While research on innovation is extensive, less is known about public sector innovations (PSI) as distributed processes based on both internal and external knowledge. Research has shown that the public sector often imitates private firms’ innovation processes, particularly when introducing new solutions such as digital processes. This occurs despite the fact that the public sector’s goal is not profit, but rather the provision of quality services. In this paper, we explore the consequences of such imitation for PSI processes related to digitalization, with a particular focus on how innovation orchestration functions in these settings.
Empirically, the paper is based on a regional government initiative that used orchestrators to intermediate innovation processes. Interviews and focus groups were the main data source, complemented by internal documents from the initiative.
The study develops five propositions that highlight the paradoxical nature of PSI orchestration: the need for top management involvement, which can also stifle innovation; the necessity of orchestrator engagement in implementation, which limits their boundary-spanning role; the benefits of organizing innovation away from daily routines, which simultaneously hampers implementation; the tendency for collaborations among diverse stakeholders to emerge as outcomes rather than drivers of the process; and how the public culture orients the process to resources rather than value.
The paper contributes to previous research by addressing PSI. More specifically, it does so by integrating such research with innovation orchestration. The paper highlights how these processes differ substantially from those in the private sector, particularly with regard to the ability to orchestrate relational aspects as an antecedent to innovation.
Introduction
The public sector in general has neither money to spend on innovation nor staff capable of orchestrating and implementing these – yet it is still under heavy pressure to improve efficiency (Main project leader, case project).
Historically, innovation and innovation research have been associated with the private sector. More recently, however, there has been a growing effort to encourage public sector innovation (PSI), along with attempts to transfer innovation practices from the private to the public sector (Bugge and Bloch, 2016; De Vries et al., 2018). Adopting private sector innovation methods in the public sector may not always create the best fit. This is particularly the case given that innovation in the public sector – defined as “ideas, processes, products (services) or procedures, new to the relevant unit of adoption, designed to significantly benefit the individual, the group, the organization in a wider society” (Wihlman et al., 2016, p. 45) – is primarily aimed at creating public value, rather than generating a return on investment (Bason, 2018; Kusumasari et al., 2019). This distinction underscores a fundamental difference between the sectors: in the public sector, financial resources are a means to achieve societal usefulness or public value, whereas in the private sector, usefulness is often a means to generate profit.
In the private sector, orchestration of innovation processes has become a popular method for managing innovations and bringing together sources of knowledge from internal and external actors. Orchestration implies third-party involvement in innovation processes in the form of a facilitating intermediary (Dhanarag and Parkhe, 2006; Sieg et al., 2010; Hurmelinna-Laukkanen and Nätti, 2018). It includes linking various resource-owners (Andresen, 2020), influencing and mapping participants’ competences, coaching and coordinating relationship development, facilitating integration and co-creation (Aarikka-Stenroos et al., 2017), and enabling resource sharing (Andresen, 2011) – in short, intermediating innovation processes and their accompanying relational aspects. Orchestration is central to many open innovation practices involving multiple parties (Aarikka-Stenroos et al., 2017), while professionalizing the innovation process through third-party intermediation (Andresen, 2020). Such orchestration has primarily been theorized in relation to private sector innovation. However, less is known about orchestration in relation to PSI, where it represents one of the methods imported from the private sector. Therefore, the purpose of this paper is to describe and discuss how innovation orchestration works in PSI settings. The underlying objective is to examine whether and how the private sector methods create a fit or misfit with the public sector, and what adjustments need to be made in PSI processes.
In this paper, we study orchestration of innovation processes that bring together internal and external actors to create PSI in a Swedish regional government initiative focused on digital solutions. Our findings are summarized as five paradoxical propositions: top managers are essential yet may constrain innovativeness; orchestrators are needed for implementation but risk losing their boundary-spanning role; and organizing outside daily routines fosters invention while hindering adoption. Moreover, collaborations often emerge as outcomes rather than drivers, and a resource-focused culture frequently overshadows the goal of creating public value.
The paper contributes to previous research by expanding knowledge on PSI. Research on PSI remains limited (Micheli et al., 2012; Salge and Vera, 2012; Bugge and Bloch, 2016; Wihlman et al., 2016; Demircioglu and Audretsch, 2017) but is thus gaining increasing attention in society. The differences between public and private sector innovation processes are of significant empirical relevance. They highlight that public sector innovators may be less skilled, less willing to take risks, and less inclined to think creatively during the innovation process – important aspects when considering the means-end logic and the characteristics of orchestration discussed in this paper. The focus on intermediation captures the PSI process, rather than viewing PSI as focused on useful outcomes as a further contribution.
The rest of the paper is outlined as follows: After this introduction, we situate the paper in its phenomenological context of the public sector by describing the specific conditions related to innovating in the sector and presenting previous research in the area. We then summarize prior research on innovation orchestration to establish understanding of the method. Next, the research design is presented, followed by the findings. The case is then analyzed, and the paper concludes with theoretical and managerial implications and suggestions for future research.
Theoretical background and framework
Public sector innovations – concepts and determinants
PSI thus refers to ideas, processes, products, services, or procedures that are new to the adopting unit and aim to benefit society in some way (West and Farr, 1990; Wihlman et al., 2016). These innovations are crucial for public sector service development in that they generate opportunities to improve service quality and efficiency (Steward-Weeks and Kastelle, 2015; Kusumasari et al., 2019), emphasizing productivity enhancement or improved services relative to invested expenses. Digitalization is a recurring approach for creating efficient processes in public organizations. PSI must be implemented within budget restrictions (Stewart-Weeks and Kastelle, 2015), since the public sector is strongly focused on budget control and typically does not allow losses (Agolla and Van Lill, 2016). This further highlights a key contrast with private sector innovation, where monetary gains are typically the primary objective and novel ideas or increased productivity serve as means to achieve financial returns. In the public sector, by contrast, financial resources are predetermined and allocated to generate value for citizens (cf. Öberg, 2020).
However, the differences do not end there (Demircioglu and Audretsch, 2020). The beneficiaries of PSI are predefined as the citizens of a country or region. Moreover, there is no traditional market (Bloch and Bugge, 2013). The budget restrictions and the organization of operations including political decision-making (Agolla and Van Lill, 2016; Wihlman et al., 2016), foster a conservative and reactive approach to problem solving (Potts and Kastelle, 2010; Palm, 2020). Consequently, the performance of PSI has often been described as low, with limited benefits for society (Albury, 2005; Meijer and Thaens, 2020).
