‘Modern talking’: Narratives of agile by German public sector employees

4.1 Methodological approach

Given our research interest in exploring different narratives of agile in German public sector organisations, we followed a qualitative and interpretative research design using narrative inquiry. In this vein, we used a non-probabilistic, purposive sampling approach (Campbell et al., 2020). The necessary condition for inclusion was that the organisations were involved in some way in the implementation of new organisational practices and/or structures that were formulated under the umbrella of agile and its diverse terminology. As argued above, agile as a management ideology has not been widely implemented in PSOs in Germany (Mergel, 2023). In consequence, we first had to identify organisations that had publicly stated ongoing activities, including the implementation of practices, or structures, that were framed under agile (government). The identification of PSOs thus was based on a pragmatic search for statements that showed an engagement in the topic of agile, using search engines and social media (e.g., LinkedIn) in early 2021. Therefore, our sampling includes elements of convenience sampling (Etikan et al., 2016). The organisations’ communications were rich in form, for instance, press releases about (past) participation in fellowship programmes (e.g., ministries), public reports on actions taken (e.g., one ministry, local governments, innovation labs), or even applications for awards (e.g., at the subnational level). This criterion ensured that the processes of narrative construction about agile among organisational members were occurring at the time of data collection, as measures to implement agile had already been taken (Maitlis & Christianson, 2014). In the process of searching for organisations, we aimed for a variety of organisational forms and administrative levels.

Once identified, we contacted those named as responsible with an initial request for an interview. The positive responses led to a sample of organisations ranging from a large state-owned enterprise to several local innovation labs and core bureaucratic organisations, namely four ministries and four federal agencies, one state authority, and two local authorities in major German cities. Consequently, as stated above, we do not claim a probabilistic sample with claims of generalisability, as we may have missed communications and positive responses were unevenly distributed across the levels. Nevertheless, our research aims to provide a comprehensive picture of narratives of agile government. We argue that the organisations under study constitute an interesting set-up that allows to better understand how agile is perceived and made sense about differently in a federal country like Germany. An overview of the cases is provided in Table 1 and in-depth case reports can be found in Appendix C.

Following our enquiry, we asked contacts to identify stakeholders who were involved in both the planning and adoption of the measures aimed at introducing what is framed as agile. Here, we followed a snowball sampling approach (Biernacki & Waldorf, 1981). This allowed us to gather a diverse picture of agile, as both promoters and resistors were interviewed to minimise a potential pro-innovation bias. In addition to the public managers and employees, we also interviewed consultants ( n = 3) and a reseacher ( n = 1) that were referred upon by the organisations as either involved in their activities or known as experts in the introduction of agile government. We consequently relied on positive responses, resulting in a different number of interviews per case, including a heterogeneous number of functions. All interviewees were promised full anonymity. A more detailed list on information of the interviewees is provided in Appendix A.

Data collection methods included semi-structured interviews, a semi-standardised online survey and document analysis. In total, we conducted 31 interviews in all included cases except one of the local authorities, where semi-standardised online surveys ( n = 24) were the main mode of data collection. The interviews took place between September 2021 and April 2022. They were conducted in different team constellations and settings, including telephone, video or in-person. All interviewees were involved in the implementation of agile elements in the organisations as consultants, managers or employees. The questionnaire included questions about the actors’ understanding of agile, its manifestations within the organisations as well as the difficulties and dynamics of implementation (see Appendix B for the interviews’ and survey’s questionnaire). All in all, we followed an explorative approach based on types of questions that allowed respondents to (re)construct their perceptions and understanding of the shift toward a more agile organisation in their responses (Nielsen et al., 2023). This allows for a retrospective analysis of possible elements of narratives that actors either disseminate themselves or report as being disseminated within the organisation. All interviews were transcribed verbatim.

To understand composite narratives, it is first necessary to understand the dynamics of the organisations and their respective environments. Therefore, we wrote a case description for all cases (see Appendix C). Secondly, following the data analysis strategies of Dalpiaz and Di Stefano (2018), and Zilber (2007), we then combined thematic content analysis and narrative analysis. This involved first sorting the data and working out the individual stories within the cases before analysing the resulting narratives in the second step. The individual steps are described in more detail below.

