trust to be established, the coherent articulation of duality, the regulation of organizational paradoxes by the overcoming of defensive management modes.

We define and illustrate the concept and show how, ontologically, paradoxes are consubstantial with administrative reform.

The concept of administrative reform

Administrative reform is defined as a significant (even radical) change introduced in a conscious, deliberate and thoughtful way in a public institution, organization or system in order to improve its performance (Birkinshaw, Hamel, & Mol, 2008; Burnes, 2004; Gow, 2012; Mazouz & Gagnon, 2019; Pesqueux, 2015).

In this sense, administrative reform actions are not limited to simply rearranging and reorganizing structures, but rather attempting a real mental and sociological revolution in order to achieve a new conception of the State and its administration and a new perception of them by citizens. It is in this sense that any administrative reform project is necessarily part of a public policy. The change introduced can be: a dynamic process synonymous with innovation (see “radical” change); an incremental process referring to the past by underlining the duality “stability-change”, “long time-short time”; an evolutionary process where we distinguish “continuous evolution” from “brutal evolution”; an identification and representation of the “tensions” that will be reflected in organizational practices by emphasizing the change management method through managerial voluntarism, or the support method; a continuous improvement approach to organizational performance following the management principles of the quality approach (Benner & Tushman, 2003; Birkinshaw et al., 2008; Burnes, 2004; Mazouz & Gagnon, 2019; Rondeau, 2008; Rondeaux, 2008; Pesqueux, 2015). As an example, Birkinshaw et al. (2008: p. 832) propose a theoretical model for analyzing the processes characteristic of “management innovation”, which is a radical type of change. Management innovation is defined as “the generation and implementation of management practice, processes, structure, or technique that is new to state of the art and in intended to further organizational goals” (Birkinshaw et al., 2008: p. 829). Examples of management innovation provided by Birkinshaw et al. (2008: p. 830) include: a new set of practices and processes aimed at improving production efficiency and reducing waste; a new organizational structure with the objective of increasing employee initiatives and overcoming problems of hierarchy; a new set of practices and processes around the job design of employees with the goal of improving their happiness at work. The framework proposed by Birkinshaw et al. (2008: p. 832) is defined around two axes: the axis of activities (motivation, invention, implementation, theorization and labeling) and the axis of stakeholders’actions (internal change agents and external change agents).

We propose this model schematized in Figure 1 to illustrate the complexity of the ontological nature of administrative reform. Indeed, as Birkinshaw et al. (2008: p.832) point out, “these processes are typically complex, recursive, and occur in nested and repeated cycles of variation, selection and retention”. Note that the figure should be read horizontally and vertically to take into account the influences of the internal organizational and external environmental contexts. Thus, for example, for the activity of “motivation” (horizontal axis), internal change agents (vertical axis) identify a new problem (a perceived shortfall between the organization’s current and potential performance) (see No. 1) that undermines the current performance or opportunities that may exist and the anticipation of environmental changes. External change agents identify threats and opportunities (see No. 5) in the organization’s external environment that they share (directly or indirectly) with internal change agents.

Paradoxes in the dynamics of administrative reform processes

Administrative reform is ontologically a complex reality and the paradoxes it faces are a consequence of this. The epistemology of complexity as we define it here is based on two key concepts characteristic of the two paradigms of methodological individualism (the deliberate) and methodological holism (the emergent), namely the autonomy of actors and the processes of emergence of a “whole” that is self-generating and self-regulating (Avenier, 2000; Boudon 1986; Burnes, 2004; Le Moigne, 1990; Morin 1984, 1990). According to Burnes (2004: p. 310), “complexity theories are concerned with the emergence of order in nonlinear dynamic systems, that is, systems that are constantly changing and where the laws of cause and effect do not seem to apply”. For Le Moigne (1990: p. 3), a complex phenomenon is one whose representations are perceived as “irreducible to a finite model, however complicated, stochastic, sophisticated this model may be, whatever its size, the number of its components, and the intensity of their interactions...” No representation can be sufficient to capture all its dimensions, comments Avenier (2000: p. 4-5) who cites the example of the implementation of an organizational change process in an organization, supported by the intervention of consultants. Birkinshaw et al. (2008) come to the same conclusion with regard to management innovation processes. The following elements illustrate the complex nature of any administrative reform, as well as the presence of paradoxes in the dynamics of its different processes.

