Adopting agile in government: a comparative case study

ABSTRACT

This study examines the adoption of agile in public administrations through the lens of Scandinavian institutionalism and translation theory. By conducting interviews in 19 German public administrations, we investigate how agile is translated into public settings and how they address associated challenges. Our findings reveal three translation modes: agile stripped down to a cultural concept, agile as governance to foster cross-functional collaboration, and agile as methodology, adhering closely to its original values and principles. As such, our study contributes to a better understanding of and theoretical advancements surrounding agile in government.

KEYWORDS:

• Agile government
• translation theory
• public sector innovation
• reform paradigm

Introduction

In recent years, there has been a mounting interest in using agile values, principles, and methods in public administrations in many countries (Baxter et al. 2023; Mergel 2023; Mergel, Gong, and Bertot 2018). This interest is largely driven by the push by many governments to become more digital, but also due to higher citizen demands (Steen and Schott 2019) and new forms of complex problems (e.g. the climate crisis or Covid-19) that require new ways of organizing an administration’s resources to be able to respond effectively and efficiently. The central tenets of agile are captured in the definition proposed by Mergel (2023, 1), who sees the concept as ‘a work management ideology with a set of productivity frameworks that support continuous and iterative progress on work tasks by reviewing one’s hypotheses, working in a human-centric way, and encouraging evidence-based learning’. Despite the roots of agile being in software development and IT (Mergel, Ganapati, and Whitford 2020), administrations not only use agile in these areas, but also in non-IT settings such as project management, acquisition, and evaluation (Mergel 2023; Mergel, Gong, and Bertot 2018). However, notwithstanding the high interest from practitioners in the concept, research on agile in public organizations remains scarce (Mergel, Ganapati, and Whitford 2020; Mergel 2016; Mergel, Gong, and Bertot 2018; Nuottila, Aaltonen, and Kujala 2016; Simonofski et al. 2018). According to Mergel and colleagues (2020, 161), ‘the canonical public administration literature has largely neglected agile and the more fundamental changes it introduces to hierarchical and bureaucratic organizations’.

Perhaps most pressingly, there is little empirical research on how public administrations can adopt agile. If we understand agile as a form of governance innovation that could potentially provide better ways and processes to address societal problems (De Vries, Bekkers, and Tummers 2016), we may argue that it is of utmost importance to better understand through which mechanisms agile can be translated into government contexts. This need for research on adopting agile is further underscored by the assessment that ‘agile is antithetical to many typical bureaucratic line organizations’ (Mergel, Ganapati, and Whitford 2020, 164), requiring substantial changes to organizational structures, decision-making, culture, operating procedures, and leadership. Another gap in the literature is the lack of theoretical grounding of many studies on agile, caused by a strong practice orientation (Baxter et al. 2023). We address these research gaps by asking the two following research questions: How do public administrations translate agile to their context? And how do they address the associated challenges? To answer these questions, we draw on Scandinavian institutionalism and translation theory (Røvik 2016; Wæraas and Agger Nielsen 2016) and conduct a qualitative comparative case study of 19 administrations at all three German government levels. We chose Germany as the setting of our study as German government is one of the most rule-oriented bureaucracies in the world (Kuhlmann et al. 2021, 3), making it an extreme case in which introducing agile should be particularly difficult.

The five chief contributions of this study are first, an expansion of the body of empirical literature on agile in public settings; second, an in-depth comparative case study of the mechanisms when adopting agile in public administrations, improving our understanding of the underlying challenges and possible solutions; third, an extension of the theoretical debate surrounding agile by introducing Scandinavian institutionalism and translation theory as well as by formulating theoretical propositions derived from our results; fourth, a refined and empirically grounded definition of agile in government; and fifth, a discussion of the potential of agile to become a new public sector reform paradigm.

Literature review

Agile in the private sector: values, principles, and methods

In private sector literature, agile has attracted the attention of scholars and practitioners for over two decades. The cornerstone of agile are the four agile values and twelve agile principles outlined in Beck et al.`s Agile Manifesto (2001) (Appendix A). Remarkably, these principles continue to provide robust guidance for software development teams and their projects (Williams 2012). While agile values (e.g. customer collaboration over contract negotiation) are the core beliefs and mindset of the agile approach, agile principles (e.g. welcoming changing requirements, even late in development) provide more specific guidelines for applying agile values in practice. Building on these values and principles, several agile methods (e.g. Scrum, Kanban, or Extreme Programming) exist. They are specifically designed to manage complex projects, originally in software development but increasingly also in project management contexts (Mergel, Gong, and Bertot 2018). Several actors play a role in each method. In Scrum, for instance, the Scrum master has the task to remove impediments but also to ensure adherence to the agile principles. The product owner represents the customer’s interests, and the development team delivers the product itself.

Scholars study agile in different settings, such as software development (Chow and Cao 2008; Livermore 2008; Maruping, Venkatesh, and Agarwal 2009; Misra, Kumar, and Kumar 2009; Tang and Kishore 2010; Tripp and Armstrong 2018) and agile project management (Gemino, Horner Reich, and Serrador 2021; Malik, Sarwar, and Orr 2021; Serrador and Pinto 2015). Some operationalize agile through agile values like customer collaboration (Misra, Kumar, and Kumar 2009; Tang and Kishore 2010), while others focus on agile methods such as Scrum, Extreme Programming (Livermore 2008), or single aspects of such methods (daily standups or retrospectives) (Tripp and Armstrong 2018). Furthermore, the combination of traditional project management practices with agile methods (e.g. Scrum) is an emerging field of research (Gemino, Horner Reich, and Serrador 2021).

Agile in the public sector

This previous research indicates agile has been studied in various application areas, suggesting their potential usefulness for public administrations. The flexibility in how agile is adopted in diverse contexts presents an opportunity for administrations to tailor agile to specifically tackle their distinct requirements and challenges (Mergel, Ganapati, and Whitford 2020). These unique requirements might include the need for stringent regulatory compliance, adaptions to complex bureaucratic structures, and the demand for transparency in decision-making. Moreover, agile could help to overcome some public administration challenges such as digital transformation (Mergel, Edelmann, and Haug 2019), administrative burden (Moynihan, Herd, and Harvey 2015), or the push to enhance user-centric approaches (Strokosch and Osborne 2020). Notably, one central facet of agile involves fostering collaboration with users, a principle that resonates with the public sector’s commitment to user-oriented service delivery (Osborne, Radnor, and Nasi 2013; Stewart and Clarke 1987). Furthermore, the studies focusing on specific elements of agile methods, such as daily standup meetings or retrospectives, suggest that adopting select components of agile methods can bring benefits such as increased efficiency (Tripp and Armstrong 2018), indicating that it might be possible to achieve similar outcomes in public settings.