Previous research on PSI has focused on economic development, growth, social issues, and digitalization (Micheli et al., 2012; Bloch and Bugge, 2013; Stewart-Weeks and Kastelle, 2015; Agolla and Van Lill, 2016). De Vries et al. (2016) summarize such research as focusing on definitions, types of innovation, antecedents, goals, and outcomes. This indicates a lack of research on the process and organizing of PSI. More recent literature has explored open innovation possibilities (Lee et al., 2012; Lopes and Farias, 2020; Mu and Wang, 2020), thus emphasizing that PSI may emerge through collaboration among parties (Sorensen and Torfing, 2011; Crosby et al., 2017) or with external actors as sourcing agents (Mergel, 2018).
In line with this reasoning, Bakici et al. (2013) refer to intermediaries in PSI but highlight private–public collaboration in such efforts. PSI may be initiated top-down, meaning that managers and politicians drive innovation, or bottom-up, where innovation emerges from incentive-driven employees or middle managers (Arundel et al., 2015; Kusumasari et al., 2019). The latter is a perspective that is, according to Lee et al. (2012), interestingly more spread in the public than in the private sector. However, given the gap between decision-makers at the top of the hierarchy and those at the bottom, Wihlman et al. (2016) and Wipulanusat et al. (2019) found bottom-up initiatives difficult to manage in public sector structures.
In collaborative efforts, not only multiple parties but also different organizational levels must interact – as demonstrated in this paper – and this is where orchestration plays a role: coordinating these efforts. In a framework inspired by De Vries et al. (2016), Kusumasari et al. (2019) argue for three levels of PSI enablers. First, the environment level portrays how innovations stem from co-evolution generated by external demands and pressures, interorganizational relations, and competition from other organizations. This category includes political, economic, social, technological, and legal factors, but also cooperation (cf. Agolla and Van Lill, 2013, 2016) and end-user involvement. A distinguishing characteristic from private sector innovation is that PSI often requires a high level of political commitment. Second, the organization level highlights how innovation originates from structural and cultural aspects, including resources for innovation, leadership style, learning, incentives, and structure. Adding to this are top management support, encouragement of staff to innovate, change and risk management, autonomy, collaboration, and norms favoring exploration – all enablers proposed by Wihlman et al. (2016). Agolla and Van Lill (2016) – stress that internal leadership commitment and leadership practices are particularly important here. Third, the individual level emphasizes the role of employees in creating innovations, based on autonomy, innovation skills and knowledge, creativity, demographic factors, and commitment.
Table 1 summarizes PSI across the three levels – environment, organization, and individuals – as a way to consolidate past insights and provide background for understanding PSI processes and their orchestration.
PSI across levels – difference to private sector innovation
| Environment related (government) | Public innovation | Note relative private sector innovation |
|---|---|---|
| Budget restrictions | PSI typically acts within budget restrictions, is focused on budget control (Stewart-Weekes and Kastelle, 2015; Agolla and Van-Lill, 2016). Innovation not prioritized under limited budgets (Wihlman et al., 2016). PSI spending stands in contrast to the public sector’s obligation to utilize resources effectively (Wipulanusat et al., 2019) | Money as end, innovativeness as means |
| Hierarchy and Bureaucracy | Bottom-up innovation initiatives difficult to manage in old-fashioned public sector structures (Wihlman et al., 2016; Wipulanusat et al., 2019). Public sectors are typically bureaucratic, hierarchical and complex (Agolla and Van Lill, 2016), a structure not well suited for innovation (Wipulanusat et al., 2019). Initiated top-down or bottom-up (Arundel et al., 2015; Kusumasari et al., 2019) | Open innovation assumes horizontal initiatives. Ecosystem innovations |
| Laws and regulations | Laws and regulation controls public sector approval processes and activities in order to prevent impulsive and unpredictable actions and to ensure uniformity (Mulgan, 2007; Wipulanusat et al., 2019). There might not be room for PSI if it does not fit existing laws and rules (Mulgan, 2007). Public sector organizations cannot change their business concept overnight due to laws etc. (Wihlman et al., 2016) | The political influence, governance and legal constraints of PSI make the challenges regarding innovation differ between the sectors (Smith et al., 2019) |
| The silo effect | Public sectors tend to be organized in divided departments, separated by walls of different power structures and finances, which hinders the spread of knowledge and cooperation, and thereby also innovation (Mulgan, 2007; Wipulanusat et al., 2019) | Efficiency focus on firm or ecosystem levels |
| Organization related | ||
| Lacking leadership | Poor leadership is a major barrier to PSI (Smith et al., 2019; Wipulanusat et al., 2019). Conflicting and irrelevant goals and messages creates problems (Wihlman et al., 2016). Frequent changes of management make it difficult to manage a full innovation process (Wipulanusat et al., 2019) | Innovation leaders. Creativity as non-hierarchical processes |
| A risk-averse culture | The public sector tends to steer away from innovative practices due to high risk and uncertainty (Micheli et al., 2012; Agolla and Van Lill, 2016). The public sector typically attracts risk-averting employees (Strow and Strow, 2018). Media coverage intensifies the risk-avoiding culture in the public sector (Mulgan, 2007) | Entrepreneurial and risk-taking employees tend to sort themselves into the private sector in search of rewards (Strow and Strow, 2018) |
| Individual related | ||
| Lack of incentives and rewards | Monopolistic sectors like the public one tends to lack competition motives (Mulgan, 2007). Traditionally greater punishments for failed innovations, than rewards for successful such (Mulgan, 2007; Wipulanusat et al., 2019). Performance pay is negatively correlated with work performance in the public one (Bryson et al., 2017) | Intensity of competition, as in the private sector, provides motive for innovation (Mulgan, 2007). Performance pay works well to motivate people in the private sector (Bryson et al., 2017) The public sector has traditionally not had the same organizational culture for rewarding successful innovation as the private sector (Wipulanusat et al., 2019; Bloch and Bugge, 2013) |
Orchestration of PSI
Orchestrating thus implies the handling of a range of internal and external actors. Orchestration is a well-researched phenomenon in the private sector, where it is strongly linked not only to open innovation but also to dynamic capabilities (e.g. Natti et al., 2014; Giudici et al., 2018; Helfat and Raubitschek, 2018; Reypens et al., 2019). Orchestrators typically play a complementary role to those developing the actual innovations, creating connections among innovating actors and guiding the process. Their intermediary positions are thereby pronounced, as are their skills in managing innovation processes and relationships among actors. Orchestrators provide a setting, or arena, offering ongoing evaluation, coaching, design-related services, and training programs, including inspiration and expert support for innovators (Klerx and Aarts, 2013; Kazadi et al., 2016; Hurmelinna-Laukkanen and Nätti, 2018). Perks et al. (2017) argue that by providing a type of platform, an orchestrator can help projects move away from typical hierarchical structures. In private sector practice, orchestrators primarily intermediate among parties within or between organizations working together on solutions, thereby emphasizing the horizontal intermediation of activities.