First of all, the axis of the agents of change in the model in Figure 1, as well as the roles that these agents play in the dynamics of the processes in this model, testify to the centrality of the “human factor” in any administrative reform process (see, for example, Brown, Colville, & Pye (2015) or Burnes (2004)). This human factor, whose lability of behavior and the multiplicity of their divergent and often conflicting interests are sources of paradoxical tensions amplified by the scarcity of resources to be shared and the competition between the internal structures of each administration as much as between the other administrations and institutions of the State. Indeed, Birkinshaw et al. (2008, p. 832) warn that these processes are typically complex. “A key feature of this framework is that it does not assume a simple left-to-right sequence of activities”, they write.

Secondly, ontologically, the public administration into which administrative reforms are introduced is organizations, in the sense that an organization is defined in management and organizational theories (Bartoli & Blatrix, 2015; Mazouz & Leclerc, 2008; Mintzberg, 1989; Morgan, 1989; Schoeneborn et al., 2019). However, these are organizations of a particular type that, compared to private organizations, possess specific attributes that reveal the permanence of paradoxical tensions (Emery, 2005; Emery & Giauque, 2005; Hudon & Mazouz, 2014; Mazouz & Gagnon, 2019; Ragaigne et al., 2018). These attributes include, for example, the ambiguity of their organizational objectives, the multiplicity of objectives, outcomes, and levels of results pursued, the permanent quest for both the global and the local, the fragmentation and centralization of the domain under consideration (education, health, security, water, energy, roads, transportation...), the position of the administration in the domain (i.e., its centrality, [example of the State and local authorities]), the management of the “political-administrative” interface, the structure of each administration, governance (involvement of stakeholders, political and decision-making accountability, justice and equity...), the identity of each administration as an organization (paradox of identity or organizational egocentrism), the permissiveness to deviance (Emery & Giauque, 2005; Kamdem, 2002; Hudon & Mazouz, 2014; Reynaud, 1997). Thus, when introducing a reform in an administration, it would be essential, observe Mazouz and Gagnon (2019: p. 10), take into account these specificities of public organization, and in particular “those associated with the mode of state ownership, the mission and objectives of general interest, common good or public services, the assessment of performance and the permanent quest for legitimacy.”

With respect to the practices characteristic of any administrative reform, as defined here, their implementation within the administrations that host them is confronted with the “exploration/exploitation” paradox (Benner & Turshman, 2003; Grimand et al., 2014, 2018a, 2018b; March, 1991; Sinha, 2015). As March (1991: p. 71) points out, exploration of new opportunities (innovate, take risks, experiment, flexibility, research...) is characterized by research and discovery, while exploitation of existing capabilities (refine, choose, produce, implement, execute, and seek efficiency...) focuses on methodical problem solving and development of the relevant administration. Exploration is associated with radical change and learning, while exploitation is associated with incremental change and learning through local research (Benner & Tushman, 2003). As March (1991: p. 72) points out:

In studies of organizational learning,the problem of balancing exploration and exploitation is exhibited in distinctions made between refinement of existing technology and invention of a new one. [...]. Finding an appropriate balance is particularly difficult by the fact that the same issues occur at levels of a nested system—at the individual level,the organizational level,and the social system level.

Ambidextry is the concept that is mobilized in the literature (Benner & Turshman, 2003; Grimand et al., 2014; Uman, Smith, Andersson, & Planken, 2020) to describe and analyze the processes put in place by the organization and to manage the “exploration/exploitation” paradox. In this regard, the concept of organizational ambidextry is mobilized to designate “the organization’s ability to reconcile exploration and exploitation orientations when using resources” (Uman et al., 2020: p.465). Uman et al. (2020) show that shared leadership of management teams is positively related to organizational Ambidextry. Similarly, they establish that NPM-inspired management control systems amplify this positive effect.

Considering management instruments and tools, they are characterized by a type of paradox that is highlighted by their simultaneously empowering character (e.g., they allow to engage, orient and justify the course of action) and constraining character (they privilege certain aspects of the situation underway) (Aggeri & Labatut, 2010; Gibert, 2009; Grimand et al., 2014, 2018b). According to Grimand et al. (2018b), an effective handling of these paradoxes must necessarily involve a revision of the traditional conception of these tools, which tends to apprehend them as “strongly coupled” systems, to develop a representation of their status as “weakly coupled” systems. The characteristics of these two types of coupling are summarized in Table 1.

