A practice approach to fostering employee engagement in innovation initiatives in public service organisations

Abstract This article adds to the public sector innovation literature by capturing the practices which act to enable or restrict employee engagement in innovation initiatives. Drawing on six cases of public service organizations in Scotland, UK, participating in an intrapreneurship programme, we identity two types of practice which formalize means of organizing resources (structural practices) and promote a positive attitude towards innovation (embedding practices). Our findings also highlight that an organizations’ strategic motives for participating in initiatives influences the strength of these engagement practices. The implications are discussed in relation to future practice-orientated approaches to public sector innovation research.


Introduction
There is a growing body of literature detailing the drivers and barriers to innovation in public sector settings (Cinar, Trott, and Simms 2019;De Vries, Bekkers, and Tummers 2016;Mergel 2018;Torfing 2019).This work has been linked to New Public Management (NPM) reforms driving the adoption of private sector practices to improve public sector performance (Osbourne 2006).While there is a growing perception that innovation can increase productivity and improve public sector services, there is considerable debate on how best to achieve heightened innovation (Clausen, Demircioglu, and Alsos 2020;Demircioglu and Audretsch 2017;Hartley, Sørensen, and Torfing 2013).
Existing work has begun to explore various initiatives designed to engage wider stakeholders in the innovation process, such as training (Brogaard 2017), innovation labs (Tõnurist, Kattel, and Lember 2017), and workplace redesigns (Lindsay et al. 2018).However, the existing empirical evidence highlights variance in the success of these innovation initiatives (Birdi 2021;Borins 2001;Dudau, Kominis, and Szocs 2018;Schultz, Sjøvold, and André 2017).This success has been associated with a multitude of organizational barriers that need to be overcome in order for innovation to be implemented (Cinar, Trott, and Simms 2019;De Vries, Bekkers, and Tummers 2016;Mergel 2018;Tremml 2021).However, one particular barrier which is undertheorized in this research is employee engagement (Yuriev, Boiral, and Talbot 2021).Existing evidence highlights that middle-managers and front-line staff drive innovation within government (Hartley 2005) This, therefore, is a particularly pertinent barrier to explore as engaged employees are more likely to display innovative behaviour which can lead to greater productivity and public service value (Fernandez and Moldogaziev 2013;Hameduddin and Fernandez 2019;Kwon and Kim 2020).Adversely, low levels of physical and cognitive engagement have been found to be detrimental to the implementation of innovation (Christian, Garza, and Slaughter 2011;Shuck et al. 2013).
Existing research typically focuses on measuring the impact of overall employee engagement against various innovation outcomes (Cinar, Trott, and Simms 2019;Tremml 2021;Yuriev, Boiral, and Talbot 2021).As such, this research fails to connect with the practice of 'doing engagement' (Fletcher et al. 2020).Extending theoretical insights into how employee engagement with a specific innovation initiative can be fostered, therefore, can extend current literature debates which conceptualize innovation engagement barriers as static and holistic (Cinar, Trott, and Simms 2019;Yuriev, Boiral, and Talbot 2021).Understanding and detailing various barriers to innovation as social practices can offer a more dynamic understanding of what fosters public innovation (Arundel, Casali, and Hollanders 2015;Cinar, Trott, and Simms 2019;Torugsa and Arundel 2016).
In this paper, we adopt a practice lens (e.g.Nicolini and Monteiro 2017), which allows us to examine the dynamic and specific drivers of engaging employees in an intrapreneurship training programme (ITP).The overall aim of this article is to introduce a practice approach to understanding barriers and drivers of innovation engagement within public service organizations.To do this, we use a theory-building multiple case study approach which allows the reasons for the relative success and failure of six public service delivery organizations in Scotland, UK, to emerge (Eisenhardt 1989;Stewart 2012).From our analysis we introduce a framework and several propositions outlining the practices that foster engagement with innovation initiatives.
This paper makes three contributions to existing research on public innovation.First, we put forward a practice perspective to employee engagement which acknowledges the construct as complex, task-specific and dynamic.Second, we position innovation barriers and drivers not as static entities that need to be overcome, but enduring practices that structure organization processes.Finally, we demonstrate the importance of understanding the strategic motives for why public service organizations engage with different innovation initiatives.
The remainder of this article begins by reviewing the current literature on employee engagement in innovation initiatives and outlining the articles theoretical background.Next, the multiple case study methodology is described, followed by a presentation of the results.The article concludes with a discussion of the theoretical framework, propositions, and contributions to theory and practice.