As intermediaries in the innovation process, orchestrators’ roles in the private sector may evolve as the innovation progresses (Hurmelinna-Laukkanen and Nätti, 2018). Research frequently reports on activities such as sensing, seizing, and reconfiguring (Giudici et al., 2018) or the related structuring, bundling, and leveraging (cf. Lindgreen et al., 2012) as part of a somewhat linear process of engaging with external parties. Dhanarag and Parkhe (2006) describe orchestration as involving knowledge mobility, innovation appropriability, and network stability. Two sets of capabilities are thus highlighted: those for organizing innovation and those for organizing collaborations (Nambisan and Sawhney, 2011) or, as Mumford et al. (2002) put it, expert and relationship orchestration. While orchestration has been shown to successfully support innovation in the private sector (Wareham et al., 2014), far less is known about how it functions – or how it must differ – in the public sector.
Given the public sector’s reverse logic between means and ends – where financial resources are the source rather than the outcome of useful services – along with political influence and limitations in competence and risk-taking, it becomes clear that PSI orchestration must account for a broader and different range of stakeholder perspectives. At the same time, orchestrators must manage capacity constraints and uphold quality standards within predetermined budgets. As a result, PSI not only introduces new dimensions to innovation processes but also challenges conventional logic. Yet our understanding of the actual and desirable conditions for effectively orchestrating PSI remains limited.
By bringing together the limited research on PSI—which emphasizes its differences from private sector innovation—with the literature on orchestration, we approach a gap in the PSI literature (cf. De Vries et al., 2016). We focus on the relational and the innovation operandi of orchestrators (Nambisan and Sawhney, 2011), exploring them in a setting that differs from most prior research. By juxtaposing the relational and innovation aspects of orchestration with the three levels of PSI outlined by Kusumasari et al. (2019), we develop an analytical framework to understand how orchestration functions in PSI across various parties and organizational levels (see Figure 1). At the same time, we consider how orchestration in the public sector may differ from its more frequently described application in the private sector.
The conceptual diagram is composed of three distinct groups of factors connected by lines. The left and right groups have the same text. These groups show three vertically arranged boxes, connected by vertical lines, labeled from top to bottom as “Environmental,” “Organizational,” and “Individual.” A curved line connects Environmental to Individual. In the middle, an oval contains two vertically stacked boxes, labeled “Relational” and “Innovation.” In between these boxes, the text “Orchestration” is written. Lines connect the central oval to each of the three boxes in the left and right groups.Analytical framework. Source: Authors’ own work
The conceptual diagram is composed of three distinct groups of factors connected by lines. The left and right groups have the same text. These groups show three vertically arranged boxes, connected by vertical lines, labeled from top to bottom as “Environmental,” “Organizational,” and “Individual.” A curved line connects Environmental to Individual. In the middle, an oval contains two vertically stacked boxes, labeled “Relational” and “Innovation.” In between these boxes, the text “Orchestration” is written. Lines connect the central oval to each of the three boxes in the left and right groups.Analytical framework. Source: Authors’ own work
Research design
The empirical part of this paper is based on a case study of a government-financed PSI program, Idéslussen, operated by the municipality of Sundsvall, Sweden. The program focused on digitalization and new digital solutions within the municipality. A case study design allows exploration of areas where research gaps have been identified, while also capturing how various elements interact (Halinen and Törnroos, 2005; Dul and Hak, 2008), a circumstance of fundamental importance when, for instance, studying orchestration of innovation. Researching a phenomenon which is practiced but only limitedly researched allows for theorizing efforts (Jensen and Rodgers, 2001; Woodside and Wilson, 2003).
The municipality Sundsvall is situated in northeastern Sweden, in a region historically shaped by paper and pulp production due to its extensive forestry resources. The area is currently undergoing restructuring in response to declining demand for products such as printed newspapers. The municipality has approximately 100,000 residents, with around 8,000 employed in the city’s administrative hub, spanning more than 300 professions. The municipal administration is organized into various service areas, including education, elderly care, housing, and public security, each tied to political decision-makers. Within this structure, Idéslussen aimed to foster and implement new ideas related to digitalization, making it a representative case of digital transformation in the municipal public sector. From a research perspective, the case is exploratory, given the limited documented knowledge of such efforts.
Data were gathered from three sources: individual interviews, focus groups, and documents. The focus groups were linked to four different sub-projects within Idéslussen, while the individual (one-to-one) interviews were conducted with members of the orchestration team. The four sub-projects were selected to represent diversity in units and operational contexts of the Idéslussen innovation process. They differed in field of interest, focus, and setting (e.g. healthcare and infrastructure), which enabled comparison and added dimensions, while they were similar in that each had completed the innovation process, being implemented or ready for implementation.
In total, thirteen individuals were interviewed, either through single-party interviews or as part of focus groups. Each interview lasted about one hour, followed a semi-structured protocol, and was recorded and transcribed. Questions addressed three main themes: (a) the innovation process—how it unfolded, who was involved, their roles, key learnings, and developments over time; (b) experiences and expectations related to the interviewee’s own role and the roles of others; and (c) digitalization as both a form of innovation and a toolkit for innovation, including its promises and challenges. For the orchestration team, questions additionally addressed coordination efforts and their perceived position within the broader municipal project.
The document sources included evaluation reports, newsletters, websites, project applications, directives, and self-evaluation reports, encompassing both official external communications and internal documentation. Using multiple types of sources enabled a deeper understanding of the processes studied and allowed for data comparison, thereby strengthening the study’s validity (Yin, 2007). Table 2 summarizes the data sources.