First, inductive thematic analysis was used to reduce the available interview and document data to those text fragments that contain statements about agile. This first step of coding was carried out separately by two of the authors. Upon completion, all authors met and discussed differences in coding. After resolving conflicting decisions, the second step of narrative analysis was carried out jointly by two of the authors. That step was divided into three stages of analysis. First, for each interviewee and each document, the story fragments that appear in the respective text segments were examined inductively. These narrative themes are, for example, “agile as external phenomenon”, “agile is not rocket science”, “agile needs change management (capacity)” (see Appendix D for a complete list of all codes). During this coding process, the authors constantly discussed which individual narratives could be identified for each case. The central narrative elements were then extracted from each narrative. In line with Dalpiaz and Di Stefano (2018), we draw upon Bruner’s (1991) approach of interpretative abstraction: Each story (individual level) and narrative (composites of stories) is structured in terms of the main actors, their quest (which overarching goal is to be achieved), the acts to achieve this goal, the instruments by which this is done and the context in which it is done (see Appendix E for an overview of the individual stories). Next, these individual narratives were combined into composite narratives as described in the coding scheme (see Fig. 1 and Appendix F for a short description). Finally, from these composite narratives, three meta-narratives were derived and discussed among all authors. The following section of findings is organised along the meta-narratives.

OPEN IN VIEWER

The individual stories of our first meta-narrative centre around what the interviewees describe as key disadvantageous features of the bureaucratic context, like hierarchies, lacking flexibility in planning, and change. They are described as obstructive for ministries to manage the “rapid changes” of the environment (I1).

Across the interviewees this can be identified as the ultimate quest for agile in a governmental context: “We recognised that […] in the VUCA2

Agile is therefore understood not only as an improvement of bureaucracy and public administration, but also as a complete renewal of the public sector and its work environment: “Agile in the context of how we did it, it’s about big changes that do a lot to everyone affected” (I26). In this setting consultants, researchers, and other external actors serve as teachers for public servants that identify willing partners and teams within the organisation to collaboratively disseminate and apply agile tools. These teams then legitimate agile methods to convince the resistant and critical bureaucrats of the methods’ necessity. Ultimately, upscaling within the organisations must be driven by top-management via change management as it is beyond the reach of the individual driving actors.

Interestingly, narratives within this context explicitly build on the contrast between (agile) units and the non-agile subsidiaries, especially within the state-owned enterprise. The storytellers presented themselves as ‘agile gurus’ living on an “agile island” (I25) of the ‘blessed’ trying to proclaim their message of agile as workplace democratisation and self-actualisation to those non-agile areas of the organisation. To achieve this quest, they act as a grassroot movement trying to find company and community through networking. In their role as strong advocates of agile, they want to give others the chance to share in their revelation of it. However, as conviction-users, their confidence sometimes turns to incomprehension when they have to cooperate with the non-agile parts of the organisation:

The new world of agile government consequently can be characterised as a democratic and digital workplace with less hierarchy, which is becoming more user-centric, flexible, and efficient through the use of agile frameworks and methods.

Narratives that can be summarised under this meta-narrative are mainly framed in the classical sense of the ‘Agile Manifesto’. The respondents understand agile as a well-suited method for developing digital products and services for their administrations in an “experimental and iterative manner” (I15). Agile methods should be employed by competent actors in an innovative environment to drive the administration’s digitalisation. This forms the main quest, especially for the interviewees of the innovation labs.

Furthermore, top management stresses a picture of agile as a multifaceted, intertwined change project to position the organisation in a complex environment, anticipating budget restrictions and resource scarcity:

Agile here refers to combined measures of project management methods or tools and structural and cultural changes to achieve their postulated quest. Thus, top management addresses a need for change that is driven by public managers, teams, and top management itself. However, the environment in which this change takes place is seen as complementary to the traditional Weberian public administration. As agile takes shape through pilot projects such as “a project management unit” (I10) or additional structures such as separately managed innovation labs where “digital projects can be launched on our own initiative” (I15), the core public administration would only need to be partially transformed to compensate for the public sector’s deficits in service development or project management.