Source: Culled from Grimand et al. (2018b: Abstract).

These theoretical foundations allow us to address the second part of our research question, which is the argumentative proposal of a framework for managing the paradoxes of administrative reform.

From the point of view of public policy analysis, the concepts of public policy, sector-based reference frame, global-sector report, and capacity to change are used here in a precise sense that should be recalled.

The concept of public policy is understood in Muller’s sense, i.e., a self-reflexive regulation apparatus (capacity for action on oneself) of the relationship between the parts and the social whole and by which: “through the State,society thinks,says and acts on itself without any reference point outside itself. It thinks and says itself according to the need to act on itself in order to take charge of its functioning” (Muller, 2015: p. 410). The sector-based reference frame is a dominant image of the sector as constructed by the actors to correspond to the perception that the dominant groups have of this sector (Muller, 1985). The Global-Sector Report (GSR) is the first element to be deciphered in order to understand a sectorial policy (Muller, 1985). It is the actors who give meaning, who “say” the needs that arise from the transformation of the GSR, says Muller (2000) see the role of agents of change in the model of Birkinshaw et al. (2008: pp. 831-839). In this conception of public policy, it is not possible to access the global (society) without going through the actors who will be “able to formulate, make legitimate and desirable a new vision of the world” (Muller, 2015: p. 412; See also Birkinshaw et al., 2008: p. 831-839).

The concept of capacity to change is mobilized with the aim of creating the optimal conditions for the effectiveness of the reforms introduced and their anchoring in the organizational culture of the host administration. It is defined as: “the deployment, combination and coordination of resources, skills and knowledge across different value streams to implement strategic objectives” (Fall, 2008: p. 20; Johnson, 2013: p. 32). The analysis here faces the challenge of confronting “prescribed work” and “real work” in its full complexity (see Burnes (2004) and Le Moigne (1990) above). In the Birkinshaw et al. (2008: p. 831) model, we are at the heart of the challenges of the “theorization and labeling” stage, which is “a social process whereby individuals inside and outside the organization make sense of and validate the management innovation to build its legitimacy”. “We expect that there will be an important rhetorical component associated with a successful management innovation”, observe Birkinshaw et al. (2008: p. 837). Soparnot (2009: pp. 111-118) identifies three dimensions of “capacity to change” that are related and co-activate: contextual, processual and reflexive. This last dimension contributes to making change a permanent state and, as a result, the time of change becomes permanent and is not confused with the time of the reform project.

We now clarify our epistemological position in this paper.

The definition of administrative reform emphasizes the deliberate, thoughtful and conscious nature of any changing action undertaken by the actors involved. These actions must therefore be based on a clear and sound argument formulated by the actors, with reference to an epistemological paradigm specified in advance (Avenier, 2011; Avenier & Gavard-Perret, 2012; Gioia & Pitre, 1990; Gioia, Corly, & Hamilton, 2012; Vandangeon-Dumez et al., 2018; Vas, 2015). We clarify our assumptions in relation to these actors and our epistemological posture in this article.

The postulate of the “socially situated” actor and “situated rationality”

The postulate of the socially situated actor, which implicitly builds on the postulate of methodological individualism (deliberation), admits that the actors involved in defining the “reality” that describes the process of administrative reform are characterized by their position or disposition in the context of action, institutions, and theoretical instruments that allow them to make “sense” of the realities they face or to take decisions (Boudon, 1986; Brown et al., 2015; Currie and Spyridonidis, 2016; Drevon et al., 2018; Friedberg, 1997; Journé & Raulet-Croset, 2008; Musselin, 2005). As Brown et al. (2015: p. 267) outline, “Sensemaking involves not merely interpretation and meaning production but the active authoring of the situations in which reflexive actors are embedded and are attempting to comprehend”. Taking into account this process of meaning construction, when he extends the notion of rationality to that of “situated rationality: the good reasons” Friedberg (1997: p. 225-229) observes that it is necessary to follow each decision-maker (or agent in action) in his or her situation in order to reconstruct the logic of his or her decisions (or acts) and thus to be able to find the “good reasons” that justify behaviors he or she displays that are apparently irrational or deviate from norms. The approach seeks the mechanism (mediation or theory) that allows us to understand a given behavior in the production of meaning and the discovery of characteristics (material and immaterial, formal and informal) of the immediate context of action (Aggeri, 2017: p. 31; Boudon, 1986: p. 294; Friedberg, 1997: p. 229; Journé & Raulet-Croset: 2008, p. 32). In the analysis or management of administrative reform, it is a matter of studying and understanding the behavior of an actor as an active and reasonable adaptation to a set of constraints and opportunities perceived in his context of action.

