The Leadership of Co-Production in Health and Social Care Integration in Scotland: A Qualitative Study

Abstract The involvement of citizens in the production and creation of public services has become a central tenet for administrations internationally. In Scotland, co-production has underpinned the integration of health and social care via the Public Bodies (Joint Working) (Scotland) Act 2014. We report on a qualitative study that examined the experiences and perspectives of local and national leaders in Scotland on undertaking and sustaining co-production in public services. By adopting a meso and macro perspective, we interviewed senior planning officers from eight health and social care partnership areas in Scotland and key actors in national agencies. The findings suggest that an overly complex Scottish governance landscape undermines the sustainability of co-production efforts. As part of a COVID-19 recovery, both the implementation of meaningful co-production and coordinated leadership for health and social care in Scotland need to be addressed, as should the development of evaluation capacities of those working across health and social care boundaries so that co-production can be evaluated and report to inform the future of the integration agenda.


Introduction
How do leaders of health and social care integration understand and seek to operationalise co-production? This article examines how co-production is interpreted by health and social care leaders in Scotland. Although there have been other studies of co-production in social care in Scotland, particularly around the personalisation of care (Flemig and Osborne, ), our research examines co-production for health and social care integration as a governance reform agenda. In this respect, the understandings and experiences of health and social care managers in embedding co-production within, and across, integration partnerships represent the dominant focus of this article.
A systematic review of co-production within health and social care settings concluded that the academic literature is weak on how to do co-production and, despite it sounding virtuous and desirable, there are significant ambiguities about what it means (Connolly et al., a). Indeed, co-production is often interpreted as a 'woolly-word' in public policy (Osborne et al., ) and vacuous policy signaling can undermine the potential for co-production to happen (Needham and Carr, ; Slay and Stephens, ; Flinders et al., ; Oliver et al., ). It is the lived experience of such a 'woolly-word' that this article seeks to examine i.e. lived by those whose task it is to promote and deliver co-production. Respondents discussed their experiences of co-production in terms of examples from their health and social care partnership area, but they also talked about co-production from a normative perspective and how this applied to the nature of their work. In short, health and social care leaders accounted for their distance from everyday interactions with citizens, and this included framing their joint working with other professionals within their employer, or the partnership, as co-production.
In Scotland, there is a strong degree of policy signalling in health and social care policy about the benefits of co-production, which call for approaches to be embedded, enabled and have meaning locally, but lack clarity on the implementation steps to deliver effective co-production (see Scottish Government, ; Christie Commission, ; Scottish Government, a; b; ; ). In Scotland, co-production is recognised by public and third sector bodies as important for achieving positive outcomes (for e.g. Alliance Scotland, ; Healthcare Improvement Scotland, ). However, little is known about how co-production is understood and the barriers to its operationalisation within the Scottish health and social care system. The aforementioned systematic review also includes a review of Scottish policy documents in relation to health and social care (Connolly et al., a). This analysis shows that co-production is present within many policy messages in Scotland yet guidance and support for undertaking co-production (including when it is appropriate to do so, in which contexts and with whom) is absent (Connolly et al., a).

The policy context
One of the key policy instruments driving the public services reform agenda in Scotland is the National Performance Framework (NPF), which presents a number of high level outcomes for public services to demonstrate their performance against (Scottish Government, a). Within this broader context the Public Bodies (Joint Working) (Scotland) Act  became the legislative framework for the integration -       of health and social care. This created new public organisations ('Integration Authorities') which aimed to break down barriers to joint working between National Health Service (NHS) Boards and Local Authority Areas (in which social care provision is historically located). This resulted in the establishment of  health and social care partnership areas, each led by a Chief Officer. As part of the integration agenda co-production is recognised by public and third sector or Non-Governmental Organisations (NGOs) organisations in Scotland as important, based on a general view that these approaches can lead to the achievement of positive outcomes for citizens. That said, little is known about how co-production is understood, implemented, and sustained within health and social care organisations. Given the normative centrality of co-production for improving public services, (Bovaird et al., ), it is timely to investigate the extent to which these approaches are understood, operationalised and sustained as part of the integration of health and social care and based on the occupational experiences of those tasked with leading and undertaking integration.