However, there is wide consensus that agile presents major challenges for the hierarchically organized and bureaucratic public organizations (Baxter et al. 2023; Lappi and Aaltonen 2017; Mergel 2016; 2023; Nuottila, Aaltonen, and Kujala 2016; Simonofski et al. 2018). First, transitioning to agile introduces new challenges for leaders. They must empower employees to make independent decisions, yet many leaders find it difficult to understand agile or see it as a threat to their role behaviour, and consequently hesitate to adopt it (Mergel 2016, 2023; Simonofski et al. 2018). Additionally, since agile encompasses a wide range of values, practices, and methods, and there is no universal guideline on how to become or adopt agile, it is often misconstrued (Baxter et al. 2023; Simonofski et al. 2018) or employees perceive agile as an additional workload that they cannot stem in parallel to existing administrative processes (Mergel 2023). Furthermore, a lack of training and change management can undermine the successful adoption of agile (Baxter et al. 2023; Simonofski et al. 2018). To address these challenges, it is important the management communicates a clear agile vision, supported through visible and communicable expressions (Baxter et al. 2023; Mergel 2023). Additionally, fostering collaborative meeting practices, such as soliciting input from all attendees, and providing examples of problems best suited for agile problem-solving, are essential for navigating the transition effectively (Mergel 2023).

Bureaucratic rules, like procurement laws, hinder the integration of agile into administrations since they conflict with the nature of agile development. When project outputs must be predetermined it becomes challenging to incorporate changing user demands or new regulations during projects, thus impeding iterative development (Simonofski et al. 2018). To counterbalance this problem Baxter et al. (2023), who investigated the adoption of the Scaled Agile Framework (SAFe) within the UK Ministry of Defense, found that rather than allowing a single large project for only one supplier with pre-defined requirements, they manage a series of short duration contracts.

Moreover it is often difficult to recruit IT talents with specific profiles that have a deep understanding of agile methods (Baxter et al. 2023; Mergel 2016; Simonofski et al. 2018). To overcome this barrier, Mergel (2016) reports how the U.S. federal government developed an innovative HR-policy that paved the way for IT talents to spend short terms in the federal government to create solutions for pressing problems.

Lastly, while public services should remain accessible to all parts of society, implementing user collaboration, a core value of agile, can be challenging in practice due to the difficulty in selecting a representative segment of society to participate (Simonofski et al. 2018).

Theoretical framework

Organizations and their employees operate within dynamic environments and relationships. While some organizations eagerly adopt new ideas, paving the way for new reform paradigms (Polzer et al. 2016) or institutional logics (Meyer and Hammerschmid 2006), others lag behind, adopting innovations at later stages (Rogers 2010). Numerous theories attempt to understand the motivations behind organizational adoption of new ideas. According to American institutionalists like Powell and DiMaggio (2012), organizations primarily adopt ideas to bolster their legitimacy. Others follow management fashions (Cram and Newell 2016) or change due to normative drivers such as increasing social equity or democratization demands (Wise 2002). How organizations subsequently translate and integrate new ideas in their existing organizational routines is this article`s central question. Prior research on public sector reforms indicates that new ideas are often ‘layered’ on top of the original bureaucratic logic, resulting in the co-existence of multiple institutional logics or organizational forms within one organization (Polzer et al. 2016). Another theoretical lens to analyse the adoption of new ideas, and the central theory of this article, is Scandinavian institutionalism and the theory of translation (Wæraas and Agger Nielsen 2016).

Scandinavian institutionalism and translation theory

Emerging in the 1980s, Scandinavian institutionalism examines how organizations incorporate new ideas into their structures, by particularly focusing on management techniques’ adoption within the public sector (Wedlin and Sahlin-Andersson 2017). In contrast to traditional diffusion models (Powell and DiMaggio 2012), which perceive ideas as static entities often mindlessly adopted by individuals (Øygarden and Mikkelsen 2020), Scandinavian institutionalism underscores the significance of social dynamics and local values to which ideas are confronted as they circulate (Røvik 2016; Wæraas and Agger Nielsen 2016). This perspective highlights that the same idea can produce different versions once it enters local contexts (Czarniawska and Sevón 1996; Wedlin and Sahlin-Andersson 2017).

Agile has been translated into public settings by numerous administrations. As such, we see translation as a sub-form of adoption. Translation theory, introduced by Czarniawska and Sevón (1996), is an essential component of institutional analysis (Nielsen, Wæraas, and Dahl 2020; Wæraas and Agger Nielsen 2016). Translation refers to the adaptation of new ideas to local contexts (Wæraas and Agger Nielsen 2016), involving actors who actively carry, reinterpret, and reconfigure new organizational elements into their specific organizational and cultural context, through activities such as workshops, meetings, and experimentation (Boxenbaum and Strandgaard Pedersen 2009; Nielsen, Wæraas, and Dahl 2020). Consequently, when organizations encounter new ideas or practices, they must ensure alignment with the existing organizational culture, norms, and structures. This task creates various challenges (Røvik 2016) as evidenced by prior studies on agile adoption (Baxter et al. 2023; Mergel 2016, 2023; Nuottila, Aaltonen, and Kujala 2016; Simonofski et al. 2018).

Especially New Public Management (NPM) has fostered many studies on the translation of private sector management techniques in public settings (Nielsen, Wæraas, and Dahl 2020). Several scholars (Radnor and Johnston 2013; Radnor and Walley 2008; Radnor and Osborne 2013) studied the adaption of lean (also referred to as the Toyota Production System or TPS) to the public sector and draw the conclusion that, instead of the promise to improve the service user’s centrality and establish itself as a new reform paradigm, lean mainly improves internal processes of public organizations and thus fails to increase value creation or customer service. One explanation provided by Radnor and Osborne (2013) is that the adopting organizations focus excessively on the technical tools of implementation without an understanding of the principles and assumptions of lean or the context in which it is implemented. This raises the question whether agile would be able to finally foster a truly user-centric approach, also given that lean, much like agile, is an aggregation of diverse tools and techniques.