Consequently, the interviewees in the innovation labs draw a sharp contrast between themselves and the non-digitalised and sometimes dysfunctional municipalities that suffer from a lack of resources. From an outsider’s perspective, bureaucracies are portrayed as systems that differ sharply in terms of culture and competences, leading to frustrations in achieving the quest of enabling innovative products and solutions. In contrast, they identify themselves as startup like entities with diverse agile mindsets and a working environment that allows for the implementation of diverse project management methods:

However, the interviewees describe that the main acts and instruments must be applied in collaboration with the administration in order to achieve the goal of digitising the administration through agile methods. Here, the innovation labs fulfil the function of a hybrid product provider. Whereas the product and projects are run outside the organisation to enable agile method adoption, the planning and implementation is an internal administrative task supported by the innovation labs. Nevertheless, the interviewees were rather negative about the ability of administrations to work in an agile way in terms of creative methods of product development, as bureaucracies tend to “detain (actors) in long projects” (I18). While agile thus helps to introduce better digital applications and processes, the method itself does not function within overly rigid bureaucratic structures as it needs “freedom to move” (I13) and can therefore only be seen as complementary to existing structures.

5.3 Integrate: Bureaucratic agile (meta-narrative 3)

The third and final meta-narrative presents agile as a novel and desirable way of working in bureaucratic organisations. In contrast to the other narratives, agile thus does not represent a turmoil for bureaucracy, nor is it seen as non-compatible and hence complementary, but as an integral part of modern bureaucratic organisations. Agile is predominantly understood as a management ideology that can improve existing daily routines and practices. The public employees’ understanding of agile predominantly relates to elements of daily communication and (digital) collaboration within and across ministerial teams. Specifically, creative and structuring methods are mentioned here that can change the meeting culture of teams in the direction of improved informative and interactive exchange with more regular feedback rounds. Moreover, vertical and especially horizontal ministerial collaboration should be enabled by a digital workplace, including SharePoint and Kanban boards to avoid lengthy voting rounds and improve “boring [and] dull” (I12) communication across teams.

The quest of the narrative is to integrate this form of agile adapted to the bureaucratic context that is described as lacking digitalisation, with slow and complex processes, and still coined by “superiors who act in the old style, the lord of the manor way” (I12). Agile thus aims at keeping up with modern times, breaking down silos and consequently implementing a modern, digital working style suitable for public authorities: “What I really associate with agile and agile working is project-based collaboration across departments” (I29). One interviewee postulates her motivation for introducing agile is based on the current working mode in her organisation that she describes as “absolute madness“ (I4). Agile is subsequently narrated as an umbrella concept from which useful measures can be drawn as long as they fit the bureaucratic context, as one interviewee points out: “To apply agile methods to projects without having to go straight to Scrum and such. It doesn’t have to go that far” (I12). Hence, agile is seen in essence as an opportunity to promote digitisation and digital tools in a modern (and partly virtual) working environment.

The driving actors in the narrative are mostly individuals, i.e., motivated employees who fight for a modern workplace with their either resistant or supportive colleagues and managers, with the final aim to create a “critical mass” of organisational employees that follow their vision of a modern workplace. These convicted individuals themselves have formed bubbles and networks in which they “stick together in a matter that is not easy” (I13), providing shelter from negative experiences within their teams. This narrative element is similar to what we identified in the first narrative, ‘renew’, namely the focus on the individual as the driving actor. In contrast, interestingly, the role of colleagues as accepting or rejecting actors is more present in the narratives here, as it is about integrating agile into daily routines. The acts and underlying instruments to achieve the quest described are bound to the communication areas and existing collaborations of these individual employees. Particularly, they describe the translation of the agile terminology and methodology within this bureaucratic context:

In addition to adapting language and methods by continuous experimentation, respondents cite building digital capacity and skills as a keyway to achieve their quest of bureaucratic agile. Employing digital tools is framed as a central foundation. In the light of creating a modern workplace, agile and digitalisation are inherently connected within this narrative. All in all, the employees suggest that they are currently in a phase of “trial and error” (I30), as they try to make sense of what works and what does not.