Co-production in the context of modern governance
The trend in West European public sector reform has been based on national authorities 'enabling' local and partnership agencies to take responsibility for service delivery (Kickert, ). The literature is clear about the importance of national policy agendas in shaping the cultures, practices, behaviours and implementation approaches to 'integration' via interagency working (Torfing et al., ). However, within empowered, localised or enabled systems of administrative governance, a great deal also depends on 'boundary spanning' public servants who can operate in collaborative and inter-organisational settings (van Meerkerk and Edelenbos, ). These are individuals who seek to galvanize silos and span boundaries. In overall terms, meaningful co-production requires agile systems, leadership, management and operational capacities and processes to accommodate such an endeavour.
Co-production has been linked to ideas of how to co-design public and third sector services with citizens as part of modern governance processes (Bovaird et al., ). A popular definition is 'the voluntary or involuntary involvement of public service users in any of the design, management, delivery and/or evaluation of public services' (Osborne et al., , p.). Yet, when it comes to a system or inter-organisational perspectives of co-production this takes us into the terrain of collectivist interpretations. In other words, as Bovaird et al. (, p.) note, collective co-production is about organising services to achieve outcomes in order to address specific concerns for particular groups or social causes. Boundary-spanning leadership within devolved complex governance landscapes (such as Scotland) requires the navigation of distributed and unequal power relationships. This article highlights the tensions   ,  .
between the central and local government, which is expressed in control, knowledge and decision-making power and has created challenges when attempting to foster a co-productive approach in health and social care integration as a reform process. The co-production literature demonstrates how complex power relationships and often competing or conflicting incentives, expectations and priorities may frustrate policy development and innovation, even in policy areas that could be seen as depoliticised (Flinders et al., ; Turnhout et al., ). Cross-boundary working is often the result of macro-level (national government level) mandates to operate co-productively. This is seen within health and social care as meso-level or partnership level actors (e.g. local government and NHS board personnel) are important agents: as functional and/or geographical 'boundary spanners' to facilitate programmes, services and initiatives (Klijn et al., ).
Based on the experiences of those tasked with leading and undertaking integration, the key objective for the study is to understand the perspectives of service planners in the implementation of co-production. This qualitative study addressed three key research questions: • What is meant and understood by the term 'co-production'? • How can co-production can be operationalised?
• How can co-production be sustained in the longer term?

