How “shared” leadership comes into being: managers’ sensemaking in networked child welfare policy Open Access

Amidst intractable public problems like climate change, increased demand for healthcare due to ageing populations, and concomitant workforce shortages, the quest for alternative leadership narratives is increasingly at the forefront of public and scholarly debates (Haslam et al., 2024). Superseding leader-centric capacities or persons, plural leadership theorizations have gained traction in leadership scholarship – giving rise to labels such as “systemic,” “collective,” “relational” and “multi-level” (Denis et al., 2012; North-Samardzic et al., 2024; Needham et al., 2025; Kuipers and Murphy, 2023; Ospina et al., 2020). Such notions transcend the agency and distribution of leadership roles and influence amongst a group of actors (Denis et al., 2012). In particular, an emergent and relational understanding of leadership emphasizes its unfolding nature across boundaries and between actors (Uhl-Bien, 2006; Ospina and Foldy, 2010). Here, tying the relational webs in which diverse actors work on issues that move beyond each one’s primary (institutionalized) responsibility is seen as vital (Clifton et al., 2020; van der Scheer, 2023).

Leadership scholars emphasize the importance of studying how leadership narratives aimed at shared responsibilities are translated from general notions into concrete meaning, and how these narratives generate reciprocal relations and commitments necessary for problem-solving (Raelin, 2011; Foldy et al., 2008). The quest for leadership narratives aimed at shared responsibilities instead of siloed organization-centered responsibilities requires a change in how actors view their own and shared positions, scope of interference and responsibilities vis-à-vis problems (Goffman, 1974). This involves sensemaking performed through interactions, and broader frames of reference in which they are embedded (Sutherland, 2016; van Hulst et al., 2024). La Grouw et al. (2024) observe that “[…] central to these interactions are the discourses the actors use, and other participants' interpretations of those discourses” to inhibit or enable problem-solving. Studying how actors make sense of a leadership narrative rooted in shared responsibilities may offer insights about the dynamics in policy transformations that intend to overcome siloed ways of working (Uhl-Bien, 2006; Waring et al., 2022). This is especially significant in policy domains that rely on interorganizational networks, like social welfare and healthcare (van Duijn et al., 2021), which is the central concern of this article.

We refer to the quest for a leadership narrative aimed at shared responsibilities as a quest for “shared leadership.” Although the importance of studying shared leadership has been addressed in (public) leadership theory, it mainly refers to team-level interactions within organizations (e.g. van der Hoek and kuipers, 2024; van Osch et al., 2024). For instance, van Osch et al. (2024) view shared leadership as a dynamic process where leadership is distributed across team members. Pearce (2004) describes shared leadership as a process of “mutual influence” among team members and beyond formal authority. This contrasts with an emergent understanding of shared leadership in this study that focuses on alternative ways of interacting and organizing across team or organizational boundaries (Ospina and Foldy, 2010). This may be of eminent reality for managers as their work increasingly goes beyond organizational practice in networks (van der Woerd et al., 2024). They must deal with co-existing (conflicting) logics like hierarchy or self-organization while traversing between the organization and the network(s) they participate in (Ospina and Saz-Carranza, 2010), continuously “translating” different (and changing) narrative elements like identities (Waring et al., 2022). This traveling may give managers a sense of “in-betweenness”: a state where leadership is not straightforward (Ospina and Sorenson, 2006). On the one hand, they work in hierarchies within organizations with a clear (control) structure and purpose, and on the other hand, they work in multiple networks with a clustering of activities with various service forms and unclear leadership roles between organizations (van der Scheer, 2023).

In this study, we seek to unravel how managers amidst a policy transformation toward a networked model of child welfare in the Netherlands make sense of shared leadership (Blanken et al., 2023). Operating in a complex policy environment (e.g. a municipal responsibility, jurisdictional boundaries, tender procedures) and confronted with budget cuts and increasing waiting lists, managers feel an urge to define what it means to share leadership with managers of different but collaborating organizations. This intentionally moves beyond redefining leadership positions or institutional mechanisms like performance metrics (Overman and Schillemans, 2022). With managers, we refer to end-responsible executives and portfolio-holding directors as distinct management levels (and refer to each specifically in the results section). We pose the following research question:

In answering this, we draw on an ethnographic study (from March 2023 until January 2025) located in an urbanized Dutch region where child welfare providers created a network to realize integrated child welfare services (care, education and social support) for youth populations. This was also policy-induced since local governments commended the clustering of services with one front office for youth populations. Fueled by an urgency to consolidate their network ambitions, managers of these organizations came together during dedicated reflective sessions, purposefully decoupled from regular network meetings (Visser, 2024). Here, they put the concept of shared leadership to the test and sought to dissociate from existing managerial attitudes, reflecting on how to transcend their own (felt) responsibility. The fieldwork here consists of participant observations of managers’ interactions during these sessions, complemented and triangulated with multiple rounds of (digital) interviews, numerous informal conversations and a document analysis of network, organizational and policy notes. The ethnographic approach employed here is in particular suitable to grasp how managers understand themselves and their position, and those of others, and where interactions take place, in their quest for shared leadership (Bucher and Langley, 2016; Sutherland, 2016). This research aims to add to ongoing debates around the quest for shared leadership in policy domains that rely on networks (Ospina and Foldy, 2010). By “plunging” into the process of how managers develop an understanding of shared leadership (Vandenbussche et al., 2020), we emphasize the processual, provisional and political nature of how managers must navigate through (conflicting) organizational traditions with routinized ways of working and networked ones.

The following section theorizes leadership as unfolding between the organization and the network(s). This is followed by an elaboration on how to study managers’ sensemaking of shared leadership. After introducing the case and the ethnographic approach, this article describes managers’ constructions of shared leadership, and how they start to engage with a new understanding of this. The study closes with a reflection on how the findings enrich insights in leadership and governance literature on dynamics in policy transformations.

Given the interdependency among actors across organizational boundaries for public service delivery (Crosby and Bryson, 2010), leadership phenomena in inter-organizational networks are receiving increasing attention (North-Samardzic et al., 2024). Interorganizational networks – henceforth referred to as “networks” – can be understood as an alternative model of service delivery compared to top-down forms, as it mobilizes (knowledge) resources that are spread across organizational boundaries (Ferlie et al., 2011). Scholars have given attention to the type of leadership needed (O’Leary and Bingham, 2007), or leadership processes within the network, as these are considered part of network functioning (Crosby and Bryson, 2010). In particular, literature on “network leadership” considers the emerging nature of leadership in a network (Endres and Weibler, 2020). Such inquiry offers valuable insight into the meaning of leadership in networks as opposed to groups or organizations – which is timely, given the upsurge of networks amidst policy transformations (Denis et al., 2012).

In (public) leadership theory, various forms of leadership, such as “integrative” and “collective” have been used to study leadership in inter-organizational or “boundary-crossing” settings like networks (Ospina et al., 2020; Parkkinen, 2025). Such accounts have resulted in valuable frameworks to identify “factors” or “characteristics” of particular types of leadership needed in interorganizational settings (Parkkinen, 2025). This study builds on these endeavors by focusing on the “relational interplay” and micro-level processes of how managers search for a form of leadership that addresses shared responsibilities in a network setting (Raelin, 2011; Waring et al., 2022). We thus emphasize managers' deliberate quest for a new understanding of leadership that goes beyond organizational team-levels (cf. van der Hoek and kuipers, 2024). By taking managers' interactions as a starting point, this study moves away from reified leadership notions as “whole” or coherent and offers a mundane understanding of how managers make sense of a leadership narrative that transgresses their (institutionalized) responsibility (La Grouw et al., 2024; Waring et al., 2022). This may refine how actors work on problems beyond their primary responsibility (Ospina and Foldy, 2010), how leadership is conceptualized at the level of networks (Ospina and Saz-Carranza, 2010) and the multi-level nature of leadership in relation to leadership functioning (Kuipers and Murphy, 2023).