Our epistemological posture: The pragmatic constructivist epistemological paradigm

In this paper, and considering the “knowledge” dimension of intervention in administrative reforms, we place ourselves within the framework of the pragmatic constructivist epistemological paradigm (PCEP). According to this paradigm, the reality remains unknowable in its essence because we cannot reach it directly without the mediation of our senses, experience, language or intentions (Allard-Poesi & Perret, 2014; Avenier, 2011; Avenier & Gavart-Perret, 2012; Le Moigne, 1995; Vas, 2015; von Glasersfeld, 2001). Our choice to adopt the PCEP is justified by two main reasons: 1) ontologically, the PCEP does not deny a priori the existence of a reality in itself. It thus offers a capacity to formulate a wide range of working hypotheses, as we have done above with the postulate of the “socially situated” actor, as well as with the systemic principles retained. Similarly, it offers a very great capacity to integrate various models of reality representation developed in other epistemological paradigms since knowledge of a phenomenon is both anchored in the phenomenon studied in its system of action, and dependent on the persons (researchers, consultants, analysts, auditors, internal and external stakeholders of the administration) who study it; 2) at the methodological level, any research and intervention method is eligible in the PCEP, including those used in the two perspectives of analysis that we have retained in our theoretical perspective, and the use of different techniques for collecting and processing quantitative and qualitative information.

The presentation is divided into four points: 1) the axes of the definition of the proposed framework; 2) the proposed conceptual framework; 3) the paradox-gene- rating effects; and 4) a discussion of the proposal.

Three related and co-activating axes

From the synthesis of the first grid, it appears that while there are many approaches to change, there is no dedicated theory of change (Autissier et al., 2018; Engelberg, 2021; Soparnot, 2009; Pesqueux, 2015). A search by Engelberg (2021) on the term “Models of Organizational Change” and analyzing nearly 395,000 works published in the period from 2017 to 2021, came to the conclusion that “the research conducted over the past 50 years has not fundamentally developed anything completely new; rather, it has provided us with the clarity needed to better understand what was developed many years ago...” (p.1). Based on our PCEP posture and following our definition of administrative reform, we propose to articulate the analysis and management of the workflow of conducting any administrative reform around three axes:

- The axis of organizational change: Change (reform) is considered here as an organizational phenomenon or as a response to a problem (See the “motivation” activity in Figure 1). How is it understood at the epistemological level? What roles are recognized for the internal and external actors of change? Who are they and what is their degree of influence? The focus is on the theoretical foundations of change at the ontological and methodological levels as defined by the researcher or consultant (see for example: the four families of types of organizational change by Ven and Poole (1995); the change management method proposed by Autissier and Moutot (2003); the one-dimensional models, including the political model by Soparnot (2003); the management innovation model by Birkinshaw et al. (2008). From the perspective of cognitive analysis of public policies, the administrative reform process has a dual function of structuring the reference framework of the intervention (for example, the GSR, the reference framework of the power of the actors…) to be carried out in the unit selected and of driving change in order to achieve the reform objectives pursued.

- The axis of the process of the changes to be introduced: This axis deals with the problem of the origin of the change (“How to produce change?”). Whether it is a question of institutional isomorphisms (DiMaggio & Powell, 1983), organizational heteromorphisms (Mazouz et al., 2005), identity projects (Rondeaux, 2008), invention in the sense of radical change (Birkinshaw et al., 2008), diffusion of the myth of managerial utopia (Metzger, 2000, 2001), or organizational learning (Argyris & Schön, 1978; Senge et al., 2000), administrative reform actions arise from the transformation of the relationship between the global and sectoral frame of reference as defined above. This process always depends on complex patterns of reciprocal connectedness between individuals and organizations. See the key role of “mediators” in Muller (1985); or the cross-cutting role in relation to all activities in the Birkinshaw et al. (2008) model. The examples of the change matrix proposed by Autissier and Moutot (2003) and Rondeau (2008) are illustrative in this respect.