Employee engagement in innovation initiatives
Within the extant public sector literature, Kahn's (1990) concept of engagement is widely adopted (e.g.Hameduddin and Lee 2021;Jin and McDonald 2017).This perspective highlights that when people engage, they 'employ and express themselves physically, cognitively and emotionally during role performances' (Kahn 1990, 694).Previous research has mainly focused on uncovering drivers and barriers, as well as determining the outcomes of increasing employee engagement (Fletcher et al. 2020).
At an individual level, engagement is thought to be driven by public sector workers' motives for public service and sense of community responsibility (Boyd and Nowell 2020;Vinarski Peretz 2020).Other studies indicate that specific job factors, such as opportunities for learning and development (e.g.Bakker and Bal 2010) or supervisor support (e.g.Jin and McDonald 2017), have also improved employee engagement at work.At the organizational level, the key role leaders and managers have in supporting engagement is well acknowledged, with supportive styles collectively believed to be most effective at fostering engagement (Ancarani et al. 2021).Other studies focus on external factors, such as the impact of cutbacks, austerity and innovation projects on employee engagement (Boumans et al. 2015;Kiefer et al. 2015).
The barriers and drivers to different types of innovation within public sector contexts are also well acknowledged (Cinar, Trott, and Simms 2019;De Vries, Bekkers, and Tummers 2016).Various incentives such as innovation awards (e.g.Bernier, Hafsi, andDeschamps 2015), innovation labs (e.g. McGann, Wells, andBlomkamp 2021;Tõnurist, Kattel, and Lember 2017), and civic hackathons (e.g.Yuan and Gasco-Hernandez 2021) have all been found to improve innovation processes in public sector contexts by engaging wider stakeholders.Training programmes have also been found to foster innovation (e.g.Brogaard 2017).Schultz, Sjøvold, and André (2017), for example, highlight in healthcare settings how formal training increased participants understanding of innovation strategy which led to the development of new ideas.
However, there is some evidence that innovation initiatives are not always effective in engaging stakeholders (Dudau, Kominis, and Szocs 2018).Birdi's (2021) evaluation of one programme delivered across 18 separate public service delivery organizations reported mixed results.The training programmes was well received by some which led to significant gains in innovation competencies.For other organizations they were unable to apply any learning from the training, representing a divergence in organizational experiences.
Within extant literature there is no clear indication as to what drives such divergence in organizations experience with innovation initiatives (Dudau, Kominis, and Szocs 2018;Fernandez and Moldogaziev 2013).Several studies, however, hint at numerous facilitating factors that can influence an employee's engagement in the innovation process.Piening (2011) acknowledges the important role managers have by encouraging the performance of either routine operational behaviours or a focus on innovation activities.Likewise, several studies show how leaders can encourage innovation through providing clear strategic goals and adequate incentives (e.g.Ancarani et al. 2021;Ravishankar 2013).
A key shortcoming with this body of knowledge is the lack of research which connects with the practice of 'doing engagement' (Fletcher et al. 2020).This can be attributable to two main factors.First, existing scholarship focuses on exploring the outcomes and not on the interrelations between various drivers of employee engagement and innovation (Rangarajan 2008;Yuriev, Boiral, and Talbot 2021).Such perspectives view both the barriers to and outcomes of innovation engagement as static and as either present or not present within an organization (Cinar, Trott, and Simms 2019;Tremml 2021).Second, engagement is captured by focusing on general worker job roles and not on specific parts or elements of a role.Considering that workers have multiple roles, responsibilities and carry out many tasks it is possible that they engage and disengage with different aspects of their work (Saks and Gruman 2014;Schaufeli and Salanova 2011).
By neglecting to explore employee engagement in specific circumstance, such as within innovation initiatives, existing research fails to capture any nuances that exist in fostering innovation engagement.To capture these nuances more in-depth qualitative research is called for to address the limitations of quantitative approaches that focus on employee engagement outcomes (Fletcher et al. 2020;Noesgaard and Hansen 2018).To facilitate this, richer conceptual understanding of the organizational practices that influence engagement in innovation is needed (Fernandez and Moldogaziev 2013;Vinarski Peretz 2020;Yuriev, Boiral, and Talbot 2021).To address this, we seek to address: What specific organizational practices influence employee engagement?How do public service delivery organizations enact various practices to foster employee engagement in innovation initiatives?

A practice approach to engaging employees in innovation initiatives
To develop novel insights into our research questions we adopt a practice lens which allows us to examine the dynamic and specific drivers of employee engagement.While there is no unified practice theory (Nicolini 2012), practice-based approaches are foreground in social practices.Social practices are defined as 'routinized regimes of materially mediated doings, sayings, knowing, and ways of relating that form the building blocks for understanding organizational phenomena' (Nicolini and Korica 2021, 5).However, practices are not discrete entities, but bundles of activities that encapsulate both the activities of actors and the enduring practices that structure organization processes (Nicolini and Monteiro 2017;Schatzki 2002).
Practice-orientated scholars generally highlight three elements of practice-based approaches -the 'what', 'who' and 'how' (Jarzabkowski et al. 2016).The what refers to the practices that are taken-up by organizations and the who refers to the practitioners who perform the practices.The how refers to the actual means in which practices are enacted in (i.e.praxis), which acts to mediate the dynamics between practice and any outcomes (Jarzabkowski et al. 2016).The how stresses the importance of context, highlighting that practices are entrenched within the organizations they inhabit (Feldman and Orlikowski 2011;Nicolini and Monteiro 2017;Schatzki 2005).Practices are informed by social rules which are explicit prescriptions or instructions of what needs to be done and are shaped by hierarchical orderings of projects, tasks and emotions (Cucchi et al. 2022).This emphasizes the importance of activity, performance, and work as ongoing recurrent processes (Nicolini 2012).
The practice approach has been applied sparingly within the public sector context to understand the 'doing' of public services, such as creating public value (Ferry, Ahrens, and Khalifa 2019;Huijbregts, George, and Bekkers 2021).In our case, we are interested in the 'doing' behind engaging employees in innovation activities.In existing practice research in public sector contexts (George and Desmidt 2014;Hansen 2011;Huijbregts, George, and Bekkers 2021), the focus is generally on understanding the 'what', 'who' and 'how'.To analyse employee engagement in innovation initiatives, therefore, requires examination of: (a) practices, the various shared routines and behaviours within an organization that influence employee engagement; (b) practitioners, those involved in engaging employees with innovation initiatives; and (c) praxis, the activities carried out to identify, reflect and decide on actions that are needed to engage employees.
In adopting this conceptual approach, two main assumptions are made about how employees' engagement in innovation initiatives is fostered.First, barriers and drivers to engagement in innovation initiatives are viewed not as something that exists, but as something that organization members (practitioners) do (practices) (Nicolini and Korica 2021).Rather than the availability of training directly supporting employee engagement in innovation, for example, it is the activities to encourage employees to engage in training (practices) and the positioning of the training in relation to other organization practices (praxis) that determine whether employees engage with innovation.Second, a practice approach considers fostering engagement as related to 'social effects generated by a practice in connection with other practices' (Corradi, Gherardi, and Verzelloni 2010, 277).Focusing on the nexus of activities that foster engagement, therefore, should directly address how various barriers and drivers interrelate, a key shortcoming in the current literature (Cinar, Trott, and Simms 2019;Tremml 2021).