Moreover, Nielsen and colleagues (2020) demonstrated how a leadership concept (‘The Leadership Pipeline’), originally conceived for private companies, was translated into Danish municipalities. They find that, contrary to the general expectation in translation theory, ideas do not always need to be entirely transformed to fit the local context: ‘When the fit between the management concept and the local context is good, and managers have sufficient hierarchical authority to control the translation process (…) reproduction could be a viable strategy. Translation under such circumstances, then, becomes more a matter of transforming the organization rather than the adopted concept’ (Nielsen, Wæraas, and Dahl 2020, 250). Moreover, the authors state that collaborative efforts, thus participation in pre-adoption activities and networks, is important to learn about other translation initiatives in the same field.

Many scholars deduce their own translation principles based on the specific contexts they investigate (Andersen and Røvik 2015; Gond and Boxenbaum 2013). Most notably, Røvik (2016) proposes a theoretically developed typology of three translation modes: reproducing, modifying, and radical (Figure 1) which we employ to examine the translation of agile into public settings. These three translation modes are further delineated by four translation rules: copying, addition, omission, and alteration (Figure 1). Røvik (2016) distinguishes between the source context, from which an idea is de-contextualized, and the target recipient where the idea is then re-contextualized. Contextualization ranges from replication or minimal changes to radical transformations resulting in local innovations (Røvik 2016).

Figure 1. Translation theory based on Røvik (2016).

Data and methods

We adopt a positivist and interpretative approach to address the research questions of how public administrations translate agile and how they address the associated challenges (Cassell, Cunliffe, and Grandy 2017; Lin 1998). According to Cassell and colleagues (2017, 18–19) ‘positivist qualitative research focuses on searching for, through non-statistical means, regularities, and causal relationships between different elements of the reality, and summarizing identified patterns into generalized findings’. However, we are not only interested in finding strategic patterns in agile adoption and suggest propositions that can be tested in other cases (Lin 1998), but we also aim to understand the mechanism behind agile adoption, thus ‘going beyond “what” has occurred to see “how” it has happened’ (Lin 1998, 167). Translation theory is particularly suitable for this endeavour since the theory focuses on how ideas are constructed and transformed within different social contexts. Similar to translation theory, interpretivism acknowledges the role of subjectivity and human agency in social phenomena. Translation theory thus serves as a guide for interpreting our data.

Semi-structured interviews with civil servants

We used a multiple case study research approach (Eisenhardt and Graebner 2007) and conducted semi-structured interviews with 23 civil servants who adopt agile in 19 public administrations at various German governmental levels (Table 1). Each administration functions as an independent analytic unit, highlighting the significance of the real-world context in which the agile adoption phenomena takes place (Eisenhardt and Graebner 2007).

Table 1. Description of public administrations.

	Public administration	Number of interviewe(es) (n=23)	Department	Position	Level of experience (with agile)	Supplementary secondary data
A	City – Small > 20k	1	Management/Strategy	Head of Department	Low	Organization chart
B	City – Medium 20 – 100k	1	Human Resources	Employee in Department	Low	-
C	City – Large > 100k	1	Human Resources	Head of Department	Low	Blog article
D	City – Large > 100k	1	Management/Strategy	Head of Project	Low	Official report
E	City – Large > 100k	2	Human Resources	Head of Project	Low	-
				Employee in Project	Low
F	City – Large > 100k	1	Management/Strategy	Head of Department	Low	-
G	City – Large > 100k	1	Management/Strategy	Head of Department	Low	Official guideline for agile initiative
H	City – Large > 100k	2	Human Resources	Employee in Department	Medium	-
				Head of Project	Low
I	City – Large > 100k	1	Management/Strategy	Head of Department	Medium	Internal strategy documents
J	City – Large > 100k	1	IT-department/Digital program	Head of Project	Medium	-
K	Regional – Federal state public administration	2	Other (Social department)	Employee in Department	Low	Presentation on agile initiative
				Employee in Department	Low
L	Regional – Federal state ministry	1	Management/Strategy	State Secretary	Low	-
M	Regional – Federal state ministry	1	IT-department/Digital program	Head of Project	High	Internal strategy documents; Presentation on agile initiative
N	Regional – Federal state ministry	1	IT-department/Digital program	Head of Project	Medium	Organization chart; Presentation on agile initiative
O	National – Ministry	1	IT-department/Digital program	Head of Department	Medium	Organization chart
P	National – Ministry	2	Management/Strategy	Policy Officer	Low	-
				SeniorGovernment Official	Low
Q	National – Federal office	1	Other (single project)	Employee in Department	Low	Blog article
R	National – Federal office	1	IT-department/Digital program	Head of Department	High	-
S	National – Federal office	1	IT-department/Digital program	Employee in Department	Medium	Official report

As suggested by Mergel (2016, 522), ‘[i]t is important for e-government and digital innovation scholars to conduct country-level studies (…) to increase our understanding of agile innovation management’. Our focus centred on German administrations since extensive online research revealed a high prevalence of agile in German administrations in contrast to other European countries. This observation can be traced back to the following development. Germany is one of the most rule-oriented bureaucracies in the world (Kuhlmann et al. 2021, 3), with ongoing discourses surrounding the administrations being overly focused on producing and following rules and functioning in silos. These factors increasingly hinder problem-solving capacities and complicate the digitalization of public services (Kuhlmann et al. 2021, 332; Mergel, Brahimi, and Hecht 2021). Consequently, the 2021–2025 coalition agreement emphasizes the promotion of agility and digitalization through interdisciplinary, creative approaches, user-centricity, and innovation (Koalitionsvertrag zwischen SPD, Bündnis 90/Die Grünen und FDP 2021, 7). Germany is therefore a highly interesting country to study agile adoption since the results of this study hold promise not only for Germany but also for less rule-oriented bureaucracies.