Despite its high promise, networks are not necessarily stable and discrete entities with shared interests, but dynamic; they interfere with and must unfold alongside organizational practice and interests (van der Woerd et al., 2024). While managers may have a clear position within their organization, their role in networks is often fuzzier. Unequal social positions; ill-fitting responsibility schemes; conflicting decision-making processes and representation forms; and ongoing negotiations over purposes, boundaries, values and rules of engagement may result in a perceived lack of leadership (Zonneveld et al., 2024; Peeters et al., 2023). As a result, it may be that no manager is taking up a leading role, or managers are hindered from doing so (van der Scheer, 2023). This sense of in-betweenness may become explicit in networks where actors attempt to address this lack of leadership (Ospina and Sorenson, 2006). Managers must navigate through a rather unstable context with the rearrangement of governance structures and working practices (van Duijn et al., 2021). Managers must foster this based on reciprocity in order for the network to flourish, while also honoring organizational positions and identities (including their own) (Ospina and Foldy, 2010). Amidst such a tension, it is particularly relevant to study managers’ sensemaking of a new leadership narrative.

Sensemaking is a social process and shaped through interactions by which managers give meaning to events, especially in uncertain times when existing working routines do not fit with the network (van Hulst, 2012). These events are situated in temporal, spatial and social settings, and shape managerial action (Weick, 1993). Sensemaking is about the processual nature of how managers navigate through competing beliefs rooted in different traditions within a socio-historical context (Bevir and Waring, 2020). Analyzing how managers make sense of shared leadership “[…] focus[es] on the manner in which stories evolve and how they co-evolve traveling through narrative landscapes, being told, retold and edited” (van Hulst et al., 2024, p. 17). Shared leadership might come with evolving interpretations; sensemaking has no clear stop (Weick, 1993). In how shared leadership as a narrative is told and the elements that are foregrounded, managers “emplot” the past (Czarniawska, 2004). Shared leadership may also be political as it conveys meanings and underlying moral imperatives (van Hulst, 2012). Making sense of shared leadership may lead to “[…] contests among diverse and contingent meanings rooted in different traditions and the dilemmas faced by actors in particular contexts” (Bevir and Waring, 2020, p. 11). It furthermore may reconfigure power positions and boundaries as certain actors or meanings are included while others are excluded. We hence critically approach how managers narrate their actions, and those of others. Taken together, these components provide a foundation for analyzing managers’ sensemaking of shared leadership (van Hulst et al., 2024; Weick, 1993): how managers interpret events within a socio-historical context; how they enact shared leadership while making sense of it; and how they construct their identity, which is shaped by and shapes sensemaking.

Although the Dutch welfare state traditionally comes with various social investment policies, its scope is increasingly under pressure. Citizens are more and more encouraged to take up care and support tasks (Goijaerts, 2022). In 2015, the Netherlands decentralized social care services (including child and family services) from national to local government levels to enable tailored service delivery by municipalities and practitioners (Visser and Kruyen, 2021). The consequent Youth Act (in Dutch: Jeugdwet) specified that services should aim to be customized. Municipalities must contract both generalist and specialized organizations to provide child and family services in close cooperation with appointed neighborhood teams (Sabel et al., 2023). As a consequence, child welfare organizations that often relate to many municipalities must invest in numerous tender procedures and personnel to make convincing bids. While they must compete with each other for long-term contracts, they also must develop coalitions to convince municipalities that they are able to provide for the needs of children and their families.

Amidst increasing social unease about child welfare bureaucracy, siloed services and the abrupt termination of care after children become adults, current child welfare policy is aimed at child service networks (e.g. the clustering of expertise to provide tailored services for children and their families) – supported by the Youth Reform Agenda (Hervormingsagenda Jeugd) as a policy plan (Nooteboom et al., 2021). These networks are considered by many as promising as they aim to exceed the expertise of individual organizations (Blanken et al., 2023). Since these organizations often serve the same clients, working as a network must prevent children from falling between organizations or having to wait multiple times for intakes, being required to share their story over and over again. However, these networks are cumbersome as they require the reconfiguration of work patterns at municipal and organizational levels amidst an institutional context that is geared for competition instead of collaboration.

This article draws on a qualitative case study from March 2023 until January 2025, located in the urbanized Dutch Riverbank region (a pseudonym, for anonymity reasons) consisting of more than 300,000 citizens. The case is here referred to as the Child Welfare Network (CWN) (also a pseudonym, for anonymity reasons). CWN is a child service network in which four child welfare organizations provide care to local child and family populations. The end-responsible executives of the organizations together form CWN's board. This board supervises the CWN director, who works alongside a group of directors that work for the four organizations. These directors supervise middle managers and local service teams in which municipal workers also take a seat. Table 1 presents CWN's participants and their characteristics. These organizations differ in terms of child welfare expertise (e.g. general, specialized), structure (e.g. clinics, institutions) and scope (e.g. neighborhoods, regional or national). For Dikeland and Coastland (also pseudonyms), the impact of CWN on their overall budget is extensive, while for Sealand and Duneland this is less. All are aimed at child welfare in the Riverbank region, although Sealand and Duneland also operate beyond the case.