Research design
Data collection involved semi-structured individual interviews, while informants were chosen using purposeful sampling. The objective was to elicit the views of meso (partnership) and macro (national) level actors on how co-production is understood and sustained across Scotland. Respondents were identified via policy documentation, web searches and through the relevant networks of the Project Board for the study. Participants were invited to take part based on their occupational positions and were offered the choice of being interviewed face-to-face or by telephone. Those who were invited to take part were given a Participant Information Sheet that included study details, and assurances of anonymity and confidentiality as well as reassurance that participation in the study was entirely voluntary. Those who took part provided their consent in writing.
The macro-level interviews secured national public sector organisations which have a leadership role in health and social care governance in Scotland, plus one major third sector advocacy body. The interviews were conducted by two members of the research team from February to December  (JC and TM). Interviews lasted on average  minutes. All interviews were -       conducted in a non-directive manner allowing participants to exercise a measure of control over the processes and encouraged them to talk openly and move relatively freely in their descriptions. Furthermore, both researchers kept a journal with field notes that reflected on the research process and potential biases that may have influenced the findings. These served to allow for internal crosschecking between the qualitative interviewers to ensure that critical reflection was an ongoing part of the research process, and this is highly recommended within qualitative research as part of verification processes (Ortlipp, ). Overall data analysis was performed iteratively with themes contained within interview questions being developed from a systematic evidence review (Connolly et al., a).
A phased research strategy was adopted in order to present anonymised key findings from the meso-level interviews to the macro-level leaders (those with policy responsibilities at a national level). The identified themes at the mesolevel stage helped to shape the macro-level interviews. These were: meanings of co-production, co-production and its association with improvement, the situational enablers and constraints to co-production, evaluative and performance management approaches to co-production. Data analysis was conducted using the  step framework approach that begins ideally with a verbatim transcription and ends in data interpretation. In between these two stages the analysis involved familiarisation with the data (reading and rereading transcripts), applying some initial coding in the first X transcripts (line by line) and checking for fit with the data, the development of an analytical framework, applying the framework to the remainder of the interviews and charting the data into the framework matrices (Gale et al., ). Data analysis was done by JC and TM who used double coding to conduct analyses of the first few interviews and to develop the analytical framework. All interviews were audio-recorded on an encrypted digital recorder, transcribed verbatim and analysed using NVivo.
In total,  in-depth semi-structured interviews were conducted at mesoand macro-level. Meso-level leaders (n=) were interviewed as the first phase of data collection; representing  of the  Health and Social Care Partnerships (HSCPs). These areas were purposively chosen to reflect the diversity of geographical and socio-economic compositions of HSCPs across Scotland; representing areas of high and low deprivation, as well as urban and rural populations from across the North, South, East and West of the country. Senior managers with responsibility for ensuring co-production of services (leadership responsibilities across NHS and local authority boundaries to deliver the integration agenda) were purposively selected to take part in an interview. Interviewees had experience of working within the health and/or social care in Scotland ranging from - years before the introduction of integration of   ,  .
health and social care in Scotland. The research also included a senior leader in a policy-facing third sector social care body. The leaders of health and social care in Scotland have, for the most part, emerged from within the system and have had considerable experience of service delivery, managerial work, leadership and partnership working over their careers so the interviewees have significant knowledge and experience with regards to service design and have talked to their lived experience of the enablers and barriers of citizen co-production.
Although the dominant focus of this qualitative study was to understand the perspectives and experiences of those leading integration within health and social care partnerships, individuals from national organisations across Scotland were also invited to participate in the research to provide insights on the research questions from the macro level of the system. Six cross-sector agencies were included: Scottish Government, Audit Scotland, the Convention of Scottish Local Authorities (COSLA), Health and Social Care Alliance Scotland, Healthcare Improvement Scotland, and the Improvement Service. Ethical approval was given for this study by the University of the West of Scotland's Ethics Committee. The project was overseen by a Project Advisory Group (made up of officials from three health and social care areas across Scotland as well as representatives from the third sector and national agencies). The group advised on the research design as well as contributing to discussions on the interpretations of data by the research team.

Findings
The meanings and understandings of co-production It became clear through the interviews that diverse meanings were attached to the concept of co-production. Several interviewees articulated that they associated co-production both with a broader change management organisational strategy to foster integration and citizen/service-user involvement. As this Chief Officer explains: : : : there will be different levels a' co-production for me. So there'd be co-production at an individual level. And, and that's where we're talking about an individual and recognising that they're the expert in their own care : : : co-production at an individual level is a bit about, yes, making sure a person's well informed but listening to what that person's priorities are and therefore together, you know, like developing what type of, you know, plan if there is a plan going forward.

(L)
Moreover, the interviewees also suggested that the term 'co-production' was an intrinsically good and beneficial thing to do. For example, L noted that: Our Chief Officer has very kind of clear views around co-production and : : : we need to be doing far less consultation stuff, which I would agree with in general, I think the -       weight should be towards stuff that is much more genuinely about co-designing ideas, concepts and building on those to deliver services and supports, etcetera.

(L)
Similarly, interviewee L said that: I don't use any references [for understanding co-production], but we're just very aware that it has to be a joint effort, it's not about what I think, or what you think, it's about the people who are going to benefit from the service needing to be involved.

(L)
The challenge with the understanding, and application, of the term co-production (and associated terms) was highlighted by L as being a word that was over-used, which impacted on the clarity and operationalisation of its meaning for those who are expected to implement it. That being said, the narratives to emerge from the interviews did suggest a role for co-productive activities at a systems level, not only involving service users in service planning and delivery; but also to support integration activities. In other words, although, co-production was seen as 'about people working together to find a local solution : : : It's about not assuming that we are the experts in any field,' (L). These meanings seemed to be widened to capture intra-organisational work between teams and inter-organisational activity, especially with the third sector. What this indicated was a propensity by interviewees to galvanise, or wrap up, managerial terms together, for example, that co-production was largely synonymous with partnership working: So co-design in one sense I see as happening in those localities very importantly. And we're on a journey there, it's not perfect. One of the things we need to do better on is how we've got good involvement from the third sector and from different providers of health and social care services, whether we're really effectively gathering the views of the communities themselves in those localities, I don't think we're there yet : : : This breadth of meaning attributed to co-production was acknowledged by interviewees, with some arguing that it's not new but rather re-packaged concepts from past practice. There were associations made between co-production and other approaches such as community development and person-centred practice. However, one interviewee highlighted that the diverse meanings connected to co-production led to differences in interpretation, with the result that many of their colleagues used the term to describe their community engagement work: What I wanted, what I was hoping was that, you know, we could've introduced this [coproduction] on a strategic basis so that everybody would be co-producing stuff all the time. I have to say though that we talk about co-production a lot, we : : : the operational staff now wouldn't try and change something or do something differently without using   ,  .
co-production. However, I don't think that they understand that what they're doing is community engagement and not co-production.