We employed purposeful theoretical sampling to choose our cases (Eisenhardt and Graebner 2007). This selection approach entails opting for cases that can provide theoretical insights into the novel phenomenon of agile adoption within public settings. The selection was based on a thorough approach, including extensive online research to identify relevant agile-related content in connection with public administrations, reviewing public sector event programmes, blog and newspaper articles. Additionally, we used LinkedIn to identify individuals holding agile-related positions within public administrations. Furthermore, we asked at the end of each interview if they knew other civil servants who currently adopt agile. We continued with this snowballing approach until none of the 23 interviewees were able to point us to any other than the 19 administrations that made up our final sample. These selected cases can be considered early adopters of agile within the German public sector. Our investigation led us to conclude that we had successfully identified all instances of agile adoption among these early adopters.

The selected administrations represent diverse perspectives, varying in organizational size, department, hierarchy level, and experience with agile (Table 1). This diversity enables us to gain insights into how agile is adopted across different public settings. By including civil servants from various backgrounds, we identified common challenges encountered during agile adoption and share potential solutions. The majority of our cases involve cities of different sizes, four regional-level (Bundesländer) and five federal-level public administrations adopting agile. The civil servants who work in these selected administrations actively lead the adoption process and apply agile in their daily work. We conducted interviews with two civil servants simultaneously in four administrations, as they sought to complement each other’s expertise.

The questionnaire was structured into three parts, the first focusing on civil servants’ previous work experience, including their prior encounters with agile. The second part delved into their understanding and translation of agile values, principles, and methods within their specific organizational setting. The third part addressed the necessary organizational changes required for agile adoption and encountered challenges. Data collection was conducted between December 2021 and May 2022, with video call-based interviews averaging one hour.

Analytical method

Interviews were recorded, transcribed, and coded using ATLAS.ti qualitative analysis software. Addressing our research questions, we employed a combination of deductive and inductive coding (Miles, Huberman, and Saldaña 2014, 237–38; Mayring 2014, 104). In the first part, we use a deductive approach, deducing categories from our theoretical framework translation theory (Røvik 2016), identifying the employed translation modes across the cases. For the second part, focusing on the challenges of agile adoption and their resolution, we adopted an inductive coding approach and inductively developed our categories from the data.

For consistency, coding was conducted by one scholar (who conducted all interviews) and cross-checked by another (who participated in a third of the interviews). More specifically, the cross-check consisted of reviewing five coded interview transcripts, checking every coded passage against the deductive coding categories and assessing the appropriateness of the inductively developed codes. No formal inter-coder reliability tests were conducted. Only minor disagreements were found using this procedure, which were subsequently discussed and resolved in the presence of all authors. This resulted in small modifications to the coding scheme, such as merging two similar categories into one to ensure that the final categories were clearly distinct from one another and discussing and deciding on the exact meaning of certain categories. Finally, we contrasted and compared the cases to identify patterns, differences, and similarities based on the translation of agile employed, as well as other characteristics including experience with agile and organizational context.

To substantiate our findings, we triangulated interview data with public administrations supplementary secondary data. This supplementary data included official reports, organization charts, guidelines, or presentation materials (Table 1). This methodological choice aimed to ‘validate information obtained through interviews by checking programme documents and other written evidence’ (Patton 1999, 1195). Validation was important as research in human resource management has shown that intended policies and lived or perceived policies do not always align, and that written documents often fall short in catalysing actual change (Khilji and Wang 2006; Wright and Hisae Nishii 2013). Written documents only gain significance when employees are aware of them and hold their own perspectives. Hence, the interviews allowed for a more critical assessment, especially with regard to the challenges of agile adoption, thus taking a more central role in our analyses compared to the written documents.

Results

Adopting agile in public administrations

We sought to answer the research questions of how public administrations translate agile and how they address the associated challenges. We begin here with the results on the first question. Using translation theory as a lens for adoption, we categorize the translation of agile into three types arising from different translation modes, agile as culture change, agile as governance, and agile as methodology (Table 2). Due to the divergent operating environment and objectives of public administrations compared to private companies, a straightforward replication of agile into the recipient administrations was never observed. Consequently, adaptations were necessary in all cases.

Table 2. Types of agile translations.

	Agile Government
	Agile as culture change	Agile as governance	Agile as methodology
Cases	E, L, P, Q	A, C, D, F, G, H, K, O	B, I, J, M, N, R, S
Translation Mode	Modifying → Omission	Modifying → Omission	Modifying → Addition
Agile Translation	Agile as a cultural concept, cultivating an agile mindset and incorporating some agile tools from methods like Scrum.	Agile as an overarching framework to facilitate cross-functional collaborations.	Agility, closely aligned with the values and principles described in the agile manifesto, involves the application of agile methods tailored to the public sector’s context through the incorporation of additional functions or tools.

Agile as culture change

The first group of cases translates agile as a cultural concept, emphasizing the cultivation of an agile mindset characterized by trust, appreciation, employee empowerment, feedback, and a positive error culture. Although they incorporate some agile tools from methods like Scrum, their main focus lies in enhancing interpersonal relationships and making them more meaningful. Tools that they mention are daily standup-meetings, retrospectives, Kanban boards, or stakeholder analysis. Despite not intentionally adhering to agile values or principles, they endeavour to customize agile to suit the unique context of public administration, as highlighted by one interviewee:

We spent a lot of time finding our values. We had a whole workshop where we thought about what these agile values actually mean in a federal administration, because many of these agile things come from software development, economy, or the startup corner. Adapting phrases like ‘fail fast, fail better’ becomes challenging within a ministry. We cannot simply say that we’re going to just throw out a law and if it’s crappy, we’ll just make another one. (Case P)

Consequently, agile undergoes significant modifications from its original context (software development and private sector), with many aspects toned down or removed. We therefore assigned the translation mode modifying with omission to these cases (Table 2). The civil servants in these cases have no prior experience in agile and have independently appropriated the concept for their administration. Notably, agile as culture change is often (in three out of the four cases, namely E, L, and P) initiated within departments including human resources, management, or strategy, while one case (Q) was a project. This focus on transversal departments can be explained by the fact that changes in organizational culture were intended to take effect across all departments, making the transversal departments ideal entry points.