Data sources
| Type | Data source Informant/Sub-project | # of informants/documents | Field | Innovation |
|---|---|---|---|---|
| Interviews | Hub employee 1 | 1 female with 20 years’ experience of regional development and innovation. She has undergone management training | Municipality administration | Main – innovation orchestrator |
| Hub employee 2 | 1 female with more than 10-years’ experience of working with strategic development processes – several of which focusing on networked innovation between public and private sector, but also within private sector | Municipality administration | Innovation orchestrator | |
| Hub employee 3 | 1 female with more than 15 years of experience from working in strategic development processes and municipality company related projects, though not specifically on innovation | Municipality administration | Innovation orchestrator | |
| Focus groups | Sub-project #1 - The Dementia process | 3 participants: 1 female head of unit related to dementia – development and care, new on the job, but with 43 years of experience from working with dementia. 1 female head of operations (dementia) with extensive project experience. I man with 7 years of experience related to regional development and 10 years’ training as assisting nurse | Healthcare | Re-organization through digitalization |
| Sub-project #2 - The Remote reader | 3 participants: 1 female, responsible for coordinating information security. 1 female, project leader. 1 man IT manager at the municipality and in one of the municipality owned companies (long experience) | Infrastructure | Developing new technology (digital application) | |
| Sub-project #3 - The Culture change process | 2 participants: 1 female head of operations related to individual and family service with a PhD in social work and long-term experience of regional R&D processes. Research in social work in between for more than 20 years. 1 female with 20 years of experience of social work of which the last two was focused on making their organizations more employee-driven and inclusive in terms of changing the organizational culture | Social work | Re-organization through digitalization of processes | |
| Sub-project #4 - The Error reporting | 2 participants: 1 female acting as head of streets department, town planning office, traffic regulation, parking operations with extensive experience from public and private sector. 1 female working as project engineer in the street department predominantly focused on sustainable urban development | Infrastructure | Development of new administration system (digital tool for error reporting) | |
| Documents/web resources | Project application for the main project and for the collaborative research project | 1 + 1 | ||
| Ongoing part-time project reports | 5 reports between 2018–2019 | |||
| Communication | Newsletters with 900 prescriptions – also linked to the website | |||
| Website | Including project and case descriptions | |||
| Final project report (3-year project) Final report related to the collaborative research project Project evaluation report | 1 written by the orchestrating team to the grant-giving organization 1 written by the members of the research team and the authors of this article 1 made by a Swedish consultant firm based on ongoing evaluation procured by the grant-giving organization |
In the analysis procedure, interview transcripts and documents were coded to capture descriptions of the project, orchestration within it, and activities at the three levels of environment, organization, and individual, as outlined by Kusumasari et al. (2019). Materials from interviews, focus groups, and documents were initially organized according to the structure in Figure 1. Exploring a field with limited prior investigation meant that codes were empirically developed at each level of organizing to create understanding of Idéslussen, while following a structure as a means of organization (Crabtree et al., 1999). In line with the purpose of the paper, particular attention was paid to how orchestrators and public actors adopted innovation tools.
Subsequently, through iterative cycles of comparison and reduction, theorizing occurred as conceptualizations of PSI orchestration and its processes emerged (cf. Gioia et al., 2013). By comparing codes across data sources and, eventually, across sub-projects—and by addressing divergences, both discussed and practiced, compared to private sector approaches—aggregate dimensions (Gioia et al., 2013) were identified. These illustrate how private sector innovation tools were adapted and what consequences these adaptations had in the public sector setting. In the analysis section, these are described as vertical orchestration; orchestrated implementation; integrated development; non-collaborative integration across stakeholders; and the means-end logic. Together, these capture orchestration in the public sector and identify specific adjustments needed for PSI. In the final stage of analysis, the findings were compared with existing research on PSI and orchestration in order to clarify the theoretical contribution. Figure 2 depicts the coding structure.
A figure is divided into three vertical sections labeled from left to right as “1st order constructs,” “2nd order themes (Axial coding),” and “Aggregate dimensions.” The “1st order constructs,” and “2nd order themes (Axial coding),” have six rows, with rectangular text boxes in each column connected by arrows. The first three boxes in the “2nd order themes” are enclosed in a larger vertical box, and the last three boxes are enclosed in another large box arranged vertically below the first larger box. “Aggregate dimensions” has five vertically arranged boxes. The first box under “1st order constructs” has the following bullet points “Political pressure but not active guidance,” “Political initiation,” “Lack of commitment,” “Adopting ways of organizing from private sector,” “Funding provision,” “Providing access to tool.” A right-pointing arrow connects to the first box under “2nd order themes” labeled “Innovation orchestration: Environment.” Two individual rightward arrows point to the first and fifth boxes in the “Aggregate dimensions” labeled “Vertical orchestration: While the inclusion of top managers is important for P S I, these managers may also hinder innovativeness,” and “Means-end logic: While P S I aims to create value for the public as its outcome, the public culture orients the process to resources rather than value,” respectively. The second box under “1st order constructs” has the following bullet points: “Obtaining funding,” “Organizing with projects,” “Hiring orchestrators,” “Culture of risk aversion,” “Few projects becoming implemented,” “Bottom-up organizing,” “Linear approach adopted from private sector,” and “Handing over to unit without involvement of orchestrator.” A right-pointing arrow connects to the second box under “2nd order themes” labeled “Innovation orchestration: Organization.” Two individual rightward arrows point to the first and second boxes in the “Aggregate dimensions” labeled “Vertical orchestration: While the inclusion of top managers is important for P S I, these managers may also hinder innovativeness” and “Orchestrated implementation: For implementation to occur, orchestrators would need to be part of this phase, while such inclusion hinders them from operating beyond boundaries in P S I,” respectively. The third box under “1st order constructs” has the following bullet points: “Limited knowledge on innovation also affecting implementation,” “Better doing nothing than doing wrong attitude,” “Need for top management support,” and “Frustration with tools and method emerged over time.” A right-pointing arrow connects to the third box under “2nd order themes” labeled “Innovation orchestration: Individual.” Two individual rightward arrows point to the first and third boxes in the “Aggregate dimensions” labeled “Vertical orchestration: While the inclusion of top managers is important for PSI, these managers may also hinder innovativeness” and “Integrated development: Organizing away from the daily activities of units fostered inventions, while hindering the implementation of ideas, thereby simultaneously enabling and disabling P S I.” The fourth box under “1st order constructs” has the following bullet points: “Challenge to involve customers and other stakeholders in process,” “Need for commitment and formal authority affecting implementation,” and “Four-year cycle of elections affecting implementation.” A right-pointing arrow connects to the fourth box under “2nd order themes” labeled “Relational orchestration: Environment.” Two individual rightward arrows point to the fifth and sixth boxes in the “Aggregate dimensions” labeled “Non-collaborative integration across stakeholders: While PSI includes a great variety of stakeholders, the collaboration efforts are not at central stage for the process, but rather possibly become process outcomes” and “Means-end logic: While PSI aims to create value for the public as its outcome, the public culture orients the process to resources rather than value,” respectively. The fifth box under “1st order constructs” has the following bullet points: “Deciding when and how orchestrators should partake,” “Decision on collaboration across organizational borders,” “Orchestrator to facilitate collaboration,” “Innovation to happen outside the hierarchical organizations,” and “Collaboration limited as process based on culture of risk aversion.” A right-pointing arrow connects to the fifth box under “2nd order themes” labeled “Relational orchestration: Organization.” Two individual rightward arrows point to the second and fourth box in the “Aggregate dimensions” labeled “Orchestrated implementation: For implementation to occur, orchestrators would need to be part of this phase, while such inclusion hinders them from operating beyond boundaries in P S I” and “Non-collaborative integration across stakeholders: While P S I includes a great variety of stakeholders, the collaboration efforts are not at central stage for the process, but rather possibly become process outcomes.” The sixth box under “1st order constructs” has the following bullet points: “Need for top management support,” and “Focus on budgets rather than building values.” A right-pointing arrow connects to the sixth box under “2nd order themes” labeled “Relational orchestration: Individual.” A rightward arrow points to the third box in the “Integrated development: Organizing away from the daily activities of units fostered inventions, while hindering the implementation of ideas, thereby simultaneously enabling and disabling P S I”.Coding structure. Source: Authors’ own work