(L)
Another interviewee emphasised how use of terms are shaped by the professional backgrounds of those expected to undertake it as part of contemporary integration work. A typical example was: 'You talk about co-production, as a known community worker I talk about community work. It's good old fashioned community work' (L). The language of co-production was also cited by some interviewees as problematic given that there were sensitivities about the use of such language on two levels. First, communities or individuals might not be ready to co-produce due to their own personal social circumstances or wellbeing. Second, the terminology used by policy makers and academics might not resonate and inadvertently create a barrier to the communities and individuals who it is intended to include and support: What I would say is that co-production, you know, once you start giving it names like that and you're trying to talk to ordinary folk, their eyes glaze over as soon as you mention it, you know, so I tend not to use the word co-production or anything when I'm talking to ordinary people : : : . I talk about working together, I talk about changing things for the better. I tend not to use the word co-production straight away, and sort of drip-feed that a bit at a time because there's been so many terms for so many things, you know, that I've noticed it, whenever I use that word, I just see the looks on folks' faces. (L) The above quote points to the barriers created by applying a terminology to co-production which implies not all will 'get' co-production and there is a requirement for some terminological navigation on the part of those leading co-production. Also, privileged elite narratives of 'knowing' co-production could perpetuate separatist power dynamics between policy designers and service users (see also Bevir et al., ).

How co-production can be operationalized
There were a number of examples provided by interviewees who were less familiar with co-production that appeared to be similar to the principles underlying productive approaches, which raises the questions of the predominance of other concepts that overlap with co-productionfor example, assets based community development. The response after discussing co-production from a Chief Officer, who expressed a lack of knowledge, reflected this nevertheless did show co-production being operationalised within the system: We have been looking at what we could've done this winter that would allow us to support more people to remain in their own homes, to depend less on hospitals and unscheduled care from hospitals where they can. So we've been, we've identified -        local people in [the area] who've got a COPD diagnosis, and we've been asking any of them voluntarily to work with us to establish improved self-care arrangements. : : : So we've been working with we think around / of those people to collaborate with them in understanding their circumstances and then co-produce a different model, which is over the course of the winter. We've worked with all of our [the area] community pharmacists : : : And it's had a dramatic effect in terms of the number of admissions to hospital for people with COPD : : : There's an example of how we have collaborated and co-produced a test of change over this winter that will lead to probably a co-produced different way of working.
(L) Yet, at a systems level, a recurring theme was the need to try to accommodate the pace of national policy change; one interviewee suggested it is like 'trying to redesign the plane when we're flying it' (L). Several interviewees perceived an expectation gap between the direction of national policies on integration and the local conditions for their implementation. Integration has also been framed as a way of managing complexity in relation to the delivery of health and social care services. This resonates with Hood's analysis of intergrated working within children's services when he notes that 'observing that much policy rhetoric and official guidance rests on the (false) presumption of controllability' (Hood, , p.), Hood () argues that the current approaches to integration in care is driven by managerial models and is concerned primarily with risk and accountability, with inadequate attention to how events unfold in unique configurations of service users, families/carers, professionals, managers and organisations. The original expectations were, according to an interviewee, based on an original assumption that 'Scotland was actually in the ideal place to make this [integration] work if people embraced it. It had the right conditions in terms of health of the nation but financial constraints and the inability to be able to run its own budgets, that's a whole other thing' (L). The interviews revealed how integration continues to be a mammoth taskone interviewee described it as one of the 'biggest shake ups in the public sector since the establishment of the NHS' (L). National pressure for advancing integration in Scotland was highlighted in one interview, which indicated that those tasked with leading integration in areas were strongly aware of the national concerns regarding the need for 'accelerating progress': Audit Scotland published their report on integration so far : : : that was then followed by the publication of the ministerial steering group kind of review of integration with, I think,  recommendations for local areas around how to, I mean one of the key things that I think was important on that is it represented a very clear commitment from government, and also from COSLA representing local authorities that integration is here to stay and we need to accelerate progress around integration and those recommendations were focused on that. : : : : : : We need to make it a success.
One interviewee noted how the Chief Officer role (the partnership area leader) is an acutely challenging role and a very unique position within the Scottish public sector; resulting in significant turnover of Chief Officers in recent years: 'As you know also there's been a number of Chief Officers who for various reasons have, kind of have moved on from their post. It is a tough gig, and it does feel quite unique, and it's an interesting arrangement' (L). The interviews gave an insight into how paradoxical tensions can breed frustration at senior levels within health and social care partnerships and that this, along with multiple/ complex accountabilities and lack of control over the decision-making levers to make change happen, can lead to a degree of stagnation and contribute to resignations.
A strong, but rather typical, example of the lack of dovetailing policy imperatives affecting integration was highlighted by the following example of the contradictory messages/directives about how change/improvement ought to happen: Some of the challenges for us is that those same messages are not always coming through : : : One part of the system is saying localise, localise, localise, and the other part of the system is saying regionalise, regionalise : : : The Scottish Government have funded a regional improvement collaborative around children's sides of things so that, it might not be explicit, but it's certainly in the region, that idea of regionalising around children's services has been endorsed by the government in a fairly substantial way. So there's a definite Scottish Government interest in endorsement of that, but that is quite a challenge. Particularly in a partnership where your children's services aren't delegated because all the decisions that are made about that are made in a council governance structure.