Agile as governance

The second group of cases has similarities to the first group, particularly in terms of embracing an agile mindset and employing customized agile tools. However, this group is distinct from the previous group insofar as these cases challenge the traditional Weberian bureaucracy by working more horizontally and cross-functionally. Thus, their primary objective revolves around leveraging synergies between departments or silos to increase speed and efficiency. Consequently, these administrations adopt agile as an overarching framework to facilitate new horizontal working structures which is why we labelled the group ‘Agile as Governance’. Their emphasis primarily centres on the agile principle of self-organized teams and partly also on internal user collaboration with other departments. Among the eight administrations in this group, three have established alternative governance structures that coexist alongside the official organization chart, sometimes equipped with their own financial and personnel resources. Thus, agile as governance enables cross-functional collaborations involving multiple stakeholders or departments, making it well-suited for complex organizational tasks ‘open to solutions’ (case D). These tasks span a spectrum of domains from process reengineering (C, K) to policy development (A, D, G). Nonetheless, interviewees emphasize the importance of maintaining standard procedures for repetitive tasks and processes requiring high legal security, such as issuing ID cards or building permits.

Similarly to the previous translation mode, agile as governance affects departments all over the administration, which is why the change was introduced primarily from transversal departments such as human resources, management, or strategy (in six out of the eight cases, namely A, C, D, F, G, and H). In one case (K) the social department led the change and, in another case, (O) it was led by the IT department. Overall, these administrations primarily focus on the agile principle of self-organized teams with an emphasis on innovative governance structures. We therefore assigned the translation mode modifying with omission to these cases (Table 2).

Agile as methodology

The third group comprises seven cases adopting agile as methodology, closely reflecting the original values and principles outlined in the agile manifesto. We labelled their approach ‘agile as methodology’, as all but one of these cases utilize Scrum and thus operate with established agile methods. One administration employs the Scaled Agile Framework (SAFe), designed for scaling agile in large organizations with multiple teams and extensive interdependencies. These cases use agile methods primarily to accelerate digital transformation initiatives, to develop digital products, and digitalize processes. Agile as methodology is predominantly adopted in IT-departments (in five out of the seven cases, namely J, M, N, R, and S), thus in software development or digital programmes. The two remaining cases adopted Scrum across the entire city administration, which is why the HR department (case B) and a management/strategy unit (case I) were charge of the agile adoption process. Prior to adopting Scrum, two cases, similar to the second group, implemented new governance structures that coexist alongside the official organization chart and deviate from the traditional hierarchical silo structure, providing greater freedom to employ agile methods.

All cases reported modifications to the original agile method to suit the specific public context. Two cases introduced an additional role within the Scrum framework (in addition to the Scrum master, product owner, and development team) responsible for bridging the gap between the agile team and non-agile departments. This role was referred to as the ‘proxy-product owner’ by one administration. Moreover, the administrations using Scrum reported to enrich the agile method by incorporating supplementary methods or tools such as stakeholder analysis, user journeys, or personas:

I claim that nobody in public administration does Scrum in its pure form, because I don’t know any project management framework that is as rigid and also as incomplete as Scrum. (Case N)

In two other administrations the level of agility in their projects fluctuated, allowing them to switch between traditional project management and agile project management (Scrum) depending on the project phase. During the procurement phase, for instance, traditional project management was deemed more suitable than agile project management. We therefore assigned the translation mode modifying with addition to these cases (Table 2).

Addressing the challenges

We now turn to the second research question and explain how the administrations address the challenges associated with the adoption of agile.

Highlighting the difficulty to translate agile to administrative contexts across all translation modes, we found that 10 cases solicited varying degrees of external support, ranging from one-day workshops conducted by agile coaches to comprehensive consulting mandates spanning several months:

We needed someone to take us by the hand. We wouldn’t have been able to do it ourselves from within the system. Their role was actually to keep pushing us so that initial developments didn’t fizzle out again and also to provide support. (Case I)

However, external support often only initiated the translation process, whereas later on, considerable resources were allocated to retaining the knowledge within the administration and cultivating agile knowledge among employees. For instance, six administrations trained agile coaches, agile pilots, or multiplicators, who actively promoted and disseminated agility throughout the administration, serving as agile experts.

For agile as culture change, a specific major obstacle relates to the difficulty of employees to comprehend the concept of agile. One interviewee (P) highlights instances where employees mistakenly associated agile with the dissolution of hierarchies or the expectation of immediate problem-solving. Moreover, employees expressed concerns about potential exclusion if they chose not to participate, and were concerned that their previous work methods may have been incorrect. To address these concerns, these administrations adopted their specific agile translation through voluntary employee engagement, facilitated by workshops and e-learning materials, without employing coercive measures. The cultural transformation unfolds as a grassroots movement, driven by tailored agile values and tools.

Agile as governance faces similar challenges to the first group. Significant resources need to be invested in communication and convincing members of the usefulness of the agile approach. One case (K) stands out as the sole case where agile as governance failed. In this case, agile was embraced at the grassroots level, yet it encountered only partial support by the head of department. Ultimately, this lack of support resulted in the failure of the agile initiative after two years because the head of department refused to provide further funding for the two employees that initiated the new agile governance structure. In the other cases, top-management support was an important factor to successfully drive the organizational changes. This included communicating about the upcoming change and new working offers, conducting personal interviews with employees to minimize concerns, and advertise first success stories:

But what is a much stronger tool is all projects that develop positively (…) and we report on these successful projects quite proactively, and that has already generated changes in others, (…) because some really lack the courage and imagination. (Case G)

Labelling the change and actively conducting targeted marketing is also part of navigating the encountered challenges:

Change management also has a lot to do with marketing. I always laugh with my colleagues that I actually see myself as a marketing manager as I try to label all the things we do. We also had the graphic designer who created our own logo for the ‘Fit in Leadership’ seminars. It is important to have a corporate design and identity. (Case A)

New horizontal working structures pose particular challenges because leaders must give up authority and empower employees to be self-organized and work autonomously. This new independence of employees brings bureaucratic constraints:

Agile teams require self-organization and personal responsibility, but civil service law often conflicts with these principles, since responsibility is associated with a certain remuneration group. How do you evaluate someone who is responsible for their own work? And who signs a team member’s vacation application if they work independently? (Case D)

Notably, discussions within the first and second group focused on the appropriate labelling of the agile transformation, with numerous civil servants highlighting that the term ‘agile’ has acquired negative connotations within the public administration. Employees often associate it with erroneous notions. One administration even held a dedicated ‘renaming workshop’ to find a more fitting designation for their translation of the concept.