Research setting
The setting for this study is Scotland, a devolved administration of the UK government, where bottom-up approaches to service improvement are promoted in a variety of service sectors (e.g.McDermott et al. 2015).This includes policies funding and sponsoring different initiatives such as technological work re-design, open and collaborative innovation, and innovation training programmes.In the UK, like many other countries facing NPM reforms, delivery organizations within sectors such as social services, health, education, and economic development have attempted to innovate due to decreasing resource availability and growing demands for services (Dudau, Kominis, and Szocs 2018;Melo, Sarrico, and Radnor 2010).In the UK an in-direct approach to delivering public services is taken which leaves room for wider stakeholder engagement (Parrado et al. 2013).As such, many social enterprises are working alongside governmental bodies to deliver public services (Calò et al. 2018).

The training programme
This research specifically focused on public service delivery organizations who engaged with an ITP.The programme provides the opportunity for front-line staff to develop ideas, develop skills and promote innovation.It is delivered by a specialist third-sector organization and was created to support public service organizations enhance innovation.It is funded jointly by the Scottish Government, charitable foundations, and participating organizations.The overarching aim of the initiative is to improve service delivery and create solutions to social problems through the transformation of internal organizational culture.The programme provides a combination of training, coaching and a pitching competition.It focuses on both direct outcomes, such as developing innovation skills amongst the workforce, and indirect outcomes, such as enabling middle managers to facilitate innovation.

Research approach and sample
To explore the practices which foster engagement within the ITP a multiple case study approach was used to build theory (Eisenhardt 1989).This approach can allow factors which can explain the relative success or failure of an innovation initiative to emerge (Stewart 2012), and fits well with exploratory work where little theory or evidence exists for a phenomenon (Eisenhardt, Graebner, and Sonenshein 2016).We use multiple case studies because they are thought to yield more robust insights than single cases (Eisenhardt and Graebner 2007).
As the purpose of this paper is to develop theory, not test it, theoretical sampling was deemed appropriate and consistent with multiple case study approaches to theory building (Eisenhardt 1989;Eisenhardt, Graebner, and Sonenshein 2016).Contrary to random sampling techniques used by deductive researchers (Gehman et al. 2018), theoretical sampling means we selected cases because they were particularly enlightening of the practices enacted to engage employees in innovation initiatives (Eisenhardt and Graebner 2007).We selected cases in consultation with ITP leaders who understood the breadth and depth of the organizations they had provided training for, and their variance in experience.These programme leaders acted as 'gatekeepers' to help the researchers identify 'polar types' of organizations who experienced the ITP differently as to make patterns in the data more easily observable (Eisenhardt 2021).
We ensured organizations had completed the programme during 2019, so that the impacts of the programme were apparent but fresh, enabling high-quality retrospective data (i.e.relatively recent -Ott and Eisenhardt 2020).All of our cases were still attempting to implement innovations developed during the programme at the time of data collection.We selected cases based on their different experiences with the ITP and sampled across the various types of public service delivery organizations that operate in Scotland.This includes large hierarchical organizations delivering large public service contracts, local government bodies delivering multiple different public services and smaller third sector organizations delivering specialist and smaller public service contracts.
We followed protocol for selecting organizations from these three categories to increase comparability (Eisenhardt 2021;Eisenhardt, Graebner, and Sonenshein 2016;Langley and Abdallah 2011).Variation across cases (organization size, structure, and service delivery) enables 'replication under different conditions and so enhances theoretical generalizability' (Ott and Eisenhardt 2020, 281).One of our six initial large hierarchical organizations was unable to contribute fully and was replaced with another large organization that had finished participation in the ITP.Our initial pairings were well matched with regards to organization size, structure, and public service offerings.However, when one of our initial large hierarchical organizations dropped out, they were replaced with another organization that was similar in size and structure but not in what services they delivered.A profile of our six case studies is presented in Table 1.

Data collection
Our fieldwork was carried out at the beginning of 2020, taking roughly four months to complete.It involved collecting data from several different sources to enable triangulation, including semi-structured interviews, corporate and media archival materials, short video interviews conducted immediately after participation, and recordings of the ITP (a summary for each case is presented in Table 1).Combining retrospective and real-time data increased the efficiency and accuracy of our case studies (Ott and Eisenhardt 2020).To further increase reliability and validity, we also interviewed multiple informants in each organization operating at different levels.

Semi-structured interviews with programme deliverers
As a starting point, the researchers conducted six interviews with the CEO, operations manager and four programme managers who were involved in delivering the ITP to case organizations.These were semi structured in nature and typically lasted between 45 and 90 minutes.The aim of these interviews was to understand programme motivations, aims, design and targets.These informants also provided valuable insight into case organizations.Another three semi-structured interviews were conducted with external stakeholders who funded and sponsored the programme.This allowed us to identify the external interest and motivation to invest in the programme and, from their perspective, how they felt the programme added public value.The perspectives from these stakeholders were compared with internal perspectives from case organizations.

Semi-structured interviews with case organization members
We conducted 36 semi-structured interviews with informants from our cases.We collected information from a range of informants, including CEO's, senior leaders, middle managers, and programme participants to enable comparison from different perspectives (Creswell 2007).Interviews were conducted one-on-one, typically took between 30 and 60 minutes, and were thematic in nature.They aimed to understand the interviewee's experiences, feelings, knowledge, and behaviour (Patton 2002).The interviews were recorded and transcribed verbatim, keeping the anonymity of our participants.Our interview guideline (appendix A of the online supplementary file) asked indirect questions looking for changes to employees, organization processes and practices, challenges faced, and outcomes experienced with the programme.To improve the accuracy of the interview data where participants recalled events from previous programme years, the authors triangulated these accounts with video recordings, document archives, and the perspectives of multiple participants within case organizations who engaged with the ITP (Langley and Abdallah 2011).