(L)
Another interviewee highlighted the inconsistencies in public service leadership policy and implementation. Scottish Government policy encourages localism based on an empowerment, or co-productive approaches, but actually operates in a highly directive and instructional manner, which t serves to constrain local governance. The entangled arrangements between local authorities and the NHS are complicated by interpretive contestations around integration as a process or an entity. In terms of instructional leadership, a strong example is the national direction to recruit more health visitors, with no additional central funding or consideration of the impact on local funding allocations within HSCPs: The types a' things that, that come into us are when we get, at times, external factors. So : : : if somebody comes to us and, as you know, they have. Scottish Government said that you need tae keep up your teacher numbers so you cannae do that. Or for us, you know, you need tae keep up x number of health visitors. And so you, you end up doing things by requirement that are not maybe what you would want tae do : : : : : : When people come down and give you know, dictats about how things have tae be, that can be -       difficult when, you know, you know as an IJB we are funded from both the council and the health board. (L) Given the tensions between centre and periphery, expressed in control, knowledge and decision-making power, challenges emerged when attempting to foster a co-productive approach to integration, and these tensions had everyday consequences in the work of those involved. Hudson () reflected, some ten years ago, on health and social care within the English context: Hudson wondered whether Scotland was heading for a 'partnership fall' given there was little evidence of previous partnership achievements to suggest that 'new heights will be easily conquered' (Hudson, , p. ). He stated that it should not be assumed that in Scotland there will be 'a trouble-free march towards a partnership nirvana', as persistent obstacles to partnership working are found in many places, including separate budget streams, different accountabilities, and inter-professional rivalries (Hudson, , p. ). For Scotland, the integration agenda for health and social care has exposed such obstacles highlighted by Hudson and many of these constraints can be grouped around cultural and systematic barriers. On this point, an interview paints a clear picture of such barriers: It must be quite difficult for the Chief Officer to manage some of that in terms of still working with a council [local authority] or a health board who don't necessarily have the same expectations either nationally or locally around working in those types of ways. : : : People obviously have quite different views about health and social care partnerships, whether they were a good idea or whether they weren't a good idea, and whether it would deliver anything better or not : : : but I think there are a number of people locally and nationally would be delighted if we failed : : : Just at the most basic level, if you do a performance report to the performance and audit committee, so you write it once for there and it goes there, and then it has to go to the Council for information, and they've got health board for information and they both then have got it in a different template so you have to redo the report : : : That's about power and local politics and all that kind of stuff.
(L) One interviewee took a degree of comfort in the fact that challenges exist in other HSCPs in some shape and form beyond their own -'you look at some of the national reports and think yeah, it isn't just us, you know, everybody's struggling with that corporate body interface, everyone, you know, it's not just, this is just difficult as well. So there's some reassurance in that fact that   ,  . everybody's struggling, I don't know if that's a positive thing or not' (L). These comments expose concerns about the systemic and leadership changes required to deliver the shift required to meet future demand. Whilst interviewees highlighted local programmes of work delivered at the local-level that adopted an integrative community approach to planning and delivery, there were examples at the meso (partnership-spanning) level: To be honest we'd, we'd done a lot a' work locally in an integrated way and again locally with an perspective and particularly with a very strong community planning partnership. So we'd worked hard at relationships because : : : it doesn't matter what structures you've got in place. If you don't have the relationships and build and a shared common outcome then it's really, really difficult to move things forward : : : And one a' the dangers for me around integration was that we could have ended up navel gazing completely : : : how do we join up health and social care and forget, actually, the role of our housing colleagues, our education colleagues, our police colleagues, fire and rescue etc. So that bit about having a community planning partnership approach tae improving the health and wellbeing a' communities was right at the heart.