The adoption of agile as methodology in bureaucratic public administrations also poses challenges. We found that user collaboration, a core value of the agile manifesto, does not play a central role in the observed agile translations, especially when considering external users or citizens. This was the case for all three translation modes, but is most surprising in this last group focusing on agile as methodology. In contrast, collaboration with internal users is a somewhat more frequent occurrence. Many administrations reported that employees exhibit reluctance to engage openly with the public, fearing the potential exposure of mistakes:

It is very difficult to let citizens participate in developments. Because people are simply afraid that something could leak out that is not going well, and […] we also have the (political) opposition, who is always very happy to take up such things. (Case M)

Collaborating with non-agile departments or teams is another challenge, as they may lack understanding of agile terminologies or be hesitant to deliver something that is not perfect. One solution is integrating an additional role in the Scrum framework to this interface. Then, recruiting staff with agile skills becomes challenging once agile teams reach high competency levels in agile methods, due to a limited pool of qualified candidates. Public administrations may not always control the recruitment process, and HR departments may struggle to grasp agile values and methods, making it difficult to find candidates who would be a cultural fit. Moreover, IT departments in public administrations face competition from the private sector, which offers significantly higher salaries. Consequently, many administrations employing agile as methodology rely on freelancers and short-term contracts:

The solution for me is what I call body leasing. Means I try to get contracts with medium-sized companies and agree on contracts for consultancy days. But consultants in quotation marks, because in the end they are software developers, but I treat them the same as permanent employees. However, the Scrum master, who leads the team, has to be an internal employee. (Case R)

Short-term contracts, however, can hinder knowledge transfer within projects, as freelancers may lack organizational culture or processes familiarity. Furthermore, civil service regulations often prevent public administrations from creating new positions with titles such as ‘product owner’. Instead, they may use the more generic term ‘IT administrator’ to describe said role. However, this can make it difficult to attract the right type of specialist, as individuals interested in product owner roles may not be attracted by a role called IT administrator. Nonetheless, some of the administrations have found ways to market their vacancies through other channels:

We try GitHub, I post a lot on LinkedIn, I’m active on Twitter, so we play all kinds of channels to be successful with recruitment. (Case R)

Discussion

With this study we sought to shed light on the questions of how public administrations translate agile to their context and how they address the associated challenges. To analyse the first question, we relied on translation theory (Røvik 2016) and identified three ways in which agile is translated into public settings. Our findings suggest that there is no one-size-fits-all approach.

The first administrations group adopts agile as a concept for culture change. They focus on cultivating an agile mindset and incorporate agile tools from methods like Scrum. However, they significantly modify agile from its original context, toning down or removing aspects. The second group adopts agile as a governance framework to foster synergies between departments or silos and enhance cross-functional collaborations to boost speed and efficiency. In some cases, the new governance structures lead to the coexistence of two organizational forms (cp. Polzer et al. 2016). Agile as culture change and as governance are mostly adopted in HR, management, or strategy departments which supports the suggestion made by Mergel and colleagues (2020) that agile government has evolved from a pure software development method to a concept applicable to generalists and managers alike. The third group adopts agile as methodology and largely adheres to the original values and principles outlined in the agile manifesto (cp. Beck et al. 2001). However, they still incorporate new elements to make agile methods such as Scrum suitable to the public context. Agile thus evolves differently in different settings (cp. Cram and Newell 2016; Czarniawska and Sevón 1996) ranging from numerous modifications to a relatively close imitation of the concept.

We suggest that the three agile types build on each other. Embracing agile as culture change sets the stage for both agile as governance and as methodology. Moreover, the cultivation of an agile culture and the establishment of agile governance lay the groundwork for the eventual adoption of agile as methodology, given sufficient time and experience. This sequential approach suggests that agile as culture change and governance can act as foundational pillars for organizations aiming to adopt agile as methodology. We suggest the following proposition:

Proposition 1.

Agile is adopted in public administrations in three different translation modes that build on each other, namely agile as culture change (focused on promoting an agile mindset), agile as governance (focused on cross-functional collaboration), which can service as foundational pillars for agile as methodology (focused on established agile methods).

Regarding the question of how public administrations effectively address the challenges associated with their specific agile adoption, we found differences between the three groups suggesting that different translations of agile led to different challenges. Nevertheless, the cases show that many of these challenges can be addressed with adequate change management and creative workarounds (e.g. proxy-product owner). Our identified challenges partly corroborate with previous findings, supporting that agile evokes fears among leaders due to its demand for a new form of leadership (cp. Mergel 2016, 2023; Simonofski et al. 2018) and often leads to misinterpretation (cp. Baxter et al. 2023; Simonofski et al. 2018). Although previous research indicates that procurement poses significant challenges in adopting agile methods (Mergel, Ganapati, and Whitford 2020; Mergel 2016; Simonofski et al. 2018), our interviewees did not mention procurement law as a challenge, possibly because most of the agile as methodology cases develop their own software or digital products, minimizing reliance on procurement processes. Additionally, one administration reported using traditional project management instead of agile during the procurement phase. Concerning the agile as methodology cases, we found that recruiting IT talents with specific profiles is a major difficulty (cp. Baxter et al. 2023; Mergel 2016; Simonofski et al. 2018), a challenge that we will address in greater detail in the following section. To summarize the challenges associated with agile adoption, we suggest the following proposition:

Proposition 2.

The main challenges for agile as culture change and agile as governance include communication and convincing members of agile usefulness, whereas agile as methodology faces challenges regarding cross-mode (non-agile vs. agile) collaboration and recruiting skilled staff.

We will now examine the question of what drives the adoption of agile in such varying ways. For the cases adopting agile as methodology, we see a combination of possible explanatory factors. First, prior professional experience and education significantly shapes familiarity with agile. For instance, civil servants with a history in the private tech or IT sectors or relevant educational backgrounds tend to seek positions in IT or digital departments. This aligns with our findings, indicating that the civil servants embracing agile as methodology typically possess medium to high levels of agile experience, which these individuals have learnt in previous positions. The second explanation concerns the government level as an explanatory factor. Agile as methodology is mainly used in IT or digital departments in higher government levels, likely influenced by recent efforts of the German government to digitalize 575 public services as part of the online access law (Kuhlmann et al. 2021, 335). This top-down approach places significant pressure on federal ministries and regional states, but leaves the process undefined for the numerous municipalities (Kuhlmann et al. 2021, 333). Consequently, higher government levels are under greater pressure to drive digitalization forward, leading to an increased use of agile as methodology. Third, regarding the size of administrations, larger administrations, such as a federal ministry, usually have larger IT departments with the resources and expertise to implement agile as methodology effectively.