A figure is divided into three vertical sections labeled from left to right as “1st order constructs,” “2nd order themes (Axial coding),” and “Aggregate dimensions.” The “1st order constructs,” and “2nd order themes (Axial coding),” have six rows, with rectangular text boxes in each column connected by arrows. The first three boxes in the “2nd order themes” are enclosed in a larger vertical box, and the last three boxes are enclosed in another large box arranged vertically below the first larger box. “Aggregate dimensions” has five vertically arranged boxes. The first box under “1st order constructs” has the following bullet points “Political pressure but not active guidance,” “Political initiation,” “Lack of commitment,” “Adopting ways of organizing from private sector,” “Funding provision,” “Providing access to tool.” A right-pointing arrow connects to the first box under “2nd order themes” labeled “Innovation orchestration: Environment.” Two individual rightward arrows point to the first and fifth boxes in the “Aggregate dimensions” labeled “Vertical orchestration: While the inclusion of top managers is important for P S I, these managers may also hinder innovativeness,” and “Means-end logic: While P S I aims to create value for the public as its outcome, the public culture orients the process to resources rather than value,” respectively. The second box under “1st order constructs” has the following bullet points: “Obtaining funding,” “Organizing with projects,” “Hiring orchestrators,” “Culture of risk aversion,” “Few projects becoming implemented,” “Bottom-up organizing,” “Linear approach adopted from private sector,” and “Handing over to unit without involvement of orchestrator.” A right-pointing arrow connects to the second box under “2nd order themes” labeled “Innovation orchestration: Organization.” Two individual rightward arrows point to the first and second boxes in the “Aggregate dimensions” labeled “Vertical orchestration: While the inclusion of top managers is important for P S I, these managers may also hinder innovativeness” and “Orchestrated implementation: For implementation to occur, orchestrators would need to be part of this phase, while such inclusion hinders them from operating beyond boundaries in P S I,” respectively. The third box under “1st order constructs” has the following bullet points: “Limited knowledge on innovation also affecting implementation,” “Better doing nothing than doing wrong attitude,” “Need for top management support,” and “Frustration with tools and method emerged over time.” A right-pointing arrow connects to the third box under “2nd order themes” labeled “Innovation orchestration: Individual.” Two individual rightward arrows point to the first and third boxes in the “Aggregate dimensions” labeled “Vertical orchestration: While the inclusion of top managers is important for PSI, these managers may also hinder innovativeness” and “Integrated development: Organizing away from the daily activities of units fostered inventions, while hindering the implementation of ideas, thereby simultaneously enabling and disabling P S I.” The fourth box under “1st order constructs” has the following bullet points: “Challenge to involve customers and other stakeholders in process,” “Need for commitment and formal authority affecting implementation,” and “Four-year cycle of elections affecting implementation.” A right-pointing arrow connects to the fourth box under “2nd order themes” labeled “Relational orchestration: Environment.” Two individual rightward arrows point to the fifth and sixth boxes in the “Aggregate dimensions” labeled “Non-collaborative integration across stakeholders: While PSI includes a great variety of stakeholders, the collaboration efforts are not at central stage for the process, but rather possibly become process outcomes” and “Means-end logic: While PSI aims to create value for the public as its outcome, the public culture orients the process to resources rather than value,” respectively. The fifth box under “1st order constructs” has the following bullet points: “Deciding when and how orchestrators should partake,” “Decision on collaboration across organizational borders,” “Orchestrator to facilitate collaboration,” “Innovation to happen outside the hierarchical organizations,” and “Collaboration limited as process based on culture of risk aversion.” A right-pointing arrow connects to the fifth box under “2nd order themes” labeled “Relational orchestration: Organization.” Two individual rightward arrows point to the second and fourth box in the “Aggregate dimensions” labeled “Orchestrated implementation: For implementation to occur, orchestrators would need to be part of this phase, while such inclusion hinders them from operating beyond boundaries in P S I” and “Non-collaborative integration across stakeholders: While P S I includes a great variety of stakeholders, the collaboration efforts are not at central stage for the process, but rather possibly become process outcomes.” The sixth box under “1st order constructs” has the following bullet points: “Need for top management support,” and “Focus on budgets rather than building values.” A right-pointing arrow connects to the sixth box under “2nd order themes” labeled “Relational orchestration: Individual.” A rightward arrow points to the third box in the “Integrated development: Organizing away from the daily activities of units fostered inventions, while hindering the implementation of ideas, thereby simultaneously enabling and disabling P S I”.Coding structure. Source: Authors’ own work
Findings
Setting the scene: starting off the project
The Idéslussen program was initiated partly to respond to political and economic pressures to cut costs and deliver more efficient services, and partly in response to a call for participation in a 3.5-year government-funded pilot project on PSI. The external pressure to increase municipal efficiency related to competitive challenges experienced by many municipal institutions from private sector operators in areas such as education and child and elderly care. Co-funding was a prerequisite in the program call, so the region, other regional municipalities, the Swedish Municipalities and Regions Association (SKR), BizMaker (an innovation hub), and Bron (a digital innovation hub and regional strategic network) were included in the application. The pilot call came during the refugee crisis in 2015, when Sundsvall municipality sought new ways to work with refugees under enormous pressure. The program focused on new digital solutions, work methods, tools and applications. The head of the municipality presented the call to the municipal board, and politicians approved the application. With this backing and support, the program was positioned as a flagship project.
Following successful funding, the Idéslussen program received a generous budget covering wages for three orchestrators and additional external experts, as well as PSI-related activities such as seminars, workshops, communication, and training programs for participants. Furthermore, the program had access to all municipality-related tools, training programs, and support systems, as well as the services and products delivered by municipal units, which in turn served as the program’s point of departure. The program included three innovation programs with 90 participants from different municipal units and operations. These programs focused on principles and processes, while all members also participated in sub-projects. Activities primarily took place within municipal premises, though there was a budget for external meetings and organizing webinars.
Between 2017 and 2020, the Idéslussen program generated and evaluated 169 cases, of which 142 were finalized and 29 were implemented. Its innovation orchestration was based on an international standard for innovation management, which drew heavily on American management literature, with particular emphasis on frameworks associated with Harvard-based research.
Organizing the project: PSI orchestration
The municipality recognized that working across organizational boundaries was important but did not know how to achieve this in practice. Its experience with systematic innovation was also limited: “The word innovation was known, but they did not fully understand what it meant beyond tools and technical stuff” (Project leader), or, as expressed by a participant in Sub-project 1: “Who knows how to do it? Employees don’t do that.” The funding institution, as well as a hired expert consultant, made efforts to emulate managerial practices and innovations from the private sector within the program.