Video archive analysis
Audio-visual archives belonging to the ITP were also analysed.The videos captured participants reflections immediately after participating in the programme.These videos were typically five to 15 minutes in length.In the context of practice research, video records allow participants to reflect and detail the evolving tasks and actions to make sense of their meaning within organizational processes (Zundel, MacIntosh, and Mackay 2018).Additionally, recordings of the programme being delivered were captured which gave researchers insight into programme participation asynchronously.These video recordings typically lasted enabled us to capture the consecutive construction of events and participants reflections of these events post-participation (Zundel, MacIntosh, and Mackay 2018).

Document analysis
We also reviewed strategic documents, media articles and archive material to understand the intended design and benefits of the ITP.For each case organization we conducted media article searches and also examined internal organization documents that covered the participation, outcomes, and experience of the ITP.The analysis allowed us to scrutinize the programme and understand motivations, aims, design and targets, as well as offering insight into participant experiences.

Data analysis
To understand how public service organizations can engage employees with innovation training we adopted a theoretical approach to practice (Feldman and Orlikowski 2011).This included interpreting the empirical material to understand theoretical relationships between engaging activities for each case.Following the multiple case study approach (Eisenhardt and Graebner 2007), we began our analysis by examining each case individually in relation to our research questions.For each case we fleshed out key events, decisions, and experiences of participants in relation to the programme.The semi-structured and video recordings were particularly useful for providing rich details, while media articles and corporate documents proved useful for understanding impacts and strategic practices.We wrote up detailed case descriptions that identified emerging patterns within each case study understanding impact and outcomes, and which practices and conditions helped facilitate (or impede) these.
We then moved to cross-case comparisons, looking for similarity, difference, consistent themes and underlying conditions and context for explanations (Eisenhardt and Graebner 2007;Miles et al. 2014).At this stage, Microsoft Excel was used as a template for organizing individual codes and quotes which we classified into appropriate categories (Saldaña 2021).We used comparative tables to list conditions and practices from each case to shed light on the intricacies of the practices.This creative process enabled us to look within pairs, across pairs and across organizations to create emerging constructs which explained how the ITP was received differently and examine the practices and conditions which provided explanation for this (Klag and Langley 2013).At this stage, consistent with practice outlined by Eisenhardt and Graebner (2007), we refined our analysis through replication logic and iteratively compared our data with existing literature to sharpen our understanding.This led to our final constructs being formed around different strategic motives for engaging with the programme (complementary or transformational) and different practices used to engage employees (structural and embedding).

Findings
To present our findings, we first demonstrate the variance in each cases experience of the programme in regard to engagement, learning and innovation.We then establish the internal conditions and practices that acted to foster or limit employee engagement with the ITP.Our analysis revealed that the two sets of practices which offer explanation for the observed variance -structural and embedding -were influenced by the underlying strategic motives for participating in the ITP.Example evidence for each of these constructs is presented in appendix B of the online supplementary file.The quotes that are used in the paper are the most representative of the research findings (Patton 2002).

Programme engagement and learning
The variation in each case's engagement with the ITP programme is summarized in Table 2.For Case A engagement with the programme was high, while Case B was low.Cases C, D, E and F had mixed experiences.
Case A stood out as having the highest level of engagement across cases.They committed to running the programme multiple times, with many staff participated in training.Senior staff were steadfast on the impact of the programme, declaring it a success: I think that the [ITP] is a fantastic lever.And it's been a fantastic lever for me to take the organisation to the type of innovative thinking that I would want to take it to.I guarantee you we wouldn't have got . . .We'd have got there in the end, perhaps, but it wouldn't have been anywhere near as successful as it has been without the [ITP] (Case A, CEO).
In contrast, the overall engagement of Case B was low.They experienced a lot of resistance to 'break from the past' and try anything new.Senior leaders rather dismissed the initiative as a short-term fix: I've been pleased with the first wave of levels of participation, but I'm also very conscious of the levels of scepticism as to whether this will work, or whether it will be another short-term fix (Case B, CEO).For cases with mixed experienced with the programme, some factions of the organization engaged with the programme, while others were reluctant.This was emphasized by the CEO of Case C: It was mixed I would say.There were some areas of the organisation that were very enthusiastic and engaged.And then, there were others that were far more sceptical and didn't engage (Case C, CEO).
Despite these mixed experiences, all organizations reported benefits to participating in the ITP.Specifically, skills development in refining and developing innovation opportunities and presenting and pitching ideas were widely acknowledged by programme participants.Additionally, Cases A, C and E reported that the ITP programme had a longer-term impact on increasing staff self-confidence.One middle manager expressed the transition that one of his team members experienced by engaging with the training: Across our cases, we found organizational learning took place in two main ways: breaking down boundaries and cross-departmental learning; and increasing awareness of strategic goals and priorities.For five out of six cases, engaging with the ITP broke down departmental boundaries, which helped staff learn about other areas in the business, communicate better and in some circumstances collaborate more efficiently with other organization units.For Case A, this was important to break down siloed thinking, as expressed by one senior leader: We had all these ridiculous silos that were inhibiting some of what we want to achieve strategically in terms of our vision and mission.So, one of the things [ITP] did has helped to . . .It broke down those barriers that existed, that people had happily wore every day, and they took them off.And so, we got synergy between roles, teams, geography, status, all of that (Case A, Senior leader).
Cases A and B were also able to benefit from increased strategic awareness.Engagement in training helped them to disseminate their strategic aims by explicitly aligning new strategic aims with the focus of the programme.Rather than seeing the initiative as something external that comes into the organization, they made sure it was tailored and purposeful, as a senior leader went on to explain: So, there's a disconnect that we wanted to narrow.We wanted to connect people into that.And by saying, okay, you've got an idea and the investment criteria for this idea, is actually the stuff that's in our strategic plan.You go, oh, what's that then?And so, you're instantly connected in.In a way that before, you never were (Case A, Senior leader).
In short, the analysis revealed that public sector service organizations experience and engage with the ITP to varying degrees.This had implications for the innovation that were demonstrated during and after the programme.For Case A, continuous service and workplace innovations were developed and implemented hand in hand with the ITP.Engagement in the programme was high, with committed employees, senior leadership, and management.For Case B many workplace innovations were developed but faced difficulties with employee engagement, implementation, and commitment from senior leaders.For Case E, a number of service innovations were developed, but faced difficulties with implementation.For Cases C, D and F, a few service innovations were developed and implemented.Examples of various innovations developed for each case organization is presented in appendix C of the online supplementary file.