Weaker adoption of agile, as in the cases adopting agile as culture change or agile as governance, can be the result of a challenge in recruiting staff with agile skills. Baxter and colleagues (2023, 5) suggest that agile is driven by ‘an emerging professional discipline that includes several professional associations and formal qualifications’. They also note that wider adoption of agile is hindered by a lack of agile expertise, namely individuals who really understand agile and can translate the concept into something tangible. This scarcity of agile expertise coupled with higher salaries in the private sector, makes it less likely that individuals with agile skills will fill positions in the public sector that are not related to IT or digital topics. Moreover, translation theory underscores the importance of ‘translation competence’, which emphasizes contextual bilingualism, implying a deep understanding of both source (private sector) and recipient contexts (public sector) (Røvik 2016). This suggests that effective translation of agile requires not only agile knowledge, but also an understanding of the functioning and particularities of the public sector. The agile as methodology cases therefore appear to benefit from employees with a solid foundation of agile knowledge who are also motivated to work in the public sector. Overall, this underscores the pivotal role of translators as ‘carriers’ of ideas (Wedlin and Sahlin-Andersson 2017) in driving organizational change.

The government level can also be an explanation for how agile has been translated in the cases of culture change or governance. Many municipal-level cases translate agile as governance and the explanation may be that municipal authorities in Germany enjoy greater autonomy in organizing their internal affairs than federal administrations, due to their extensive public duties and guaranteed self-governance rights (Kuhlmann et al. 2021, 40, 113, 114). Agility as governance, i.e. increased cross-silo cooperation with various actors and stakeholders, can thus help the municipal level to manage its numerous public tasks more efficiently. However, it is worth noting that this observation may be influenced by the higher number of cases at the municipal level compared to the other levels, potentially introducing bias.

Finally, the extent to which top-management supports agile can explain the extent to which agile leads to real structural and organizational change in the administration. In case K, the absence of backing from the department head led to the failure of agile governance, with refusal to allocate additional funding to the initiative. Conversely, cases B and I stand out for their adoption of agile as methodology throughout the entire city administration, distinguishing them as the only instances where agile as methodology is not confined to IT or digital departments. We attribute this broad implementation to the extraordinarily strong support from top-management, combined with a good level of agile experience (case I) and strong external support (case B). These findings emphasize the need for supportive hierarchical structures to facilitate successful organizational transformations (cp. Baxter et al. 2023; Mergel 2023; Nielsen, Wæraas, and Dahl 2020).

We also found that administrations often solicit external support when adopting agile, which resonates with previous research from translation theory proposing that individuals can serve as carriers of ideas, transferring them between settings and shaping them into successful models through editing (Wedlin and Sahlin-Andersson 2017). Similarly, Nielsen and colleagues (2020) showed how a leadership concept was translated into Danish municipalities through collaboration between consultants, researchers, and public sector managers. Therefore, it is conceivable that also in our cases consultants, or other individuals from the external support, have already translated the agile concept in some way before it was proposed to the respective administration. Furthermore, it is plausible that internal agile multiplicators (which many of our analysed administrations trained after soliciting initial external support) then made further modifications to the agile concept over time. We assert the following proposition:

Proposition 3.

Initially, in all three translation modes, public administrations rely on external support to initiate the adoption process, whereas subsequently, administrations expend great efforts towards internalizing and retaining the agile knowledge by training employees in-house.

Regarding user collaboration, our findings contradict the Mergel and colleagues (2020) observation that in agile government, external users play a central role in decision-making. Instead, they resonate more closely with the observations made by Mergel and colleagues (2019), who found a notable absence of citizen involvement in government digital transformation efforts. Our research underscores that agile in public administrations predominantly prioritizes internal operations, particularly evident in cases emphasizing agile as governance, even if collaboration with internal users is prevalent to some extent. Nevertheless, it must be emphasized that the focus on internal processes can still have a positive impact on citizens and taxpayers, as costs can be saved, and goals can be achieved more quickly by increasing internal efficiency. Normann (1991) coined this as the ‘virtuous circle’ of service improvement which highlights that the quality of internal processes significantly influences and contributes to the quality of external services (Osborne, Radnor, and Nasi 2013). This leads us to the following proposition:

Proposition 4.

Given that agile adoption is still a relatively new phenomenon in many public administrations, a key focus is on internal culture, process-optimization, and structures, whereas external user collaboration (e.g. with citizens) is not (yet) at the fore in many cases.

Taking these findings into consideration, we propose the following refined, empirically grounded definition of agile government: ‘Agile government is a form of governance innovation consisting of organization-specific mixes of cultural, structural, and procedural adaptations geared towards making public organizations more flexible in changing environments, ultimately pursuing the goal of increasing efficiency, effectiveness, and user satisfaction’. It is grounded in a set of values and principles from agile software development, focusing on cross-functional and increasingly user collaboration, self-organization, early feedback and learning, sustainability and simplicity, and it is implemented with varying adoption scope depending on the organization.

Implications for theory

We observed that the first group adopts agile as a catalyst for cultural change (cp. Mergel, Ganapati, and Whitford 2020). However, this cultural shift does not significantly alter organizational structures or administrative routines. In contrast, agile as governance or methodology brings, in some cases, substantial changes departing from Max Weber’s bureaucratic model, challenging fundamental characteristics like the division of labour. Our results thus corroborate Mergel and colleagues’ (2020) assertion that agile represents ‘A New Way of Governing’, capable of reshaping government and public management. Following Mergel and colleagues (2020) suggestion that shifts in government operations often evolve through public management reforms, the question arises: Can agile government emerge as a distinct reform paradigm, addressing limitations of previous models like New Public Management (NPM) (Stoker 2006)?