Within this overarching narrative, a more critical perspective also emerges, mainly complaining about agile as a fad, while longing for similar goals as the other storytellers. The narrators here see a dysfunctional, bureaucratic organisation that lacks basic digital tools and an appropriate feedback and error culture, as stated in one interview: “I think a better management culture would be necessary. Then there would be no need for new desks or a [fancy] phone booth in an open-plan office” (I28). Hence, they want more pragmatic solutions and a functional work place without any need for fuss. For them agile is not more as remote working in a new workplace that offers sufficient collaboration tools and better methods of communications. Consequently, their actions consist of “nipping change in the bud” (I28) to expose agile as a buzzword used by top management as a fig leaf to cover up the shortcomings of what they see as the current way of working in bureaucratic organisations.

6.1 Modern talking and narratives of agile

In our study, we show that the idea of agile as a management ideology (Mergel, 2023) is understood differently by employees, managers, and externals (such as consultants and researchers) working in, with and for German public sector organisations. With this interpretive work based on a narratological perspective, we thus contribute to the literature on agile government, in particular by adding a critical perspective on how this ambiguous management concept is perceived in practice. Assuming the performativity of language (Loewenstein et al., 2012; MacKenzie et al., 2007), how a concept is talked about locally, affects its subsequent implementation through the dynamic interplay of actors’ acceptance, adoption, and rejection of an idea within its specific context (Waldorff & Madsen, 2022). In the following, we discuss our analytical findings alongside three perspectives on the examined narratives of agile government: the narratives’ attitudes towards agile’s degree of fit with bureaucracy, agile’s connection to digital transformation as well as the implied actors’ constellations.

First, all three narratives convey, to varying degrees, how agile government fits with the narrators’ perceived organisational realities of bureaucratic organisations. For some (‘renew’), agile represents the requisite disruption that challenges and changes bureaucratic functioning and thinking wherever possible and necessary to survive in the VUCA world. For others (‘complement’), it is seen as incompatible with core bureaucracy in the Weberian sense. Hence, in this narrative contrasting the first, bureaucracy is not subject to change. Rather agile, is presented as a very useful and necessary management framework for achieving what bureaucratic routines and practices are designed to constrain: experimental thinking and flexible action. Thus, agile takes on functions in complementary settings, e.g., in laboratories or projects. In the third narrative (‘integrate’), and in stark contrast, agile is presented as a means of achieving a better working culture, including more ‘modern’ collaboration and communication practices. Here, agile government is about changing burdensome internal practices and routines to facilitate and improve working structures, such as meetings or interdepartmental collaboration.

Consequently, we find that agile government, as an arguably ambiguous management fad, provides a rich projection surface for what the narrators hope to improve within their individual realities of bureaucracy. They span all levels of narrative abstraction, from the VUCA world to the way meetings are facilitated. Our analysis highlights that the narratives of agile government function particularly in structural relation to the perceived (mal-)functions of bureaucracy. The solutions (quests) provided by agile as proposed in the narratives are consequently directed at different levels and locations, depending on the extent of the perceived challenges. They range from an abstract focus on changing entire organisations to adapt to the ‘new realities of crisis’ (Janssen & van der Voort, 2020), to adding new temporal structures to be able to work in a more experimental way (Tõnurist et al., 2017), to changing practices and routines to modernise the way of working (Mergel, 2023; Mergel et al., 2021).

This finding that, although to varying degrees, agile in the narratives is always thought of in relation to the core ideas of bureaucratic organisation and also builds on these, is consistent with research on earlier management trends such as NPM (Polzer et al., 2016). In this respect, Heidlund and Sundberg (2023) show that digitalisation strategies also predominantly refer to increasing efficiency as a central objective. To be recognised as an idea and reform, new trends must incorporate and build on certain taken-for-granted ideas to take hold in administrations (Polzer et al., 2016). Moreover, our analysis shows how the practitioners’ narratives resemble the scattered picture of what agile is in government in particular, and how it manifests as depicted in the literature (McBride et al., 2022; Mergel et al., 2018; Soe & Drechsler, 2018). In doing so, we contribute to the literature on agile government with a more structured picture of the diversity of how agile is narrated in practice within the context of bureaucracy.