- The axis of the challenges of the effectiveness of the expected results: As we have stated, reform processes are complex and recursive; it would be unrealistic to expect to achieve meaningful results by assuming otherwise. This axis focuses on monitoring the implementation of planned actions and tracking the embedding of the results achieved in the organizational culture. See, as an illustration, the challenges of the “refreezing” step of Karl Lewin’s three-step (unfreezing, moving, refreezing) model of change in which these challenges also relate to organizational norms, policies and practices (Burnes, 2004: p. 313). The aim here is to provide answers to the questions of How do you bring about successful change? How to determine if a change has really occurred? How to create the optimal conditions for anchoring the results obtained? The axis highlights three types of issues that challenge the dual problem of “change management capacity” and “managing the capacity to change” (Soparnot, 2005), namely: 1) the adoption of a national structure in charge of administrative reform (NSRA) in a cross-cutting coordination role within the government; 2) the management of this structure and that of the administrations hosting the reforms; and 3) the professionalism of the experts credited with specific “know-how” who are going to intervene at the operational center of the NSRA and as members of the project team of the administration hosting the reform. One of the implications of complexity theory drawn from the literature by Burnes (2004: p. 318) suggests that “In achieving effective, order-generating rules have the potential to overcome the limitations of rational, linear, top-down, strategy-driven approaches to change”. The three contextual, processual and reflexive dimensions of the concept of “capacity to change” and their mutual influence dynamics illustrate the importance and necessity of this axis which includes actions corresponding to the activities of implementation, on the one hand, and theorization and labeling, on the other hand in Birkinshaw et al. (2008).

The proposed conceptual framework

We summarize this theoretical discussion with the diagram in Figure 2, which represents our proposed conceptual framework for intervention and analysis of paradoxes in administrative reform.

Potentially paradox-generating effects to consider

When applying this framework to understand the reality of the workflow of an administrative reform in order to analyze the phenomena of paradoxes, it is necessary to take into account several effects that influence the mechanisms of communication, sensemaking, interpretation of the observed phenomena, analysis of the collected data and learning, and that are likely to generate paradoxical tensions. We have retained the following four categories of effects, which are integrated in Figure 2.

1) Situation effects. Including: the position and disposition effects (see situated rationality) and the perspective effect induced by the positive or negative consequences, as perceived by an actor, of a rationalization practice (Boudon, 1986). This effect exposes the actors to the trap of denial or concealment according to their strategies of the time and the particular location.

2) Communication effects. Including: the attribution of meaning and value to what is considered the reality of the administrative reform workflow; the structure of the message as a source of paradoxes; the paradoxical message is given verbally or covertly (Bernard, 1998; Gramaccia, 2011; Gaulejac (de), 2010; Keller, 2004; Moisander et al., 2016; Mucchielli, 2006; Schoeneborn et al., 2019; Watzlawick, 1978). As an example, the reform host organization may find itself, as Gaulejac (de) (2010: p. 90) points out, practicing “exclusion” to improve its performance, causing, for example, the identity paradox (see Jarzabkowski et al., 2013). This is also the case for paradoxes resulting from the rhetorical, constitutive, and narrative power of emotions in institutional work at the discursive level (Moisander et al., 2016).

3) Epistemological effects or E-effects. These relate to the epistemological framework that the analyst uses or that underpins the global referential of the committed reform policy (Boudon, 1986; Foudriat, 2013; Muller, 2015; Vas, 2015; Watzlawick, 1978), and reflect theoretical choices made by the analysis of the specialists who drafted the global referential more or less consciously. Very often, these specialists admit as such, the frameworks they use, without discussing their epistemological foundations. Some models contain invisible (non-explicit) hypotheses that escape the theoretical framework adopted (see the examples of pragmatic paradoxes). For example, when a speaker invokes the concept of “bureaucracy” in an argument, he or she implicitly admits the stability and unpredictability of the analyzed phenomena and at the same time refutes the hypothesis of emergent phenomena of the methodological holism assumption. Hence the importance for the researcher or consultant to clarify his or her own epistemological stance in advance and to try to decipher those of the other stakeholders in the reform process (see the dialogical approach).

