Understanding and supporting change in health systems using the strategic action fields framework: the availability and origin of sources of authority

ABSTRACT

Policy-driven change in public service systems is difficult to implement. We focus upon ‘sources of authority’, which figure within the Strategic Action Fields Framework (SAFF) as resources mobilized by socially skilled actors to catalyse and sustain change. Drawing on qualitative case studies evaluating a large-scale policy-driven change in the English NHS, the Vanguard New Care Models programme, we (1) expand the definitions of ‘sources of authority’ and (2) argue that ‘sources of authority’ vary across two intersecting dimensions. This expands the utility of the SAFF beyond understanding and evaluating policy impact, and offers practical opportunities for achieving change.

Video Abstract

Read the transcript

Watch the video on Vimeo

© 2024 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group

KEYWORDS:

• Strategic action fields framework
• policy implementation
• interviews
• sources of authority
• public policy

Introduction

Change in public sector organizations is complex, difficult to manage and may well generate unintended consequences (Best et al. 2012). In this paper, we take an existing framework designed to support the evaluation of policy-driven change and use evidence from a case study in the English NHS to further develop and refine it. In particular, we focus upon the ‘sources of authority’ (Moulton and Sandfort 2017) that local actors charged with initiating and maintaining change can mobilize and provide a typology designed to support their proactive use by public sector managers.

‘Policy-driven change’ is change introduced by local, regional or national bodies with legitimate public-sector policy-making authority in a particular geographical area or over particular types of services (Boden et al. 2004). This definition would include change such as a smoking ban introduced by a local government, but would exclude a smoking ban introduced by individual businesses. In exploring such large scale policy-driven change, in addition to understanding the micro-level processes by which change occurs (May and Finch 2009) it is also important to take account of the meso and macro-level contexts within which the inherent contestability of policy intersects with questions of resource allocation, political will and power (Campos and Reich 2019; Manna and Moffitt 2019).

The Strategic Action Fields theory was initially developed by Fligstein and McAdam (2011) and builds on the concept of ‘fields’ from organizational sociology (Martin 2003). A Strategic Action Field is a ‘meso-level social order where actors (individual or collective) interact under a set of common understandings about the purposes of the field, the relationships in the field (including who has power and why), and the field’s rules’ (Fligstein and McAdam 2011, 3). Strategic Action Fields help to make sense of the interactions, conflicts, and collaborations in various spheres of social and economic activity, and have been applied to study higher education institutions (e.g. B. J. Taylor 2016), urban citizenship and movement (e.g. Anghel and Alexandrescu 2023; Domaradzka 2017), industrial development (Normann, Vasstrøm, and Garmann Johnsen 2022), public sector policy implementation (Smith, Girth, and Hutzel 2021) and other topics. Notable attempts have been made to expand and adapt the Strategic Action Field theories; for instance, Parker and Corte (2017) combine Collaborative Circles theory and Strategic Action Fields to explain differences in creative potentials between small groups. One such adaptation, combining ‘Strategic Action Fields’ (Fligstein and McAdam 2011) with public administration research (Ostrom 2011), alongside sociological theories of social skill (Fligstein 2001) was developed by Moulton and Sandfort (2017). It offers a multi-level, multi-actor framework for exploring the complexities of policy implementation.

In our framework, implementation actors – be they policymakers, public managers, service providers, funders, or direct service staff – work within bounded social settings. They employ social skill to interpret and adjust a public service intervention in ways that build common understanding and reconcile competing sources of authority to enable collective action. Different social dynamics across Strategic Action Fields lead to variations in how interventions are brought into practice, which, in turn, can contribute to variations in the outcomes in the system and for the target population (Moulton and Sandfort 2017, 145)

Importantly, the resulting ‘Strategic Action Fields Framework’ (SAFF) conceptualizes the wider political and policy context as dynamic and changing rather than as a static backdrop facilitating or impeding action (Cane, O’Connor, and Michie 2012), and provides a means to account for variations between settings. In studying public service interventions, Moulton and Sandfort (2017, 148) argue that the first task is to define the intervention, including processes of change (and associated assumed causal logics), methods of co-ordination, and the desired impact on outcomes. The framework then identifies three categories of ‘drivers of change or stability’ (148). These are as follows: sources of authority; social skills; and exogenous shocks. They argue that socially skilled actors interpret and wield sources of authority and respond to exogenous shocks (i.e. unpredictable or unexpected events external to the organization), using these to motivate local action (Moulton and Sandfort 2017, 154). Of the three elements, socially skilled actors have received the most attention (e.g. Biazzin, Sacomano Neto, and Candido 2020). For instance, our previous work shows that the ability to exercise social skills and enact change depends not only on possession of social skills but also their interplay with the actor’s position, personal characteristics, behaviours, fixed personal attributes and local context (Coleman et al. 2022, 965).