The tender procedures for child services in two sub-areas of the Riverbank region stimulated the constituent organizations to intensify their efforts toward network formation. From October 2022 until March 2023, they worked on their plans and subsequently won the tender procedure of one of the two sub-areas in which they competed. They institutionalized their ambitions in CWN, which was established in January 2024. Given their long-term responsibility until 2032, managers initiated a search for shared leadership in reflective sessions to discuss the consequences for organizational and managerial practice. These serve as the basis for this study's case engagement.

Central to the data collection in this study is an ethnographic approach to study how managers make sense of shared leadership (Schatz, 2009). Table 2 presents all data collection moments and Figure 1 overviews when these took place.

Firstly, data collection entailed 33 h of participant observations of managers' interactions during dedicated reflective sessions, including pre- and post-meeting (Visser, 2024). The sessions took on two different formats with particular compositions. First, eight sessions with a timespan of 4 h with end-responsible executives (between June 2023 and September 2024), and second, two sessions with a timespan of 4.5 h with 12 portfolio-holding directors who work for CWN's organizations, including CWN's director (March 2024 and July 2024). During each session, the researchers took unstructured field notes individually. These notes were combined and worked up into observational reports afterward (e.g. distinguishing thematic discussions, elaborating speech quotations) (Emerson et al., 1995). The researchers used booklets to keep track of interactions, enabling less invasive engagement with attendants and a greater focus on managers' sensemaking of shared leadership. The second author guided the meetings, while the first author mostly functioned as record keeper. After every session, all attendants received a summary of what was discussed, including agreements. This summary helped at the start of the following session as a basis to reflect on. Our position as researchers here can be understood as an intervention to stimulate “embedded reflection” among managers, while also sharing our own reflections (Duijn, 2018). Importantly, we organized time after the sessions had taken place to critically reflect on our own experiences of the sessions.

Second, parallel to the observations, this study involved semi-structured (online) interviews with executives (n = 18) and directors (n = 15) during multiple rounds, mostly in-between pre-planned observational moments. Each reflective session was preceded by an (online) interview with two executives or directors. The interview questions started by asking how managers experience ongoing CWN developments. We elaborated on the tensions they experience in their quest for shared leadership both on network and organizational levels. We asked how managers interpret particular events like network decisions, and how they follow up on agreements made. The interviews provided a more detailed contextual understanding of how managers try to define what shared leadership means to them prior to the reflective sessions. This fostered our understanding of managers’ sensemaking of shared leadership over time (Weick, 1993). The interviews lasted between approximately 45 min and 1.5 h. Most of the interviews were carried out digitally, but some were physical. All interviews were recorded with permission and transcribed verbatim. If recording was difficult to realize, for instance via telephone, notes were made and added to the digital case documentation.

The observations and interviews were complemented with a document analysis of child welfare policy documents produced by local and national policymakers, newspaper articles and (website) texts of CWN. These documents helped clarify CWN's scope, the differences amongst organizations and how the child welfare domain functions.

Furthermore, this study is built on numerous informal conversations with managers pre- and post-meeting, or via telephone or e-mail. This helped to clarify the statements made and to provide ongoing discussions with subtle background information.

This combination of data sources enabled the capture of meanings managers bring from one situation to another, and how they interact from various perspectives (Sutherland, 2016). For all data collection, respondents gave consent to the research. All data sources were collected in Atlas.ti software.

The data analysis followed an iterative approach (van Hulst and Visser, 2025), going back and forth between preliminary findings and leadership studies that converge with the public management field's interest in networks (e.g. Ospina and Foldy, 2010). The data analysis involved several steps taken carefully.