Strategic motives for engagement
Strategically, the ITP was either regarded as 'transformative' or 'complementary'.Motives for engagement were either to bring about cultural change or focus on staff development.A summary of the strategic conditions in each case study is presented in Table 3.
For three of our cases (A, B, C), the ITP was seen as 'vehicle for transformation', something that could unite organizations behind new strategic agendas.These organizations were going through various changes which impacted strategic direction, such as changes in leadership or restructuring.For these cases, engagement and disengagement in the programme seemed much more intense.The changes organizations experienced created uncertainty for workers, which senior leaders looked to address through promoting new innovation initiatives.
So, there'd been some upheaval that probably had some . . .I don't want to say a hugely negative impact, but a bit of shock and awe impact actually.Yes, it was disruptive.But we knew that wasn't going to be the end, so we needed something that would help us get over the disruption (Case A, Senior leader).
For Cases D, E and F engaging with the ITP was seen as complimentary to current strategy.In relative terms, these organizations appeared 'stable' in comparison to the other cases.The focus of the programme was to improve staff engagement, or indeed the means by which they engaged with staff.The impetus was not for change but more of an incremental improvement to existing practices.The intensity of the ITP was much less and the impact various practices to foster or restrict engagement didn't seem to have as much of an impact as organizations looking for transformation.The workforce for complimentary cases was seen as a resource, which when engaged with in more efficient ways could improve services and organization processes.As one CEO states: The good ideas come from both the top and the lower reaches of the organisation, where people are actually engaging with the front-line issues.So, I've always known that that was the case . . .And so, it was an opportunity for us to extend and engage the whole team in something around about developing.And we're an entrepreneurial organisation, so actually it's the thing we should be doing (Case E, CEO).
The ITP, therefore, was seen as a novel way in which improvements could be made without the strategic imperative for 'transformation'.

Structural practices to enable engagement
Within each case, we identified practices that aimed to formalize engagement in innovation initiatives.We labelled these practices as 'structural' as they were material and aimed to enable employee participation in pursuing the development and implementation of innovation.The analysis revealed four main structural practices for enabling engagement: resource investment, organization restructuring, implementing complimentary initiatives, and forming networks and partnerships.Adversely, in each case there was evidence of practices that restricted employee engagement in the innovation initiative, specifically: organization decoupling and constraining resources.These practices were predominately enacted by management and senior leaders.A summary of the structural practices evident in each case is presented in Table 4.
For Case A, a number of changes were made.The biggest change involved restructuring, where they created an innovation team, operating in a matrix structure across organization departments to create and maximize innovation.The role of this team was to take innovations developed during ITP and implement them throughout the organization.As described by senior leadership, this aimed to cement their innovative thinking: I wanted the [ITP] to give us that impetus to create that sense of a creative, intrapreneurial thinking within the organization.So, it has done that . . .So, what I decided to do this year was that we need to now make that something that the organization does as a way of working . .

. So, we've developed an innovations development team . . . [To use] as a portal for constant innovation (Case A, CEO).
This was backed with the commitment of resources to support the development and implementation of further innovation ideas: You've got to build on it.And that's a game where I'm seeing just now is that you cannot just be allocating £25,000 to that process and think that that's going to be enough to get your organisation developing what you want (Case A, CEO).
As well as these structural changes, senior leaders ensured that innovation was embedded within strategy.As well as formalizing this through strategic documents and messaging, they also incorporated a number of complimentary initiatives into the organization.These included leadership development programmes and peer learning networks.Other cases demonstrated similar practices, but without the scale or commitment demonstrated by Case A. Case C, for example, included the ITP amongst a whole suite of policy initiatives aiming to develop entrepreneurship and innovation.This included expanding their network to include public and private partners and developing incubator and accelerator programmes to help drive innovation in the local authority.Practices that were identified as restricting the effectiveness of the ITP included constraining resource and organizational decoupling.All cases experienced a 'disconnect' between different levels and units of the organization which impeded the implementation of innovation.These disconnects related to breakdown in communication or a misalignment of motives and actions between different areas of an organization.The most restricting practice our data revealed was a decoupling between front-line workers with middle managers and leaders.This was reflected by a participant in Case C who explained why the innovation idea she developed in the ITP was not implemented: It wasn't chosen in the final.We didn't get the money.What we did get was a commitment to going forward.There was already something happening.So, instead of being on one system or another, we would get much more of a collective [impact].That was happening in the background.We were told once that had happened, we would be brought in to maybe look at that front facing bit, which it would be the app, which would work.Unfortunately, there's been quite a lot of issues there.Although the backend system is still on the agenda and senior management are still taking that forward, they've never got to the stage of actually getting us in now to do the frontend of it (Case C, Participant).
In this quote, she highlights that she did not get the resources needed to deliver the initial innovation.She also highlights a failure to connect with another department to develop a joint product, as well as disconnect with senior leaders of the organization.She also alludes to a lack of strategic commitment from leaders to develop the innovation, using a vague notion that the 'back-end' product is still on the agenda, but her innovation is in 'limbo' having not gone anywhere.Cases B, D and E expressed constraint in the resources they wished to invest in implementing innovation ideas developed during the ITP.