The decision to recruit external parties to orchestrate the innovations stemmed from this lack of knowledge and from the perceived importance of cross-boundary organizing. Three orchestrators were recruited to the program. The main project leader was certified as a PSI orchestrating intermediary by the governmental research institution Research Institutes of Sweden (RISE), having undergone a specialized training program provided by this financier. The other orchestrators had participated in related courses and lectures and felt well-prepared for their roles. They facilitated a program that served as a shared platform for cross-boundary collaboration, where value creation and innovation activities could be co-developed. This approach emphasized innovation processes occurring outside the traditional hierarchical structures that typically shape day-to-day operations within the municipality.
The orchestrators aimed to continuously adapt their actions in response to emerging needs, guiding the sub-projects toward long-term sustainability while addressing the diverse values and requirements arising throughout the innovation process. Intra- and interorganizational interactions were implied, with outcomes including new ways of cooperation throughout the municipality and the development of digital devices or services as innovation outputs. This approach departed from the notion that innovations are constrained by day-to-day operations, allowing for a separate unit in which cross-team interaction could occur. The Idéslussen program thus operated outside the organizational structure as a standalone initiative, yet collaborated with municipal institutions through employees’ renewal ideas, offering a “free zone” for development: “As of today, [the public sector] is so inflexible and limited in our tasks, and laws and budget never give you this test zone, free zone … That is what we’re trying to create” (Idéslussen member).
The program adopted a bottom-up design to spur ideas among employees. Early on, communication structures were established, including a dedicated website to promote the program and allow employees to submit their digitalization ideas, as well as a newsletter open to all employees and external subscribers. A municipal communicator provided ongoing communication support throughout the program. Employees were invited to contribute ideas, which they then championed within task forces composed of colleagues and other key stakeholders. The program and its sub-projects collaborated daily with various municipal units and external firms, and this collaboration was identified as a vital determinant of innovation. As one participant from Sub-project 3 noted: “We were supported related to ideas and input from others within the innovation program.”
Involvement of end-users and consideration of all stakeholders’ views were regarded as important for performance but challenging to manage. Most participants came from middle management and saw the program as a means to develop their own knowledge and careers: “Many of the middle managers participated. Of 300 champions, 150 were middle managers using participation as a development tool” (Main orchestrator). Participants also emphasized personal growth associated with involvement in innovation work. A member from Sub-project 4 observed: “It was great personal development to be involved in development projects.” Another added: “It’s incredibly rewarding to immerse yourself in your role—even when the workload increases—and to gain new knowledge.”
Identified teams had access to information and communication technology and mediating tools provided by the funding and supporting organizations for use in PSI sub-projects. However, the orchestrators increasingly felt frustrated with the tools and methods available, which seemed ill-suited to the public sector context. Despite this, the program was designed to adapt to these approaches, even though they had been imported from the private sector rather than specifically designed for PSI.
Sub-projects: operationalizing the program
At the start of the program, the orchestrators facilitated mini-seminars, workshops, workplace meetings, breakfasts, and information sessions, through which they identified the sub-projects. Knowing how to invite individuals to the program and making them feel valued was perceived as essential, as was identifying others with key competencies to contribute. Sub-project 3, for example, involved a collaborative setting with participants from two external organizations who engaged to provide support. As one project member explained: “We focused on what we wanted to learn and, together with them, agreed on the learning goals for the case, experimented, and then continued developing the work within our own organization” (Project Member, Sub-project 3).
The orchestrators organized the sub-projects according to a three-phase process, a linear approach adopted from the private sector and innovation management literature, which was adapted along the way. The first phase consisted of problem identification, including identifying needs, potential value added, beneficiaries, and the interests of the champions. Additional team members, resources, and competencies were identified at this stage. The second phase focused on exploration, targeting problem-related knowledge development through benchmarking, observations, seminars, and interviews facilitated collaboratively by orchestrators and champions. During this phase, further tools were identified to meet the sub-projects’ specific requirements. The third phase, experimentation and testing, led to implementation, when the new solution was handed over to the municipality unit that would use the innovation. Throughout these phases, the orchestrators acted to varying extents as a business unit, providing ongoing evaluation, feedback, design-related services, training, expert support, and an innovation-facilitating setting for ideas and champions to develop collaboratively.
Sub-projects also included user participation in interviews, testing, and feedback on new solutions. Before innovation transfer for implementation, the orchestrators facilitated relationship-building activities to support PSI, fostering both social relationships among municipal actors and business relationships with external actors with the potential to contribute.
While each sub-project faced unique challenges, some factors were consistently central to PSI success. Resources and funding were critical, as expressed by one participant: “We could never have done it [the innovation process] without Idéslussen facilitating competencies and resources”, or by a participant in Sub-project 1: “Project grants may be needed to move from the current state to a new situation.” Participants in Sub-project 2 emphasized the importance of external support: “It went very smoothly, and the reason for that was that Idéslussen stepped in and helped us”, and “It’s great when external expertise is available to help narrow things down, and once we had that in place, we could continue working independently.” Despite acknowledging the orchestrators’ role, participants remained highly focused on budgets.
However, Idéslussen limited its involvement once an idea was formed and evaluated, before it was transferred to the municipal unit responsible for implementation. One orchestrator emphasized: “It is still the organization of the sub-project that has a responsibility to implement things into their organization and activities; it is not us.” This boundary created concerns during implementation. Leadership and top management commitment emerged as critical factors for successful implementation but were sometimes perceived as barriers. Participants in Sub-project 2 noted: “Leadership is very important—there needs to be an interest in working with innovation, which isn’t very strong in our organization”, and “If we hadn’t had support, we wouldn’t have been able to do this.” Challenges faced by middle management were also highlighted: “Middle managers have the hardest time, facing pressure from both directions” (Participant, Sub-project 3). Sub-project 4 participants similarly emphasized the importance of informed leadership: “It’s extremely important to have a management team that is informed.”
Even with employee-driven sub-projects, a lack of top management support, encouragement, and norms favoring exploration often caused innovation processes to stall or end before or during implementation. The project leader noted the difficulty for units to manage development work alongside normal operations. One participant stressed: “All commitment emerged from leadership […]. If you have leadership without commitment, it is difficult to push forward.” Despite aiming for a bottom-up orientation, sub-project participants expressed concerns about top-down decision-making in the public sector. Idéslussen attempted to bypass this through its bottom-up design but recognized its limitations: “What hinders implementation is the way Idéslussen is rigged through this employee perspective: ‘Come with your idea as an employee; it does not matter whether your manager thinks it’s a good idea or not.’” Participants in Sub-project 1 emphasized the importance of formal authority and support from higher organizational levels: “We must have a decision and a mandate to work on the issues” and “There is no lack of commitment or ideas, but it’s those above us who make the decisions.”