4) Learning effect: In the sense of the issue of change and transformation of the administration that hosts the reforms, on the one hand, of the learning and adaptation capacities to its environment to keep up with the ongoing developments, on the other hand (Birkinshaw et al., 2008; Burnes, 2004; Senge et al., 2000). Paradoxes, in fact, offer opportunities for further reforms and innovations (Morin, 2011: p. 132, 297). This learning concerns all actors in each reform project and relates to all activities. At the moment, point out Birkinshaw et al. (2008: p. 839) “we know little about the relative effectiveness of different sequences of activities [in an administrative reform process], which makes it difficult to offer any coherent advice to managers about how to improve the quality of their interventions [along the actual workflow]”.

Tools can be used to diagnose the organizational processes of communication and learning in order to identify the mechanisms that generate the paradoxes as presented above. For example, Mucchielli (2006) proposes a survey guide based on the “actionist analysis” approach to diagnose the overall communication process of the organization. Senge et al. (2000) discuss the concept of “organizational learning architecture” from which a table can be developed that could be called the “learning profile” of the administration hosting the reforms, linked to the dimensions of “organizational capacity to change”, which will be used to diagnose the dimension of organizational learning processes (Ngouo, 2018: p. 285). This profile is articulated in two dimensions: 1) the theories, models and tools used for reflection, conceptualization and conversation within the organization; 2) innovations in learning infrastructure which are new initiatives or measures that the organization takes to make available to staff, the means that will allow them to invest in the organizational learning process mobilized. These means may include strategies for moving from a traditional design of management tools and instruments to an approach that emphasizes “Weak coupling” systems (see Table 1).

Figure 2, and the application tools mobilized, whatever they may be, serve as a background for the work of interpreting and formulating the “meaning” of the paradoxical phenomena that are based on relating the phenomena to their context in an administrative reform process approached from the inductive perspective.

Discussion about the application conditions of the proposed framework

Any reform has multiple effects on the behavior of the actors and recipients concerned, and in particular, the effects of resistance to change when the expected results are perceived as threatening to their professional achievements or interests. These behaviors of resistance to change can manifest themselves in various ways widely documented in the literature (see for example: Abouem à Tchoyi & M’Bafou (dir), 2013; Autissier & Moutot, 2003; Burnes, 2004; Mazouz & Gagnon (dir), 2019; Ngouo, 2008, 2018; Ondoua, 2015; Pichault, 2013; Rondeau, 2008; Rondeaux, 2008; Saner, 2005; Soparnot, 2003). These undesired effects create many paradoxical tensions that can profoundly and permanently compromise the achievement of the expected results.

Managing these perverse effects of any administrative reform project is the main challenge to be taken up by the managers of the project’s host administrations, in order to ensure the optimal conditions for achieving the expected results, and above all for anchoring the results that have been effectively achieved in their administrative culture. In reference to the paradigm of the “capacity to change”, we situate this discussion at the heart of the real work of administrative reform; where locally the actors reconstruct their organization in a way that is not prescribed and not planned in advance (Journé & Raulet-Croset, 2008). From this viewpoint, Niesten and Stefan (2019: p. 148) emphasize the importance of the relationship of trust between actors by observing that this factor can cause a dynamic of the evolution of paradoxical tensions that moves “from a virtuous to a vicious cycle, but only when the partners in a trust relationship are confronted with formal control tools that can stimulate the opportunistic behavior of one of the partners”. Gramaccia (2011: pp. 263-265) observes that trust and loyalty are the characteristics of a shared relationship to risk and to collective time, and that the affirmation of trust serves as a model for action in contexts of uncertainty (See also: Kamdem, 2002; Lee & Ryzin, 2020; Nguyen, Leclerc, & LeBlanc, 2013; Niesten & Stefan, 2019). In addition to this trust factor, the literature highlights the negative impact of two other bureaucratic attitudes of actors mobilized in collective action. These are the refraction to flexibility and the resistance to double-loop learning. Indeed, innovation or reframing requires an attitude of flexibility oriented towards the search for a compromise between a diversity of procedures and multiple conflicting interests, both institutional and individual in the context of public administration (Feldman & Pentland, 2003). Flexibility is also essential for the development of the networking culture. In the same way, a strong orientation towards the “double-loop” organizational learning process (Argyris & Schön, 1978; Birkinshaw et al., 2008; Mazouz & Gagnon, 2019) is essential to manage paradoxes.