The first part of the thematic analysis was to “familiarize” ourselves with the data by ordering the observational reports and interview transcripts into broad strokes of developments in how managers make sense of shared leadership with related quotes and excerpts (e.g. surprising events, striking interactions) (Naeem et al., 2023). In doing so, we specified managers' embeddedness on various “process levels” (Vandenbussche et al., 2020): their own experiences (e.g. managerial beliefs); the intra-organizational level (e.g. organizational traditions for service delivery); the network level (e.g. changing forms of representation and decision-making) and how these relate to broader child welfare policy developments (e.g. increasing needs of youth populations). These levels functioned as thematic codes to order the lines of interpretation and their interrelatedness in how shared leadership is constructed.

Second, whilst developing a socio-historical understanding of the case, the process analysis was deepened by a focus on “critical moments” that illustrate a particular aspect of how managers make sense of shared leadership. Examples are agreements about agenda-setting and rules of engagement, or fierce discussions about child welfare policy. Moving from specific moments to more abstract interpretations (Naeem et al., 2023), we sought to develop a nuanced understanding of how managers make sense of shared leadership (see Table 3). This required sensitivity as to how differences were processed, how problems were (re)framed and how beliefs and feelings were expressed.

Third, informed by leadership theory in the field of (network) governance (e.g. Ospina and Foldy, 2010), the analysis focused on the forms of engagement with shared leadership that can be distinguished (see Table 3). This unveiled how the constructions distilled in the previous analytical step work out in real-life examples, yielding an understanding of how shared leadership is enacted differently and how the forms of engagement relate to one another (van Hulst et al., 2024; Weick, 1993).

Throughout this analysis, we carried out member checks with respondents by presenting our take on ongoing CWN developments and ambitions for shared leadership at the start of most reflective sessions. Managers’ reflections helped to refine our analysis.

This section describes, first, how managers craft a “reflective time-space” for meaningful dialogue – enabling them to make sense of shared leadership over time. Second, it elaborates on managers’ sensemaking of shared leadership.

Amidst an institutional context that is aimed at individual responsibilities and accountabilities that revolve around the performance of child welfare organizations, initiating a change in how managers perceive their own and shared responsibilities is not straightforward. While referring to themselves as “the four musketeers” (fieldnotes), the four end-responsible executives craft a “reflective time-space” – that is: a socially bounded setting physically separated from their normal working environment and distinct from existing fora that address day-to-day concerns like project management (Bucher and Langley, 2016). In doing so, they facilitate a learning process that enables them to “find a shared language that aligns with the shared responsibility they hold for child welfare problems” (fieldnotes). Here, managers constantly move back and forth between a visualized future of shared leadership and a reality fierce with network tensions to explore what shared leadership means to them. They consider this process meaningful, as it creates spaces for dialogue and opens up a door to reflect on their own managerial attitude:

Despite the policy-induced creation of CWN, the consequences for executives and their organizations are yet to be experienced. Confronting each other's intentions is seen as vital for the problems they face. The reflective space helps them to reorient toward each other and the problems at hand. Although the space is developed exclusively for the executives, they realize that others in their organizations face similar changes (e.g. directors and care workers). They also need time and space to adjust to a reality of working in an organization and a network simultaneously, leading to organized reflection among affected directors in March 2024.

Crafting the reflective time-space allows managers to carve out an unconventional attitude addressing shared responsibilities amidst a fragmented child welfare system. Its origin and anticipatory nature illustrate a desire to transcend governance interventions like contracts with a more felt change, denoting an in-between state marked with no clear ending, as what shared leadership means for themselves, their organization, and the network must find out yet.

This section turns to managers’ sensemaking of shared leadership, and how they start to engage with a new understanding of this. Table 3 summarizes the main findings.

A pattern emerges here in which managers radically “open up” about personal stories and feelings that show their affections with child welfare, and define mutual relationships; from organizational interests to conflicting personalities, from historical events to current dilemmas, and from extended career experience to uncertainties while networking. Here, the reflective time-space is typically used by managers to repeatedly articulate organizational issues, and how these affect CWN. This opening up by managers indicates that shared leadership is portrayed as a mode of interaction as they engage in a process of “relational transparency.” In doing so, they get to know each other beyond formal positions, legitimizing their points of concern. However, such a mode of interaction contrasts with the interactive forms they are used to, based on respect for organizational interests and jurisdictions, which you typically don't share with networking actors:

Here, the underlying issue is that child and family expertise in Dikeland is not yet being used properly, putting pressure on their involvement in CWN. Reaching for radical openness through relational transparency is not self-evident; managers try to understand that what affects themselves may affect others too. The reverse side is that others may interfere in what used to be managers' own, demarcated, organizational concern. This means that shared leadership does not only entail developing a sense of responsibility beyond organizational boundaries (namely CWN), but also about making other networking actors co-responsible for organizational matters. Managers try to give substance to their two areas of responsibility: for their own organization, and for CWN. This attempt aims to contribute to a trustful relationship that can stand networking difficulties (van der Woerd et al., 2024). Such “trust-work” helps managers' shifting identity to encompass their own organization and CWN, allowing for horizontal (shared) and vertical (hierarchical) leadership to co-exist since the home organization is hierarchically organized. Shared leadership hence does not dissolve hierarchies.

Characteristic of managers' in-between state is that they become engaged in identity enactment (Endres and Weibler, 2020). This refers to the ways in which managers perform and reframe their personal identities in light of CWN. Managers use terms for their shifting identity like “ambiguous” and “boundary-spanning” (fieldnotes). They appeared to transcend a perception of “I as part of CWN” by developing a sense of “we as CWN,” for instance when managers questioned previously made decisions:

This excerpt shows that executives want to understand each other’s concerns to feel each other’s personal commitment to CWN. Exchanging what they fear and hope for in CWN, is not what they are used to, but what they feel is necessary. From a participant-observer stance, this was more about showcasing trust rather than demarcating boundaries of interference.

Shared leadership as a mode of interaction is seen as an improvement from past times when managers mainly reasoned from their own organizational perspective. It comes with a search for how CWN becomes part of, and thus refines, managers' identification. They realize that the required change is a “shared task” that goes beyond organizations' individual tasks, and that they themselves are CWN and thus are the change.

In contrast to constructing shared leadership as a mode of interaction, managers also portray shared leadership more structurally as an organizing method. Here, the reflective time-space is typically used by managers to address and identify which work processes and decision-making structures in CWN must change to accommodate an alternative way of service delivery. In doing so, managers start to engage in what we here refer to as “responsive organizing”; a continuous process of reconfiguring CWN by making iterations that impact patterns of behavior (Feldman and Pentland, 2003). This was observed, for instance, when executives expressed concerns about the behavior of directors during CWN meetings. In response, they propose a reordering of consultations:

The excerpt also shows that advocating for structured consultations amongst executives and directors is considered an important step to further professionalize CWN. As such, CWN starts to become a “brand” of significance, calling for adjustments in routinized service delivery.

Underlying this lies a recurrent observation that CWN is considered both as a means and an end. For instance, some directors perceive CWN as a network entity part of a broader portfolio of network involvement – often across the Riverbank region. For them, CWN is a demarcated aspect of their everyday work. For other directors, CWN is perceived as part of a broader policy transformation that centralizes local support in child welfare and with a high threshold to institutionalized care. These perceptions of CWN exist side by side, but over time, clash with one another as they come with different expectations:

Here, the Dikeland director signals that social work in schools is currently out of CWN's scope, while such early support may present a solution to increasing (institutionalized) child care. To meet the purpose of CWN, the organization should be broadened. The excerpt also shows clashing perceptions of CWN, which gives the directors a sense of “feeling lost” (fieldnotes):

The directors experience a lack of decisiveness in decision-making, giving rise to the previously mentioned reflection of the end-responsible managers to reorder the consultations with directors.

Shared leadership as an organizing method emphasizes the structural iterations required for working in a network like CWN, meaning continuous adaptation, fitting the purpose and matching interorganizational dynamics. Shared leadership is used to shape new forms of child welfare services, but also creates confusion because this happens in parallel to existing forms.

Whilst the previous section emphasizes structural notions, the construction of shared leadership as an ideal-typical future is culture-oriented as it foregrounds how participants work together, which is about manners. Here, the reflective time-space is used by managers to question and reconstruct managerial beliefs, and how these guide future courses of action. In doing so, they engage in a process of “moral intentionality” to establish a moral standard that fits with CWN's purpose (Zonneveld et al., 2024), exploring how a new managerial attitude will help them in future scenarios compared to old ones. Through such envisioning, they “emplot” what type of behavior is desired for in CWN (Czarniawska, 2004). Looking ahead (visioning) helps managers to formulate which actions to take in the present.