Embedding practices to facilitate a culture of innovation
Evident across cases were practices that were more discreet than structural practices and helped to embed a positive perception towards innovation.The aim of these practices was to increase the credibility of the ITP.Practices aimed at promoting innovation included: championing the programme, tailoring initiatives, and visual reinforcement.Adversely, practices aimed at preventing or reducing the credibility of the ITP included: trash-talk and failing to promote the programme.These practices were evident across organization members and not just limited to management.A summary of the embedding practices evident is presented in Table 5.
Case A's senior leaders worked hard to increase engagement and acceptance from the wider organization.One senior leader who championed the initiative early on explained how she had to work hard to increase exposure and bring attention to the programme: It took a lot of hard work to convince people to actually voice their ideas and that they would be believed, valued, and legitimised . . .Lots of our middle managers weren't open to this, so we talked about the [ITP] a lot.We talked about the possibilities.In all our learning programmes, we talked about the [ITP].In our induction programme, we talk about the [ITP] (Case A, Senior Manager).
Through building the ITP programme up, they were able to promote the benefits and opportunities it provided, which helped to create a sense of excitement.One board member, after viewing a pitching session expressed: The staff really appreciate it because they get a buzz out of it, there really is a terrific buzz.I have noticed that every year.Just talking to members of staff and they are all saying how much they enjoyed it (Case A, Senior Leader).
This sense of excitement helped to create credibility for the ITP in the organization, which helped to increase engagement.After participants had been through the ITP, they passed their experiences and the benefits they received to peers, effectively acting as intrapreneur champions, as illustrated by one ITP participant:

I am really passionate about [ITP] as I've obviously been through [ITP] and seen how successful our project is, so I'm a real advocate for [ITP]. I'm always encouraging our other staff and our other teams to put themselves forwards (Case A, Participant).
After engaging with the ITP for three years, however, senior leaders in Case A acknowledge that to further embed intrapreneurship into their organization they would need to create a sense of distinctiveness around intrapreneurship development.As illustrated by the CEO, they were looking to incorporate the initiative with their own mark to further cement the benefits: I think organisations need to take it and say that's been really successful.I'm going to make that the way that we work.Do you know?And when you do that, you've got to put your own mark on it.You've got to put your own brand on it, do you know what I mean?(Case A, CEO) These practices worked hand in hand with the structural changes they made, such as resource investments and restructuring to cement their learning into the organization.The collective activities helped them to maintain the credibility of their attempt to transform organization culture.This is illustrated by the international reputation some of their innovations were receiving, as highlighted by a participant: The woman that pitched that and got the funding for it, one of the [service] advisors, she was at the [health] Europe conference . . .And she was making a presentation -it was a poster presentation she had -on [her innovation].And she won second prize for the best poster presentation on the actual project in Europe.That type of thing, which got its funding through [ITP], is being recognised as a great development (Case A, Participant).
This success acted as a visual reinforcement for other organization members to align behind.Other means to reinforce the ITP programme included promotional posters and communication through emails.Some organizations engaged in practices which fostered reluctance, scepticism, and restricted credibility.For Case B, they had a historical backdrop of failed initiatives that looked to reform their services and organizational culture.This was noted by the new leader who wanted to create change: 'The context and the backdrop is an organisation that has seen many initiatives that have failed to be followed through without proper commitment to the senior leadership team'.There was also a sense of arrogance from Case B, with regards to the need for ITP, they viewed themselves as already knowledgeable about the subject.The CEO questioned the need for the ITP programme, believing the organization was 'above' the level of the training: The question is why do we even need the [ITP]?We are the people who are driving workforce innovation.Why don't we just turn our specialism on our service and mainstream it for ourselves?Why aren't we doing that?' (Case B, CEO).
The senior leader's 'trash-talk' trickled down to participants who generally had mixed feelings about the ITP programme and shared in the practice of speaking negatively about the programme.One participant from Case C highlighted how her colleagues would speak badly about the programme: "For the ones that have been involved it's been amazing.For others that haven't been involved and don't know anything about it, not so much.I just overheard the conversation, some other people saying they talk nonsense.People not knowing that I was involved talking at the desks around me" (Case C, Participant) A lack of attention to employees that were pursuing innovation within the programme was also evident.This practice restricted front-line staff's awareness and understanding of the programme, emphasized by one participant who by chance came across the programme: "I work in entrepreneurship so it's right up my street and it wasn't even on my radar.If Andrew hadn't sent around directly to us do you want to pick this up?It would have bypassed me and everyone in our team as well."(Case C, Participant)

Discussion
The existing literature highlights the importance of various initiatives to engage employees and wider stakeholders in public sector innovation processes (e.g.Brogaard 2017;Lindsay et al. 2018;McGann, Wells, and Blomkamp 2021).Within this body of literature it is recognized that there is a lack of research which connects with the practice of 'doing engagement' (Fletcher et al. 2020).We attempt to contribute to this work by taking a practice approach to explore how six organizations in Scotland, UK, fostered employee's engagement in an ITP.Our findings and subsequent model offer insight into our two research questions: What specific organizational practices influence employee engagement?How do public service delivery organizations enact various practices to foster employee engagement in innovation initiatives?