In reference to the Epistemological effect, as highlighted above, let us recall with Mintzberg (1989: p. 198) that any bureaucracy is ontologically resistant to change, because, quite simply ignoring the real work which is the essence of the management situation, “[Its] spirit is to create a way and to stay on it by making sure that everything that can result from it has been wanted? Bureaucracy means no surprises”. The very strong identity of public agents to the “bureaucratic thought” makes them almost incapable of questioning its basic postulates, thus exacerbating the preference for a relation of certainty to reality, to the doubt that comes from its complexity. It is therefore a matter of any manager and of the internal actors of any public administration, in an attitude of reflexivity, to leave the bureaucratic logic and to step back from the usual injunctions and the various usual bureaucratic techniques in order to engage in a distanced and in-depth analysis of the managerial dispositifs as defined by Aggeri (2017: p. 43), as “the arrangement of heterogeneous elements—discursive, cognitive, material—by managers aiming at framing, orienting and guiding the conducts of subordinates towards assigned ends”. The effectiveness of the administrative reforms undertaken in each administration is at this price.

From the point of view of organizational analysis and public policy, participation has a double paradigmatic value: an instrumental paradigm for its mechanism and a rationalist paradigm, in the sense of “situated rationality”, for the effective, efficient and relevant involvement of actors.

Participation seen as a temporary work mechanism at the service of a policy, as described by Friedberg (1997), has an instrumental character, in that it is necessarily self-interested, in particular through the selection of actors, which combines a logic of participation with the mobilization of the actors’ skills and a logic of representation of interests, which is done on an egalitarian and defensive basis (Dietrich, 2000; Friedberg, 1997). This leads to recourse to the mode of regulation through hierarchical control to manage the various types of tensions that result (Bennani et al., 2021; Hudon & Mazouz, 2014; March & Simon, 1969; Ogien, 2001). Participation is seen as a process that must facilitate the self-determination of actors, stakeholders, so that they become involved in the co-production, facilitation of change, extension and dissemination of the results of the reform project undertaken. The involvement, commitment and adherence of actors to co-production depend very closely on the ability of the administration hosting the reforms to legitimize its objectives and strategies to the stakeholders of the reforms undertaken and to co-construct with them the notion of “the general interest” (Trosa, 2017; see discussion of the proposed framework in Figure 2). The analysis of the “rational” behavior of each actor must then take into account numerous constraints (Dietrich, 2000; Emery & Giauque, 2005; Giauque, 2004; Grimand et al., 2014, 2018b; Mazouz & Leclerc, 2008; Mazouz et al., 2012; Mazouz & Cohendet, 2013; Mazouz & Gagnon, 2019; Reynaud, 1997; Saner, 2005; Watzlawick, 1978). Reynaud (1997: p. 113) states in this respect that “insertion in a decision-making circuit or the influence on a complex decision cannot easily be translated into quantifiable measures” (See for example, the theme of motivation towards public service, Giauque (2004); Perry and Wise (1990); Vandenabele & Hondeghem (2004)).

The application of the participation paradigm in conducting administrative reforms thus leads to a combination of the three theories of management, stakeholder, and paradox (Pinto, 2019). Stakeholder theory is used to take into account and manage “the multiple and often conflicting objectives of different stakeholders, while paradox theory suggests ways to achieve the challenging outcome of simultaneously satisfying conflicting priorities” (Lindgreen & Maon, 2019: p. 140).

The success of any administrative reform—and therefore of the introduction of RBM—also depends on the involvement, in order to co-create the value that will benefit everyone, of the internal units of the host administration, as well as of the other national administrations or international organizations that are stakeholders in the reform (Prime Minister’s Office, beneficiary ministries, ministry in charge of planning, ministry in charge of finance, national structure or ministry in charge of the administrative reform, international organizations…). Taking this dimension into account calls into question the quality of the relations that these different administrations have with each other. Indeed, these relationships are not immune to the reflex of each stakeholder to unilaterally derive individual benefits and appropriate the co-created value. This results in paradoxical tensions between “co-creating value” and “capturing co-created value”, illustrated, for example, by the opposing strategies required to create and capture value respectively. As defined by Niesten and Stefan (2019: pp. 234-235), value co-creation refers to collective processes that generate common benefits shared by all stakeholders in inter-administration relationships (IAR), with one or more other organizations. Value capture is the ability of stakeholders to unilaterally extract individual benefits and appropriate relational rents. Drawing on Smith and Lewis (2011), Niesten and Stefan (2019) propose a conceptual framework to analyze this type of tension. We propose it in Figure 3 as a tool that can be adapted to the conduct of administrative reforms by applying the participation paradigm.