(L)
A programme leader from another health and social care area elaborated on this way of working but stressed the importance of working with the voluntary sector and the role they play supporting co-production: Integration is about leadership. It's about leadership without authority. It's about leadership within a complex system and a leadership whereby you are providing reassurance and context for people who on the whole don't want to change, don't like change. The management and the supervision of staff systems is what it is but the role of leaders and leadership within this complexity is hugely important. Whether that's leaders within Scottish Care who aren't always as helpful as they could be at a national level but for us at very localised level are nothing but supportive and, and helpful and engaged and part of our leadership team. And the CVS, third sector interface would be the same. So that localised leadership is hugely important to actually manage people's expectations whether it be public, whether it be staff and to provide reassurance that we are moving in the right direction. That we are managing the complexity of governance, accountability, and financial management as well as the operational delivery.

(L)
The quote demonstrates the leadership and systems deficiencies with regards to co-production and that leadership towards collective co-production (i.e. engaging with local organisations) requires both humility and support to those organisations to develop their own knowledge and their voice. Yet a major concern confronting all of partnership areas has been the matter of health inequalities and the role of partnerships in addressing the wider social determinants of health: Unless we can tackle the issues around employment, poverty, housing, environment, then if we can't get into those kind of social determinants of health then our ability -       to tackle the health inequalities is always gonna be limited. So I suppose from that perspective the health and social care partnership seeing itself as a player with other community planning partners and being round those strategic tables is really important, so I like to think of the health and social care partnership as a partnership, but also as working in partnership with those other kind of key players, and that's something that's really important for us.

(L)
How co-production can be sustained Given the roles of national bodies tasked with supporting or scrutinising the implementation of integration, one of the key themes to arise from the macrolevel interviews was the need for sustainable resources to embed co-production. As actors in these settings, their varied encounters with local actors afforded them particular insights into patterns of practice across Scotland. Interviewee N stated that, even nationally, 'one of the myths that perpetuates in here is that there are dozens of people working in integration and there aren't' (N). There were also concerns expressed from local respondents regarding the cluttered 'middle ground' of intermediary organisations. Interviewee N also acknowledged that although budgetary constraints represented a barrier to co-production, there was also an issue with not having localised infrastructure or capacity to undertake policy, planning and evaluation and that it might be a question of investment -'there's something to be said for the extent to which councils and the health board are actually funding more than just like the salary of the chief officer in IJB [Integrated Joint Board] meetings'. N rather candidly, while acknowledging the value of concepts like co-production and consequences for the evolution of health and social care, compared the limitations of integration to Winston Churchill's famous saying about democracy, noting that 'But I think it's a little bit like what Churchill said about democracy, you know. It doesn't work but it's better than the alternatives'  .
At a systems level, N reported that in terms of challenges and their composition, they are not all the same and 'there's a whole legacy behind every little stone you lift up'. This means that 'it's very difficult to start to change in any meaningful way'. In many respects, efforts to empower communities and embed co-production in a sustainable manner cannot be divorced from the extant cultures, behaviours and practices that uniquely exist within organisations and specific places. An Audit Scotland () report also discussed the degree of reticence amongst senior individuals in the public sector who did not want integration to happen because iot was percieved that integration enabled a shift in power. One interviewee also highlighted the problems with a culture of blame avoidance and the suspicion associated with national auditing. They expressed some bemusement at the conflicting behaviours seen within health and social care partnerships, whereby senior managers appear to be actively promoting   ,  .
transparencyin this instance, from individuals in unnamed NHS territorial boards. However, in practice, often defensive behaviours are demonstrated: The culture : : : to me was quite shocking, quite shocking. In organisations that on one hand would talk about openness and engaging the public and transparency. And I know from my work in here that it's not as open and transparent as we would like. And then seeing their reaction to inspections where we had chief execs turning up to feedback in meetings where it was way below their pay-grade, sitting very close to inspectors, questioning judgements on very junior inspection teams, disagreeing with things that they had found, not wanting to hear the story at all. So the strength of what we do, I think, is we come in and say to these people actually that's not very healthy, and because of the independence we can say that publicly, we can talk about those things. : : : To be able to talk to chief officers and for them to tell you honestly what the problems are, and that kind of takes a long time to build up, but it's something that we kind of instil in folk here.