A model of the practices used to foster employee engagement in public service innovation initiatives
Our findings, summarized in Figure 1, explain the conditions in which employee engagement in innovation initiatives can be fostered.
We identify two sets of practices for fostering employee engagement -structural and embedding.Structural practices, which were material and aimed to reorganize organization processes to enable engagement, were enacted by the leaders and managers of organizations who had control of resources and decision-making power.These practices either enabled or restricted employees from engaging in innovation initiatives.The organizations that looked to enable employee's engagement looked to free-up extra resource, change organization structures and work patterns.Practices that restricted engagement included decoupling organization units to prevent ease of participation and withholding resources.In light of these findings, we put forward the following propositions: Proposition 1a: Organisations that enact enabling structural practices are likely to have higher employee engagement.

Proposition 1b:
Organisations that enact restricting structural practices are likely to have lower employee engagement.
Embedding practices looked to either promote or prevent innovation engagement.These practices were more discreet than structural practices and looked to influence people's perceptions about the innovation initiative.They were adopted across all levels of organization and included championing, tailoring the initiative to make it fit the organization context, as well as using visual reinforcement to align and enthuse employees.Adversely, prevention practices were highlighted which included trashtalking the programme and failing to internally promote the initiative.In light of these findings, we put forward the following propositions: Proposition 2a: Organisations that enact embedding practices to promote innovation are likely to have higher employee engagement.Proposition 2b: Organisations that enact embedding practices to prevent innovation are likely to have lower employee engagement.
Our findings also highlight a moderating role that the initial strategic motives for engaging with the ITP had on the relationship between the various practices.Two motives were identified -transformative and complementary.For the three cases which viewed the innovation initiative as complimentary to existing strategic approaches, mixed engagement levels were identified.For each of these three cases, various practices to foster or restrict engagement in the initiative had less of an overall impact.Adversely, for the three organizations that viewed the programme as transformational to strategic approaches, the practices enacted had a greater impact.That is, practices to limit engagement seemed more damaging to the ITP, whilst practices to foster seemed to be much more rewarding.In light of these findings, we put forward the following propositions: Proposition 3a: The practices enacted by organisations with a transformational strategic motive have a higher impact on employee engagement.

Proposition 3b:
The practices enacted by organisations with a complementary strategic motive have a lower impact on employee engagement.
Our framework which details how public service organizations can foster employee engagement in innovation initiatives offers insight into why public service organizations may experience divergence in the outcomes of innovation initiatives.For Case A, practices enacted mainly enabled and promoted the innovation initiative, and paired with a need for strategic transformation to increase the intensity of these practices, created the internal conditions for employee's high engagement.Adversely, Case B enacted mainly restricting and preventive practices, which were heightened by a motive for transformation, ultimately led to low engagement with the initiative.For cases with mixed engagement levels, evidence of practices that both enabled and restricted employee engagement were evident.Effectively, organizations that were able to engage employees demonstrated greater learning and innovation outcomes, while those that were unable to engage employees demonstrated lower learning and innovation outcomes.

Boundary conditions, limitations, and future research
We identify some important boundary conditions for our model which are relevant to theoretical generalizability, and also offer promising directions for future research.Size, structure, availability of resources and policy pressures seemed to influence the level of impact various practices aimed at fostering engagement had.The smaller specialist delivery organizations appeared to lack the resource commitment to fully capitalize on the ITP.Alternatively, the larger local authority organizations appeared to be too diffused in their organization structure and service delivery to unite behind new strategic initiatives.The large hierarchical organizations appeared to have the best conditions for engagement.They were large enough to commit resources, had an external pressure to innovate and had focused services that benefited from the added value of innovation.
The geographic and sectoral contexts in which innovation initiatives are implemented are also important boundary conditions.While our model focuses on the internal practices which foster engagement, we also acknowledge the importance of external influences.Practice-orientated scholars stress the importance of both history and the external environmental context in driving various practices (Feldman and Orlikowski 2011;Nicolini and Monteiro 2017;Schatzki 2005).There are differences, for example, in the delivery of public services in centralized state administrations such as France and Germany with the UK (Parrado et al. 2013).It is likely in these contexts that strategic motives, structural practices, and embedding practices vary.
It is also possible that specific sector characteristics can influence the motives and practices which engage employees in innovation initiatives.While both Cinar, Trott, and Simms (2019) and De Vries, Bekkers, and Tummers (2016) in recent reviews do not find sector characteristics as a commonly identified barriers or drivers to innovation, there are likely shared practices across sectors which vary.Although the specific practices identified in this study may vary, the constructs are likely generalizable across these sectoral differences.Future research can explore the influence that such contextual conditions have on how employees engage with innovation initiatives across locations and sectors.
Considering the work that various stakeholders have to do to implement change in public sector contexts (e.g.Cloutier et al. 2016), our findings are likely generalizable to other innovation initiatives.Various programmes which drive organization change and require engagement from employees, such as technologydriven reform (e.g.Lindsay et al. 2018), can benefit from identifying the various structural and embedding practices that can facilitate implementation.Additionally, the structural and embedding practices likely apply when public service organizations are looking to engage wider stakeholders in the innovation process, such as citizens (e.g.McGann, Wells, and Blomkamp 2021).Nevertheless, future research can empirically validate the various practices that work to foster engagement in various innovation initiatives.
Further empirical research is needed to test the constructs outlined in this paper on wider innovation initiatives, stakeholders, and contexts.To do this there is work needed to operationalize the constructs uncovered in this study.First, future research needs to account for the potential of multiple different strategic motives for organizations to participate in innovation initiatives.It is important to consider various internal motives for programme engagement (i.e.culture change, learning and development, operations efficiency, service improvement) as well as externally mandated performance measures.These different motives are likely to frame the same innovation initiative differently across different public service delivery organizations.
Second, further research is needed to capture the various social practices enacted by different organizations across wider contexts.While the majority of practice-based researchers advocate qualitative 'on-site' research methods (e.g.Nicolini and Monteiro 2017), quantitative survey instruments can be used to capture the frequency and diversity of practices performed, and their common variants and characteristics across populations (Browne et al. 2014;Fernandez and Moldogaziev 2013).An example question for embedding practices could be 'how many times today have you seen an advertisement or received a reminder of the [innovation initiative]?'.The development of a survey can be used to empirically validate the presence of various structural and embedding practices and their relationship with various innovation and learning outcomes across larger samples.