Surprisingly, citizen-centricity was very rarely mentioned in the narratives, despite being central to agile as an ideology (Mergel, 2023). Improving citizen-centricity appears to be less of a priority for our interviewees. Instead, they present agile methods primarily as a way to change bureaucratic structures and practices, especially in meta-narratives 1 and 3. Consequently, without a focus on agile methods for software production, the user perspective (with citizens as service consumers) plays only a minor role in these narratives. We can only assume that this might be the result of our case selection, as few organisations provide services to citizens at street level. If so, the employees are seen as users for improved agile practices. In meta-narrative 2, user-centricity is indeed presented as a standard element of agile product development that can be pursued outside of traditional administrative structures. Interestingly, however, citizens as users are hardly mentioned.

Building on the literature on sensemaking and sensegiving (Maitlis & Christianson, 2014; Vaara et al., 2016), this equivocality (i.e., many voices on the same issue) may pose a key challenge for implementing agile in practice. In the absence of a compelling narrative, agile remains an individually constructed and understood concept, aimed at different problems with different solutions. Or one that is rejected altogether. This is in line with Mergel (2023, p. 13) who explains the consequences of the absence of such change-supporting top-down narratives in the implementation of agile practices as such:

Second, agile as management fashion is narratively connected to further diffusing fads in the public sector. Particularly one megatrend emerges that we find to be interwoven and at points almost synonymously employed with agile: digitalisation including different forms of remote work. In more detail, we identify three distinct connections of agile and digital change. The first mechanism characterises a digital organisation (understood as remote working, the existence of digital tools and applications, and sufficient IT infrastructure) as the necessary foundation to collaborate and communicate in an agile way. Vice versa, the introduction of agile can secondly function as an enabler of digitalisation if agile working teams increase the demand for the improvement of digital processes, or the implementation of digital collaboration tools. Lastly and thirdly, as mentioned above, the organisational discussions on agile as a broader change-endeavour can serve as a bandwagon for digitalisation efforts. If measures to structurally introduce agile are discussed, acts for digitalisation are automatically included or even seen as less disruptive for some actors, improving agile’s chances to be implemented.

This finding that agile and digital are inherently linked does not seem surprising given that agile, in its most famous approach, is a software development method (McBride et al., 2022). Also, in the early literature on its application in the public sector, it is framed as addressing “the need to innovate digital service delivery in government” (Mergel, 2016, p. 1). However, on closer examination, the analysis shows that agile is not only talked about in this way, reflecting its recent call for implementation in other projects and government routines (Mergel, 2023). Here, therefore, agile can both serve and benefit from digitalisation in government, particularly in relation to a digital organisation, including digitalised back office and project work. Agile working teams utilise and demand digital tools and processes that others in the organisation can then benefit from. However, this requires both technological resources and competencies that need to be harnessed and built (Franco et al., 2022; Tallon & Pinsonneault, 2011). Agile as a management framework, though not a technological reform per se, is thus a central discursive element of the digital transformation discussion, and vice versa.

Third, our analysis shows that the narratives predominantly thrive on a distinction between those who act for the implementation of agile (us) and those who are resistant or neutral to it (them). This distinction seems central to the storytellers. The belief to act for or against an ideology determines which actors are presented as allies and which are still to be convinced. This applies regardless of the background of the interviewees. However, it seems that the more distant the storytellers are from those who are not acting as agile as desired, the more abstract the opposite party (‘the rigid bureaucrats’) becomes. Although the storytellers do not confront public sector employees personally and also positively highlight motivated internal allies, the narratives reflect what we know from the stereotype literature to be prominent features of the portrayal of bureaucrats as inflexible and overly rule-oriented (Dinhof et al., 2023; Willems, 2020). As with what we discussed earlier, particularly in the ‘renew’ and ‘complement’ narratives, at times the bureaucratic mindset is characterised as a hindrance to the agile quest. This resonates with the tones of either the need to renew the bureaucracy or to complement it, as within the bureaucratic organisation staffed with bureaucratic minds, agile cannot unfold its value.