(N)
Interviewee N also reinforced the analogous messages that emerged from the other national and local interviews, particularly that integration needs to be supported nationally and that leadership is vitally important. This interviewee also provided an example of a time when there was a realisation that the statutory and third sector needed to co-produce with each other (N). The interview concluded with the point that co-production, as one of the approaches advocated by the Christie Commission () subsequently promoted by the Scottish Government for improving health and social care, has to be seen in the context of a perception that the system is 'buckling'. As a proponent of co-production, they believed that there has been a lack of willingness at national government level to have 'honest conversations' about the current system, which lacks the resources and capacities to meet demand (N).

Summary of findings and contribution of the research
In overall terms, this research advances the study of co-production of health and social care by focusing on the systemic issues in relation to leading integration, including the dynamic between levels of governance (meso and macro). As Table  below demonstrates, understanding the challenges within the policy system itself, and the relationships between levels of the health and social care system within a policy environment that emphasises empowerment and localism, is critical for successful boundary spanning to deliver care. Studies of co-production are often 'bottom-up' in focus (for example, co-producing the personalisation of care within specific settings). Yet the research underpinning this article has shown how co-production is shaped by those leading health and social care integration and how cultures, systems and national policy agendas shape the opportunities for co-production to happen within and across governance systems.  Awareness of co-production The interviewees, both at a meso and macro-level, demonstrated an awareness of co-production as having broad applicability, both in an intra and inter organisational sense, and in terms of user-engagement in codesign. The meanings of coproduction appeared to be shaped by specific professional backgrounds and other practices associated with conventional activities, such as community development and asset building.
With widespread use of the term co-production, potentially diluting its intentions, and given the impact of the pandemic on health and social care, we recommend applying a deliberative approach to bring together citizens, practitioners, local planners, third sector providers, policymakers and support staff from intermediaries, to revisit the principles of co-production, as expressed in the Public Bodies (Joint Working) (Scotland) Bill Policy Memorandum , as a way to renew interest and share power in co-production and underpin post-pandemic enhancement. Entrenched systematic obstacles The majority of interviewees highlighted a number of systemic factors that have a direct impact on the work being undertaken to integrate health and social care. Some of these are about local national relationships (i.e. national policy-makers understanding meso/micro conditions) and the need for sustainable funding models. The expectations regarding the need to advance health and social care integration in Scotland, using coproduction approaches, are out of step with the major extant complexities of the Scottish public sector, which were not considered sufficiently before the roll out of health and social care integration.
In parallel to reconsideration of a recommitment to coproduction for health and social care, we recommend building on the  Audit Scotland progress report, a mapping exercise that outlines and distinguishes the national systemic features from localised systemic issues in advance of the development of a contribution analysis framework e.g. the differing regulations governing the treatment of finance between the NHS and Local Government, and the interoperability of local IT systems; or the introduction of other national policy initiatives such as the educational regional planning frameworks that do not align with local community planning partnerships. Unfortunately, there has been a very high turnover of Chief Officers in recent years and an explanation for this is that the acute systematic challenges, and frustrations, relating to progressing integration referred to above are taking a tolllet alone nurturing coproductive practice.
Health and social care integration would benefit from a greater level of priority at a macro-level to enable health and social care partnerships to access evaluation/improvement support based on a 'decluttering' exercise of national agencies so that areas know where and how to access support. There needs to be a review of the model of public services in Scotland, including the establishment of a national care service which raises the profile of social care and is led by senior leaders from local government and the NHS which is also part of a wider review of how public services in Scotland can be better organised in line with post-Covid recovery reviews. Chief Officers need be part of a formalised national governance agency (such as a care service) and have a space for continuous leadership and development training in partnership with academic institutions in Scotland.