Research contributions
Our findings and subsequent model make three main contributions to the literature on public sector innovation.First, we advance a practice approach to explore how employee engagement with specific tasks is fostered (George and Desmidt 2014;Hansen 2011;Huijbregts, George, and Bekkers 2021).By viewing engagement as something that can be fostered through practice, we detail actions that can be done to mobilize employees to pursue innovation.This moves away from existing conceptualizations of employee engagement as a static variable (Rangarajan 2008;Yuriev, Boiral, and Talbot 2021), by highlighting the complex and dynamic nature in which engagement can be continuously worked upon.We also highlight the importance of viewing employee engagement as task-specific, challenging existing holistic conceptualizations (e.g.Ancarani et al. 2021).Viewing engagement as a more specific construct can provide researcher with greater precision on how to align and engage employees in certain situations, such as during technology-driven reform (e.g.Lindsay et al. 2018).
Second, our findings advance the innovation barriers literature by identifying various practices which act to engage employees.These practices which exist across two dimensions -structural and embedding -were intrinsically linked with the innovation barrier (employee engagement).As such, it is likely that various organizational practices co-evolve with various barriers which together can act to restrict or drive innovation.Barriers such as budget constraint, lack of support or low employee engagement (e.g.Cinar, Trott, and Simms 2019;Demircioglu and Audretsch 2017), therefore, should not be considered in isolation from the organizational practices which are enacted to maintain them.This positions innovation barriers and drivers not as discrete entities, but the enduring practices that structure organization processes (Nicolini and Monteiro 2017;Schatzki 2002).Future research should therefore view engagement, learning and innovation as evolving practices within public sector contexts.This can help create a more dynamic understanding of existing barrier and driver templates (Cinar, Trott, and Simms 2019;Tremml 2021).
Finally, we contribute to the existing literature that focuses on how innovation can be fostered in the public sector through various initiatives, such as innovation training (Brogaard 2017), innovation labs (McGann, Wells, andBlomkamp 2021;Tõnurist, Kattel, and Lember 2017), or innovation awards (Bernier, Hafsi, and Deschamps 2015).Not only do we identify practices which can foster their impact, we demonstrate the importance of understanding the strategic motives for why public service organizations engage with innovation initiatives.Understanding how innovation initiatives fit in with wider organizational goals and objectives is important as it influences the strength that various practices have on engaging employees.This has implications for current research as it stresses that strategic approaches frame innovation practices and levels of engagement (e.g.Clausen, Demircioglu, and Alsos 2020;De Vries, Bekkers, and Tummers 2016;Demircioglu and Audretsch 2017).This advocates that when it comes to engaging with innovation initiatives, a one-size-fits all approach is not feasible and must be tailored to organization contexts.

Implications for policy and practice
The results from this study have implications for public sector practitioners faced with incentive to innovate.First, innovation initiatives should not be viewed as blanket interventions that can be applied to increase innovation performance.Rather, organization conditions and existing social practices will influence the ability of employees to engage in the innovation process and must be taken into consideration when designing initiatives.The practices outlined in our study can be viewed as a blueprint to foster employee engagement and maximize benefits.Second, it is important to consider the size and scope of public service organizations who engage with innovation initiatives.In our results, larger public agencies delivering focused public services appeared to be able to benefit the most from participation, having the ability to commit resources and apply specific innovations to improve service delivery.Finally, our results indicate that innovation initiatives are most effective when viewed as commitment devices to enable strategic transformation.Practitioners, therefore, should not look at innovation initiatives as short-term incentives but as commitment devices that need to be embedded in longer term strategic plans.

Conclusion
This paper explored how public service organizations foster employee engagement in innovation initiatives.We adopted a practice lens to explore the experiences of six case organizations participating in an ITP in Scotland, UK.We illustrate how various structural and embedding practices act to either restrict or enable employee's engagement.These practices are influenced by an underlying strategic motive for participation.This study offers one of the first theoretically informed and empirically grounded investigations into how public service organizations 'do' employee engagement.In contrast to existing studies that look at innovation barriers and drivers as static, we emphasize that various barriers should not be considered in isolation from the organizational practices which are enacted to maintain them.This positions innovation barriers and drivers not as discrete entities, but the enduring practices that structure organization processes.We hope our findings will encourage further research using a practice lens to understand public sector innovation processes.

Figure 1 .
Figure 1.A model of the practices used to foster employee engagement in public service innovation initiatives.

Table 1 .
Description of sample firms and data collection.120 minutes.All cases had video material, except for Case E. To mitigate this potential weakness, we conducted several more interviews for this case.The combination of participant reflection videos and training session recordings

Table 2 .
Summary of ITP engagement and learning experiences for each case study.
* d represents innovations that were developed in the ITP, but never implemented.** i represents innovations that were developed in the ITP and implemented within organizations.
had seen her at the[ITP]because her confidence and her stature . . . it felt to me like it was a TED talk I was watching, which is that the TED talks, your top dogs, your people you believe, you absorb that and you think, my goodness, they know their stuff inside out.And that's what  [participant]looked like, and that was the first time I'd seen her be a peacock and stand and just say, this is my message and you better listen (Case E, Manager).

Table 3 .
Summary of strategic conditions for each case study.

Table 4 .
Summary of structural practices evident in cases.

Table 5 .
Summary of embedding practices evident in cases.