The results-based management (RBM) approach is one of the emblematic axes of NPM reforms (Bezes & Mussilin, 2015; Drumaux, 2011; Emery & Giauque, 2005; Mazouz & Leclerc, 2008; Mazouz (dir), 2008; Mazouz (dir), 2017; UNDP, 2009). Following a logic of institutional isomorphism (constraint, mimetic, normative), its introduction in an administration brings a radical change aiming at moving from a culture of management by means (priority to efficiency) to a culture of management guided by results (priority to effectiveness and optimization of resources). The sole mode of regulation of the administrative workflow by hierarchical control should be replaced by a hybrid mode including autonomous regulation, regulation by negotiation and regulation by the public service users’ demand. We choose this reform to illustrate the analysis of the participation paradox.

The introduction of the reform is carried out in phases, each of which includes several stages. In this regard, and emphasizing the importance of “real work” in comparison to “prescribed work”, UNDP (2009: p. 23) concludes that “Good RBM is an ongoing process. This means that there is constant feedback, learning and improving. Existing plans are regularly modified based on the lessons learned through monitoring and evaluation, and future plans are developed based on these lessons.”

For the purposes of this presentation, we have identified three phases, linking each of them to the effects shown in Figure 2.

v Project planning, which corresponds to interactions 1 and 2 (see E. P. C. DISP effects). The primary goal here is to correctly identify key stakeholders who may have a strong interest in or ability to influence the planning object.

v Monitoring implementation, which corresponds to interactions 3, 4 and 5 (see E. C. DISP effects). This involves bringing the theoretical, “prescribed” world of planning closer to the “real” world of stakeholders, as socially situated actors, by mobilizing them as best as possible.

v Consolidation of the results obtained, which corresponds to interactions 1 and 6 (see the effects E. P. DISP. C. L). Consolidation places us at the two levels of effects and impacts of the RBM results chain. Stakeholders need to be intensely mobilized for the ownership of reform outputs following a logic of developing the “capacity to change”. [See the discussion above on the optimal conditions for applying our proposed framework].

The emphasis on ownership of reform outcomes raises another type of tension reflected in the paradox of time: the time of the project, which must be completed by a predetermined deadline (prescribed work); the time of the reform, which takes into account the contingencies of the real workflow of the reforms and focuses on the embeddedness of the changes introduced to promote a new administrative culture. The time of reform places RBM in an always-on process and in the space of change that is not necessarily linear (see the iterative processes in the trial-and-error activities of the Birkinshaw et al. (2008) model, or the dynamics of Karl Lewin’s three-stage model of change, or the U-shaped change models in Mahy & Carle (2015)).

Table 2 summarizes our analysis by identifying, by phase, the roles and responsibilities of stakeholders in terms of the commitments they make during information and communication activities related to the phase under consideration, the paradoxes to be managed and the levers of action that can be mobilized to manage these paradoxes. The elements in the third column are the diagnostic hypotheses of the anticipated paradoxical tensions that may arise in each phase in the field. Those in the 4th column are suggestions for strategies to manage these paradoxes, if they are confirmed in specific cases (contextual analysis) in the field, in order to capitalize on their creative potential. The formulation of these hypotheses and strategies is based on current knowledge of management theories, stakeholders and the paradox approach, particularly the theories of “organizing” (Smith & Lewis, 2011) and those relating to the concept of “management situation” which mobilizes the concepts of “sensmaking”, “strategizing” and “decision making” (Journé & Raulet-Croset, 2008).

The table refers to the polyphonic leadership style as a lever to manage paradoxes in the monitoring phase of project implementation. Indeed, Pichault (2013) distinguishes two extreme types of leadership style: the panoptic style, which leads to seeing and controlling everything and leaves little room for the confrontation of ideas, and the polyphonic style, which leads to the dialogue of several voices. The polyphonic style allows for a diversity of rationalities; it is likely to better deal with the conflictual phenomena that arise during change projects (see dialogical processes).

Regarding the management of the paradox of co-creation and value capture, Niesten and Stefan (2019) propose a typology of factors rendering tensions salient and factors spurring virtuous cycles. We offer a synthesis of this typology in Table 3 to guide the search for appropriate strategies to manage this paradox.

Source: Made by the author.

Source: Synthesis and adaptation of the author based on Niesten and Stefan (2019: pp. 239-245).