Politics for public services reform
A key barrier to sustaining progress in health and social care, was political interference linked to making the case for Scottish independence that could be seen to be 'popular' whilst, at the same time, failing to make difficult decisions about public sector governance. This is in line with the recommendations of the Feely Report (Scottish Government, b), which calls for a national care service and systems change.
There is a need to build on the deliberative approach outlined above, review the mix of national infrastructure and resources with the aim of streamlining support for the purpose of evaluation and improvement.

Conclusions
This study provides novel insights into the health and social care integration system in Scotland. Both the health and social care partnership (meso) level and national (macro) level interviewees generally recognised the potential benefits of, and attached value to, co-production as a means of supporting service improvements. A policy disparity has arisen in Scotland around parallel normative aims, empowerment and accountability. Those working in partnerships struggle to navigate through often conflicting policy agendas and ambiguous policy narratives for change (e.g. improvement, localism, and empowerment).
One of the key barriers to sustaining co-production, highlighted in both local and national interviews, was the 'cluttered landscape' of national improvement agencies in Scotland. This is problematic for national agencies themselves given that it is difficult for them to understand their impacts and contributions to national-level outcomes when their support efforts are intended to also be co-productive. This cluttering has also been challenging for local areas, as it creates confusion as to where to find the appropriate information and supportparticularly with regards to co-production, improvement, and evaluation. The findings did not suggest there were too many agencies, rather, there was a need to align the work of these agencies around their distinctive contributions and to have this alignment reflected in the policy and governance levers to support cross-fertilisation, consistent policy messaging and support. This research suggests that if the Scottish Government is to continue to undertake policy signalling to promote co-production as key to the change health and social case integration aspires to, then operational guidance, training and support should be provided to partnerships on its practical application and evaluation, while being sensitive to local contexts. The scope for delivering co-production also needs to be seen in the context of meso and macro governance relations. As one interviewee noted, 'the biggest difficultly you're going to have is that in order for this to work, you're gonna have to see a manifestation of leadership in the public sector which is co-dependent. Which has people sharing power' (L). Evaluating co-productive approaches within integrated service areas will be important to understand their value, but national actors need to take responsibility for being co-productive partners with meso-level leaders in order to build capacities within the system itself and improve multi-level relationships.
The study provides key lessons for other state contexts, particularly given the patterns of European governance in areas of social policy, which reflect a move towards increasing empowerment, decentralisation and disaggregation (Connolly et al., b). Furthermore, the COVID- pandemic has substantially tested the resilience and responsiveness of health and social systems worldwide. These systems are now likely to be operating under increasingly stringent conditions with reduced public finances, post-pandemic demands and increased   ,  .
expectations. As public sector entities they will have to reflect again on how to better move forward in such a dynamic landscape. However, in Scotland, the high prevalence of COVID- in care homes has been a marked failure of governmental crisis management (Public Health Scotland, ). Such a situation will be the result, in no small part, due to the systemic and cultural challenges within Scotland's health and social care system, which should feature as part of a future public inquiry.
Within the current social, policy and political context, there are a number of capacity challenges to sustaining co-production in health and social care. These capacity challenges have implications for national public service leadership and have essentially produced capacity gaps at both levelsi.e. at both meso and macro levels. The cumulative result of this is that meso and macro relations have been burdened by the multiple policy agendas and challenges. The fact that integration was one of SNP Government's flagship policy initiatives is an overriding political driver for its continuation. Yet there remain fundamental institutional and cultural issues that need to be addressed. Given this, it could be argued that policy expectations about the public sector in Scotland being institutionally fit to adopt integration were, and remain, problematic.