Solely within the ‘integrating’ narrative, which is predominantly shared by the core administrative staff, we examine a rather differentiated picture. The dominant actors consist of a trio: the driving employees or managers, who believe in the value of agile, the either resistant or neutral colleagues, and the respective (team) leadership. The latter is presented as the ‘tip of the scales’ in achieving the goal of being able to sustainably integrate agile into the daily work routine. However, this (rhetorical) guidance seems to be lacking according to the interviewees. Without leadership support, the storytellers argue that they are unable to act due to the lack of a sensegiving process initiated by more powerful actors (Balogun et al., 2015). From the perspective of the literature on narratives, three main tasks emerge for public managers as sensegivers (Maitlis & Lawrence, 2007) if they or their organisational leadership pursues the adoption of agile. First, they function as a sieve for narratives external to the organisation, following the managers’ own interpretation of the phenomenon (Waldorff & Madsen, 2022). As the second function, they are expected to weave a compelling narrative that integrates the local organisational traditions and bureaucratic routines with what is viewed as legitimate from the outside (Day et al., 2023; Logemann et al., 2019; Maitlis & Lawrence, 2007). Lastly, this narrative must be adapted over time to fit the internal discussions (Nielsen et al., 2023). Respectively, the management must be capable of orchestrating the narrative development process, which includes pushing and probably resisting voices over time while ensuring the compliance with the overall strategic goals of the change initiative (Vaara & Rantakari, 2023).

Our study contributes to the discussion of agile government by introducing a narrative perspective. We show three prevalent meta-narratives of agile, which are positioned differently to bureaucracy, and which present different visions of agile public sector organisations. In summary, we show that processes of sensemaking about what agile might be for different actors, are in full swing. However, if not responded to or orchestrated by management, these narrative dynamics could catalyse a fragmented picture of how agile manifests in organisations (Mergel, 2023), or encourage processes of symbolic decoupling (Waldorff & Madsen, 2022).

6.2 Limitations and future research

Next, we discuss the limitations of our study as well as directions for future research. Firstly, interpretive narrative analysis puts the researcher in the position of analysing the social realities of the actors under study. Although this is not a limitation per se, the extraction of narratives and their analysis is subject to individual analytical biases, and one may interpret the data differently. In addition, as interviewers, we implicitly convey our personal beliefs about the topic in our questions and responses, which can potentially result in response bias. Nevertheless, our description of the processes of data collection and analysis, and the provision of the questionnaires, as well as the coding scheme aim to increase the reliability of the findings. Future research could therefore critically assess our findings and derive comparative accounts of narratives of agile across different countries (e.g., where agile is part of public policy), areas of administrative action, and over time (Nielsen et al., 2023). Furthermore, as narratives are complex entities that often contain implicit contextual cues that may be overlooked, the narratives derived here may be oversimplified. Using novel methods such as structural topic modelling (Guenduez & Mettler, 2023), future research could extend the findings of this paper by exploring narratives about agile based on larger n-studies.

We do not claim generalisability with the selection of interviewees and their organisations. The main interest of this paper was to provide an in-depth analysis of existing narratives about agile in organisations that have begun to engage differently with this management ideology. In doing so, we face the risk of selection bias, and management scholars should be cautious about extrapolating our findings to other contexts. In organisations where agile is not being applied in any form, the narratives may unfold very differently. Nevertheless, our respondents come from a wide range of public sector organisations, and we provide a first empirical picture of the narrative manifestations of agile. Finally, it is beyond the scope of this paper to draw conclusions about the potential advantages or disadvantages of agile for public sector organisations. Following previous research in this area (Mergel et al., 2021), scholars could explore how organisational performance or public value delivery changes following the introduction of agile measures (McBride et al., 2022), also incorporating the perspective of citizens. However, researchers should be cautious about the individual nature of what agile is or is presented as in the examined case.