With such interventions that refer to wished-for future scenarios of timely and accessible child welfare services, executives put into question managerial beliefs that a continuation of current service provision is sufficient. Instead, they foresee that reasoning backward from CWN's long-term dream to what this requires of manners in the here and now is more productive:

The excerpt shows that, for the executives, CWN is about much more than a restructuring of child welfare services. They want to change ways of working, and this requires the personal and moral commitment of all involved.

One of the reflective sessions was dedicated to the development of a so-called “ethical framework” that consists of a template which executives can refer to and use to account for their actions (Feldman and Pentland, 2003). The subject of this session was initiated after executives came to the realization that, in certain cases, they do not behave in line with their ideal of shared leadership. As stated by Dikelands executive: “Shared leadership needs to take place at all different [management] levels, we must jointly start a new culture.” The executives attempt to break with dominant routines within their home organizations and make alternative routines less voluntary. One of the formulated ethical rules entails the lines of representation in CWN, referring to clarity about who is representing who, and on which occasions. For some organizations, it's common that directors mainly interact with directors (and less with executives), while for others, directors are positioned more closely to executives, and even take over their role on “outside” network fora. The executives hence formulated an envisioned routine; professional expertise is leading rather than (formal) position, and as such, executives and directors can mingle. Another ethical rule concerns the executives' scope of interference. While they focus on main issues like budget negotiations with the municipality, directors are given leeway to shape CWN's underlying working procedures for child welfare services. Here, the executives came to the agreement that appropriate behavior entails that executives do not take over the directors' work when things get difficult, but do provide (moral) support.

Shared leadership as a moral standard illustrates how networking actors work on a coherent plot by prescribing how it “should” be. This plot is in-the-making, and therefore requires both envisioning and internalizing new ways of working and thinking.

Leadership scholars emphasize the importance of studying how leadership narratives aimed at shared responsibilities are translated from general notions into concrete meaning (Raelin, 2011; Foldy et al., 2008). This article has aimed to offer an understanding of how managers – in this case, end-responsible executives and portfolio-holding directors – make sense of shared leadership amidst a network-dependent policy transformation in Dutch child welfare (Blanken et al., 2023). This was achieved by conducting an ethnographic study in a Dutch region where child welfare providers seek to collectively provide support to youth populations. Such an engaged approach allowed for the following reflective sessions, where managers put the concept of shared leadership to the test and tried to dissociate from current managerial attitudes that favor organizational performance (Vandenbussche et al., 2020).

These findings contribute to leadership scholarship by showing what sensemaking of a new leadership narrative can bring about (Ospina and Foldy, 2010). We argue that making sense of shared leadership is both a means and an end; it attempts to instigate change and is (part of) the change at the same time. This comes with clashing managerial cultures; an “old” management attitude reasoning from individual responsibilities leads to new problems. This became visible in the considerations made about competition or networking, control or support, and representation of the home organization versus the network. Managers realized that shared leadership can only emerge if performed and internalized by all organizational levels, indicating its systemic and multi-level nature (Kuipers and Murphy, 2023). It requires congruence in doing and thinking and in speech and performance. A leadership narrative must have “real” consequences for actors, transforming mutual relations from shaky, negotiation-driven orders into productive yet dynamic partnerships.

In conclusion, this is the potential of what a new leadership narrative can bring about in networked policies. Bringing shared leadership “alive” is a political act as it brings underlying moral imperatives about “good child care”, “good management”, and “good networking” to the fore (van Hulst, 2012). This fuels an ongoing process of negotiation and contestation among networking actors. Rather than creating and maintaining a new governance order, it is precisely this provisionality that is characteristic of shaping a leadership narrative (Haslam et al., 2024). This extends leadership concepts at the network level (Ospina and Saz-Carranza, 2010; Endres and Weibler, 2020) by pointing to actors' continuous traversing between their own organization and the network, each with their own (and conflicting) traditions. This is illustrative of a leadership narrative in the making, rather than adopting reified leadership notions.