The program supported employee-driven sub-projects, but many were ultimately terminated by top managers during the implementation stage, resulting in wasted time and resources. One participant noted: “We live in four-year cycles in public administration. You work out good ideas, and then—after elections—you have to go back to doing what you’ve always done.” The orchestrators summarized this as: “We can only lead them to the well; they have to want to drink themselves.” Despite these terminations, the sub-projects contributed to a more innovation-oriented culture, as many participating employees gained insight into innovation, related tools, processes, and PSI prerequisites: “The perception of the concept is broader now and linked to ways of working, processes communicated in terms of simplification, renewal, improvement, and change to make it understandable” (Project leader).
Sub-projects that successfully reached implementation were those in which top management of the receiving units was change-oriented and willing to take risks. These sub-projects received autonomy and opportunities to experiment within certain boundaries, expanding employees’ culture and skills in innovation while providing incentives to innovate. A sub-project participant emphasized: “[What motivated me] was that I had participated in the training and wanted to do a test. Try it out and see if we could get this innovation process into our organization.” However, few managers were willing to assume the associated risks.
Recognizing the importance of top management commitment and risk tolerance, orchestrators noted the need to engage managers in dialog to enable innovation, appealing to them rather than bypassing them. Orchestrators also emphasized the importance of flexible guidance, “humble guidance, making people do things without too much control”, being able to manage multiple tasks simultaneously, and understanding when to intervene or step back. Early efforts focused on engaging participants, generating energy, showing opportunities, providing support, and planning with options for the champions.
Participants confirmed the positive perception of the orchestration team. A Sub-project 2 participant noted: “They contributed a lot of positive energy from the beginning. It felt safe to have them there”, and “They stayed present throughout, were supportive, and offered help during meetings with decision-makers.” Others added: “They were open—they didn’t try to steer us in any particular direction”, and “We would have liked to have them with us the whole way, if that had been possible.” A participant from Sub-project 3 remarked: “They also asked the ‘stupid questions’ – but it’s different when they ask them. It opens up the conversation and makes it easier to engage the managers involved” and “They compiled, clustered, analyzed, and created polished slides that we could bring back to our employees.”
Participants also emphasized understanding municipal logic as key to effective orchestration. As described by Sub-project 4 participants, the orchestrator was someone who “leads the process”, “ensures momentum”, and “tests, monitors the flow, and brings everything back together without long investigations.” Later, the role expanded to provoking, breaking norms, rearranging, addressing new challenges, and adapting to different contexts.
Analysis
The Idéslussen program thus illustrates how a municipality invested in a PSI initiative focused on digitalization. This was achieved by adopting methods and processes from private sector innovation, including the use of orchestrators to facilitate innovation activities organized separately from the municipality’s day-to-day operations. Orchestrating innovation is a well-established method in the private sector (Andresen, 2020; Andersen and Ljungkvist, 2021), popularized through open innovation (e.g. Natti et al., 2014; Giudici et al., 2018; Helfat and Raubitschek, 2018; Reypens et al., 2019). Orchestrators typically facilitate two main roles: organizing innovation and organizing collaborations (Mumford et al., 2002; Nambisan and Sawhney, 2011). While orchestration has been popularized in the private sector and described as a successful way of organizing innovation (Wareham et al., 2014), the Idéslussen case reveals key differences when applying this method in the public sector.
More precisely, captured as the aggregate dimensions of our analysis (cf. Gioia et al., 2013), the following key deviations from private sector orchestration emerged, highlighting areas of misalignment between established private sector orchestration and the realities of the public sector context: vertical orchestration; orchestrated implementation; integrated development; non-collaborative integration across stakeholders; and the means-end logic. Each is below formulated as a propositional paradox, illustrating how these dimensions not only diverge from the private sector but also generate tensions inherent to PSI.
To elaborate on the vertical orchestration, orchestration is mostly known from the private sector to facilitate innovation in the intersection of parties collaborating on horizontal levels within or among firms. This refers to how parties that partake in exchanges or individuals as part of a specific organizational level work across to come up with new solutions. The findings related to Idéslussen indicate how orchestration would need to entail the coordination among organizational levels. This in turn reaches beyond Kusumasari et al. (2019)’s description of the three levels of PSI enablers as these levels, including the environment, the organization and the individual, only answer up to the resource provision, in the Idéslussen program described in terms of the budget and provision of tools and methods. To actually produce PSI, the coordination would mean that top managers are on the same boat as the employees, are risk-taking and change oriented. This again requires coordination between employees and top managers that comes forth strongly in the case of Idéslussen. Not only was the program oriented to bottom-up innovations, but it actually to a large extent attracted middle managers. Still, the sub-projects often failed in the implementation phase. In a context characterized by risk aversion and a public sector culture where there is a strong fear of making mistakes, the role of top managers play is critical in supporting and legitimizing employee-led innovation. However, with these top managers being as embedded in cultures of risk aversion, they could well have a negative impact on the innovations, leading to the first proposition:
While the inclusion of top managers is important for PSI, these managers may also hinder innovativeness.
Vertical orchestration therefore differs from private sector practices by emphasizing cross-level coordination, which includes shaping mindsets toward risk-taking. Even bottom-up initiatives are influenced by top-down structures, including political oversight and budget constraints.
As the second aggregate dimension—referred to as orchestrated implementation—reveals, the orchestrators were not involved in the implementation phase, having stepped away beforehand. This often led to implementation not happening. It relates to lacking competence as well as to fears of leading change among participants. Hence, there was a continuous need for the orchestrators to be parties willing to carry the risk, also seen in quotes above referring to resources as prerequisites for the innovations and thereby how others needed to carry responsibilities for the innovations. Meanwhile, orchestrators should normally not be those inventing but act their sovereign role as intermediaries to be able to, outside boundaries, facilitate relation and innovation coordination. This leads to a second proposition:
For implementation to occur, orchestrators would need to be part of this phase, while such inclusion hinders them from operating beyond boundaries in PSI.
As our third, related, aggregate dimension —integrated development—reveals, the sub-projects were organized as separate, standalone units, disconnected from the everyday operations of the departments expected to implement the innovations. This structural separation contributed to a sense of detachment, reducing the willingness or capacity of operational units to adopt innovations developed outside their immediate context. With that said, organizing away from the daily routines was described as a prerequisite for any innovation to take place, according to the interviewees. Moreover, the reluctance to implement innovations—often described by the ‘not invented here’ principle—was not solely due to the fact that these innovations were developed elsewhere. Instead, it was thus deeply rooted in the public sector’s general aversion to risk-taking (cf. Potts and Kastelle, 2010; Agolla and Van Lill, 2016; Palm, 2020). This leads to a third proposition:
Organizing away from the daily activities of units fostered inventions, while hindering the implementation of ideas, thereby simultaneously enabling and disabling PSI.