Beginning with NPM, criticized for its intra-governmental focus in an increasingly complex world (Kuhlmann et al. 2021; Osborne 2006, 2018), contemporary discussions on reform paradigms indicate a shift away from singular reform approaches. Instead multiple paradigms, such as digital era governance (Dunleavy et al. 2006), public value management (Stoker 2006), and collaborative governance (Greve 2015), now offer a dynamic agenda for public administrations to tailor to their unique needs (Greve 2015; Ritz, Sinelli, and Neumann 2016). However, achieving a user-centric approach remains challenging despite successive waves of public service reforms (Strokosch and Osborne 2020; Osborne et al. 2021). Agile offers promise in finally embedding this user-oriented perspective within public organizations, fundamentally aligning them with service provision (Osborne et al. 2021).

Nonetheless, our findings reveal that administrations adopting agile mainly focus on enhancing internal capabilities and governance, lacking readiness for extensive engagement of external stakeholders or citizens in public service delivery (cp. Strokosch and Osborne 2020). This echoes prior debates on the lean approach, having been proposed to drive cost savings and quality enhancement in public organizations (Radnor and Walley 2008). Lean has also been advocated as a strategy for public service reform (Radnor and Osborne 2013), thus sharing similarities with agile, both originating from the private sector and offering diverse techniques with varied adaptations across different public organizations (Radnor and Walley 2008). However, lean has been criticized for its limited focus on service users (Radnor and Walley 2008; Radnor and Osborne 2013). The similarity between lean and the current agile discussion becomes even more evident when substituting lean with agile in the following excerpt from Radnor and Osborne (2013, 279): ‘Our contention, though, is that if lean (agile) is implemented (…) with a focus only on internal, rather than external, users, then this will severely limit its effectiveness as a public service reform strategy’.

Summing up, even though agile government may currently lack emphasis on external user collaboration, its significance could grow as internal governance structures become more deeply embedded (cp. Normann 1991; Radnor and Johnston 2013). Thus, while agile government may not (yet) offer a definitive solution for genuine public service logic (Osborne 2018), it can still complement other reform paradigms. Agile as methodology, for instance, has the potential to enhance digital project governance, thereby serving as an extension of digital era governance (Dunleavy et al. 2006).

Limitations and future research

This study has three main limitations. First, our interviewees were closely involved in the adoption of agile, likely making views idealistic, whereas other employees affected by the change may hold different opinions. Future research could therefore also assess critical opinions. Second, the present study merely captures one point in time. Future research could focus on adoption dynamics and study how agile diffuses and institutionalizes inside organizations (Colyvas and Jonsson 2011). Moreover, future research could analyse how external consultants and internal agile multiplicators influence agile concept translations. We also recommend conducting more research on outcomes of agile in public settings, such as job satisfaction, performance, innovation capacity, or user centricity. Thirdly, our research focused on 19 administrations in Germany, which are currently frontrunners in agile adoption, and might restrict our findings’ generalizability. Thus, future research could explore whether the three translation modes of agile hold true across a broader range of administrations and countries. Lastly, we noticed that our cases have a high geographical proximity and are mostly located in western Germany. Future research could study the impact of the behaviour of proximate organizations as suggested by De Vries and colleagues (2018).

Conclusion

Our empirical investigation of agile translations in public administrations reveals that while the initiation of a cultural shift is a pivotal starting point, the true transformative potential lies in the adoption of agile as governance framework or as methodology. Agile methodology has the potential to enhance the governance of digital projects aligning with the demands of the digital era. While agile government currently prioritizes internal aspects over stakeholder engagement, it demonstrates the potential for future development in user collaboration. Future research is essential to fully understand the impact and long-term implications of agile government, and to ensure its effective integration into the broader spectrum of reform paradigms.

Author’s contribution

Oliver: Conceptualization, Data collection (conducted interviews together with Pascale whenever possible), Design of methodology, Supervision (as Pascales’ PhD supervisor), Writing – Review and editing. Pascale: Conceptualization, Data collection (conducted all interviews), Data analysis, Design of methodology, Writing – original draft, Writing – Review and editing. Carina: Conceptualization, Design of methodology, Writing – Review and editing. All authors regularly exchanged about the progress of the manuscript, provided critical feedback, helped shape the analysis, and participated in the revisions.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Notes on contributors

Oliver Neumann

Oliver Neumann is an assistant professor in the Swiss Graduate School of Public Administration (IDHEAP) at the University of Lausanne, Switzerland. He received his PhD in Management at the University of Bern, where he also worked as a postdoctoral researcher in Information Systems. His current research interests include public sector innovation, artificial intelligence, agile government, behavioral public administration, and digital transformation.

Pascale-Catherine Kirklies

Pascale-Catherine Kirklies is a doctoral researcher and graduate assistant in the Swiss Graduate School of Public Administration (IDHEAP) at the University of Lausanne, Switzerland. Her research interests include public sector innovation, agile government, and citizen-state interactions.

Carina Schott

Carina Schott is an assistant professor at the Utrecht School of Governance. She conducts research at the interface of Public Management and HRM in the public sector. Specifically, her research focuses on employee motivation, social belonging, professionalism, and leadership in various public sectors.

Appendix A.

Agile values and principles

Table

Agile Values

(1) Individuals and interactions over processes and tools (2) Working software over comprehensive documentation (3) Customer collaboration over contract negotiation (4) Responding to change over following a plan

Agile Principles

Our highest priority is to satisfy the customer through early and continuous delivery of valuable software. Welcome changing requirements, even late in development. Agile processes harness change for the customer’s competitive advantage. Deliver working software frequently, from a couple of weeks to a couple of months, with a preference to the shorter timescale. Business people and developers must work together daily throughout the project. Build projects around motivated individuals. Give them the environment and support they need, and trust them to get the job done. The most efficient and effective method of conveying information to and within a development team is face-to-face conversation. Working software is the primary measure of progress. Agile processes promote sustainable development. The sponsors, developers, and users should be able to maintain a constant pace indefinitely. Continuous attention to technical excellence and good design enhances agility. Simplicity, the art of maximizing the amount of work not done, is essential. The best architectures, requirements, and designs emerge from self-organizing teams. At regular intervals, the team reflects on how to become more effective, then tunes and adjusts its behaviour accordingly.