Furthermore, these findings offer insights into the mundane dynamics of the making of a new leadership narrative amidst network-dependent policy transformations (La Grouw et al., 2024). What stands out is that networking actors must navigate through a “web of meanings” with regards to their own and shared responsibilities (Crosby and Bryson, 2010). These are moreover subject to change, and become explicit at critical moments. For networking actors, this means being able to combine both “vertical” and horizontal’ leadership capacities, persevering through both work in a hierarchical organization and an intended horizontal network (van der Scheer, 2023). Whilst scholars emphasize the importance of diplomacy as a leadership capacity (Needham et al., 2025), this study shows that actors develop the ability to understand their own position and the positions and interests of others. Relational transparency is the first step in this process. Opening up allowed managers to reframe their own leadership role in a “shared” leadership context in which mutual influence becomes possible (Raelin, 2011; Foldy et al., 2008). This fosters a sense of responsibility both for the network, and one's own organization and offers a concretization of the less tangible relational aspects that are part of the “transformational work” in policy transformations (Waring et al., 2022). Policymakers could learn from managers' interpretations of shared leadership; they may develop a sensitivity for cultural change beyond formal governance interventions (Clifton et al., 2020).

In particular, these findings illustrate the importance of organized reflection amidst policy transformations, which is not self-evident given the intricacies of public problems that put pressure on managers' time investments. Inspiration can be taken from the efforts of managers and researchers here to foster “environments” to figure out how to move beyond each one's (institutionalized) responsibility (Overman and Schillemans, 2022). The reflective time-space served many purposes: it functioned as a “safe space” to build trust and getting to know each other personally; to address emerging tensions; to challenge network perceptions and performances; to redesign network structures; and to work on a culture and moral story that fits the network, as well as constituent organizations (van Hulst, 2012). The findings showed that managers' reflection helped facilitate “learning” in anticipation of future scenarios in child welfare policy (Czarniawska, 2004). The iterations over time (i.e. shifting identification, changing notion of organizing, internalizing networked ways of working) were forged by a learning environment that stimulated listening, curiosity and sensemaking. The place and time for reflection beyond team-level settings (cf. van Osch et al., 2024), and how this contrasts with other (network) meetings, hence matters in the making of a leadership narrative (Visser, 2024).

The ethnographic approach employed here sheds new light on the making and meaning of a leadership narrative that supports the changing norms and beliefs (Sutherland, 2016). As such, we had to interpret evolving understandings of shared leadership, as these are not static in place and time, especially being aware that leadership is often ill-defined (Denis et al., 2012). This informs process researchers that immersion in the field – and the shifting positionalities this may require – also relates to cogenerating a new leadership narrative (Vandenbussche et al., 2020). The participant–observer fieldwork amidst the reflective time-space was possible after developing good field-level relations. This allowed us to become engaged in, and thus part of, their quest for shared leadership. We were hence instrumental for the field in shaping a new leadership narrative. We were the “mirrors” they looked for in reflection.

These contributions come with several limitations. While our observations focused on managers' constructions of shared leadership in a self-initiated reflective time-space, we paid little attention to other levels of interaction like (policy) system or shop-floor levels. We have tried to mitigate this by considering multiple perspectives from different data sources; for instance, informed by policy documents, we asked managers about their take on (historical) events and interactions with the municipality. This helped to partly grasp the perspective of key actors “outside” the reflective time-space. We encourage future research on (policy) systems and shop-floor levels to unravel how shared leadership comes into being. Furthermore, this analysis focused on a particular health system developed for competition. This may inhibit the transferability of the findings to other contexts. Nonetheless, this study provides insights into the mundane dynamics of how a leadership narrative that transcends organizational boundaries is made concrete through networking actors' interactions during policy transformations – something that is becoming pressing in other policy areas and contexts too.

The data obtained in this article is in line with Dutch research law and regulations and does not have to be tested. All respondents were asked for consent, and this was received each time. More details about data collection can be found in the method section.

The authors would like to thank the research participants for their time and effort. We thank attendees of the Critical Interpretative Public Administration (CIPA) and Public and Political Leadership panels at the Netherlands Institute of Governance Annual Conference (2025) in Ghent for their useful feedback on an earlier draft of this article. We also thank members of the academic community of the Erasmus Centre for Healthcare Management and the Healthcare Governance section for their helpful suggestions in developing this article.