The orchestrators have thus previously been described as relational and innovation operandi, or the organizing of collaboration and of innovations (Nambisan and Sawhney, 2011). These operandi would not be unconnected but affect one another. Collaborations would foster further innovations but potentially make them increasingly streamlined. Meanwhile, innovation may lead to collaboration. The public sector, being organized in a way that includes both political and service providing organizations, seems to suggest that the collaborations are partly put out of place. Although the Idéslussen program entailed collaborations among units, these were strongly oriented towards the innovations and would not last overtime. Within the units, the hierarchy of top management and employees suggested quite lengthy distances, not to mention to the politicians. Thus, politicians may enable innovations, but they would not take part in the innovation process (cf. Kusumasari et al., 2019), nor intend to create collaborations with innovating employees or the various municipality units of implementation. This leads to a fourth proposition:
While PSI includes a great variety of stakeholders, the collaboration efforts are not at central stage for the process, but rather possibly become process outcomes.
Continuing with the influence of the public sector, the delicate issue with money being the means to achieve processes, products, services or quality improvements to benefit society to some part (West and Farr, 1990; Wihlman et al., 2016) thus creates the reverse logic to the private sector’s goal of financial return (Öberg, 2020). Again, this is nested into budgets as a key tool of operations (Stewart-Weeks and Kastelle, 2015; Kusumasari et al., 2019) and practiced in the form of not allowing spending excessive to budgets. This has previously been referred to as leading to a reactive, conservative, risk-reluctant public sector (Palm, 2020; Potts and Kastelle, 2010), but also needs to be considered in the evaluation of innovation and what they should accomplish. Top managers focusing on budgets, and albeit the Idéslussen program entailing quite an extensive budget, meant an orientation to costs rather than benefits, as did the use of tools from the private sector. This means-end difference between the private and the public would hence be important to consider in any PSI process, leading to the last proposition:
While PSI aims to create value for the public as its outcome, the public culture orients the process to resources rather than value.
This proposition also highlights participants’ continuous requests for additional resources, often using resource limitations as justification for not innovating.
Conclusions
This paper described and discussed how innovation orchestration works in PSI settings. The underlying objective was to examine whether and how the private sector methods create a fit or misfit with the public sector and what adjustments need to be made in PSI processes. The case of Idéslussen illustrates that key definitional characteristics distinguish the public sector from the private sector in ways that directly influence innovation processes. These differences are evident not only in how innovation is approached, but also in the nature of the outcomes.
More specifically, any PSI process requires coordination among vertically organized actors—from politicians to employees—extending beyond each level while ensuring enabling traits at every stage (cf. Kusumasari et al., 2019). Orchestration that is both horizontally and vertically oriented is therefore crucial. Resources play only an enabling role, while the specific characteristics of the public sector necessitate a reconsideration of the means and ends of innovation (cf. Öberg, 2020). This underscores the limitations of directly applying private sector models to PSI, while highlighting how budgetary constraints influence innovation orientation and risk-taking behavior.
Furthermore, the hierarchical structure of PSI settings, combined with political governance, laws, regulations, and the absence of traditional market incentives, does not provide the same motivators or legitimacy for innovation as the private sector. Innovation may even be perceived as threatening due to potential loss of status, power, or identity for individuals and departments affected by change (Stensaker et al., 2020). In addition, PSI participants may fear an increased workload. Given these comparatively weak incentives at both individual and organizational levels, orchestrators play an even more critical and challenging role in PSI than in private sector innovation.
Theoretical contributions
This paper provides new insights into PSI, specifically regarding the orchestration of innovation processes where digitalization and digital tools shape outcomes. By juxtaposing the dimensions of innovation and the relational operandi of orchestration with the public sector’s means-goal distinction from the private sector, the paper suggests that relational aspects of orchestration may emerge as outcomes of successful innovation processes rather than serving as their starting point, as is typically the case in the private sector. The emphasis on vertical coordination further highlights the role of means, illustrating that innovations are not only intended to improve service quality for citizens but must also engage with political governance.
The paper contributes to existing research by theorizing the conditions of PSI, including the need for vertical coordination across enabling levels, and by showing how risk aversion necessitates top-level commitment, since digital innovations may be perceived as threatening to individuals rather than purely as tools for improving services. The five propositions developed in the paper highlight how PSI orchestration involves inherent tensions that shape both processes and outcomes, revealing the ways in which private sector logic can produce contradictions when applied to PSI.
Managerial implications
The empirical findings of this paper show that innovation processes in the public sector differ substantially from those in private firms. Hierarchical structures, limited knowledge and understanding of innovation processes, and the absence of clear mandates or top management support can all act as barriers to PSI. A first recommendation is therefore to critically evaluate any adoption of private sector methods or tools for PSI. Key questions to address include: What is needed to create better-quality services for citizens, and how is such value generated? How can actors across organizational levels, including political representatives, collaborate to create this value? How can risk aversion be managed in relation to innovations? And how can separate innovation processes be better aligned with the units responsible for implementation?
To address the observed differences from the private sector and the implementation challenges identified in the case, one potential solution is to integrate political and top management representatives into the orchestration unit. This form of top-level intermediation could help counteract risk-averse organizational cultures, where inaction is often favored over the possibility of failure. Additionally, financial resources could be allocated through stepwise approval processes, with evaluations at each stage assessing projected benefits relative to the investment. Ideally, such evaluations would involve citizen representatives to ensure democratic legitimacy and relevance. Finally, the inclusion of innovation coaches to support employees and middle managers could foster creativity, encourage risk-taking, and generate more effective innovations, while also challenging the prevailing culture of avoiding mistakes through inaction.
Limitations and future research
This paper is based on a single case study, albeit including several sub-projects. The case study design helps to explore and connect various aspects of the phenomenon (Halinen and Törnroos, 2005; Dul and Hak, 2008) but the limitations of a single case are evident. It could be argued that the findings are specific to this case and may not be directly generalizable to other PSI programs or settings.
However, to counter this concern, our findings align well with previous descriptions of public sector characteristics, even though such research has not specifically focused on innovation or orchestration. To strengthen the validity of our findings and expand upon them, further research is though needed. Future studies could include additional cases, quantitative evaluations, or comparative analyses across geographical regions and non-digital contexts to address potential context-specific limitations. It would also be valuable to compare municipalities—or even states—that have advanced in open innovation practices, as well as organizations that use orchestrators versus those that do not for innovation and relational goals, as well as to elaborate on the role of employees’ experience with innovation, the public sector, and orchestration in the failure or success of PSI processes.
More broadly, we hope this paper stimulates further interest in research on PSI and encourages continued exploration of this important area.