An important element of the SAFF that has been relatively under-explored in the literature is that of ‘sources of authority’ (discussed below). Both the SAFF and sources of authority tend to be used retrospectively rather than pre-emptively, with a focus on understanding what has happened (Høiland and Willumsen 2016; Stanczyk et al. 2018), rather than proactively seeking to support socially skilled actors in implementing change. In our study, we found that some sources of authority are easier to mobilize than others, and that some important sources of authority may not be under the direct control of those at the local level.

The aim of this paper is to use these insights to consider sources of authority in more detail. Drawing upon the findings from a national evaluation of a large scale change programme in the NHS in England, we apply the SAFF and propose a new approach to classifying available sources of authority to facilitate their mobilization.

What follows is divided into five sections. An initial background section briefly describes sources of authority as part of the SAFF. This is followed by a description of the research setting, the Vanguard New Care Models, and an account of our methods. In the results section, we briefly summarize the sources of authority that we saw in our research sites, before proposing an approach to classifying sources of authority which may make them more easily mobilized by those involved. A final discussion reflects upon the contribution of this paper in the broader context of policy implementation studies.

Sources of authority in the strategic action fields framework

While the SAFF is characterized by three categories of ‘drivers of change or stability’ - sources of authority; social skills; and exogenous shocks (Moulton and Sandfort 2017, 148) – in this paper, our focus is upon sources of authority. There are multiple sources of authority within any field, which can be complementary or conflict. Sources of authority from different fields also interact and guide behaviour. Drawing upon the work of Ostrom (2011), Moulton and Sandfort (2017) highlight the polycentric complexity present in most institutional contexts, within which overlapping systems of rules and norms will operate.

Sources of authority do not function independently within Strategic Action Frields; socially skilled actors draw upon, or respond to a variety of sources of authority in order to initiate, validate and maintain change programmes. Moulton and Sandfort (2017) identify four broad, categories of authority, set out in Table 1:

Table 1. Sources of authority(Moulton and Sandfort 2017).

Type of authority	Definition
Political	‘Political authority defines what is legally permissible, establishes mandated processes, stipulates which public institutions are vested with legal responsibilities, and delineates hierarchical governance relationships between fields’
Economic	‘Within implementation systems, economic authority is often quite visible, creating incentives tied to specific processes or outcomes. Yet, signals often conflict, and consequences are not inevitable but shaped by actors within a particular context’
Beliefs and Values	‘Shared beliefs and values reduce uncertainty, operating as a cognitive framework that provides a means of sensemaking, often in light of what is understood about the past (Khademian 2002; Weick 1995; Yanow 1996). For example, moral categorizations of both workers and the target population may shape how an intervention is implemented (Mettler 2007; Schneider and Ingram 1997; Soss 2005). Shared beliefs about the efficacy of an intervention may shape whether or not actors integrate new activities into daily practices (Damschroder et al. 2009; May and Finch 2009)’
Norms	‘Norms associated with professional expertise may be a particularly potent form of authority in particular fields. Expressed through professional codes, evidence-based practices, or industry standards, norms can create binding expectations of what needs to be considered during implementation activities. They are demonstrated when powerful field actors issue white papers or offer training programmes’

Sources of authority figure in implementation as rationales to act by being mobilized in particular ways in particular settings. Exactly which are mobilized in what way will depend upon the characteristics of the field. Sources of authority are linked to the concept of legitimacy but are conceptually distinct. Legitimacy refers to the perceptions of those in receipt of implementation efforts (Huy, Corley, and Kraatz 2014), with particular actions by local leaders seen as more or less legitimate. This legitimacy will in part depend upon the mobilization of sources of authority but will also be influenced by other factors such as the social position of the leaders or their previous actions (Coleman et al. 2022). Importantly, sources of authority may conflict, and individuals will make choices over which to mobilize in particular circumstances. For example, in their study of the entry of wind power into energy markets, Stafford and Wilson (2016) explore how different sources of authority were mobilized as wind power was added to US energy markets, highlighting the fact that whilst political authority was clearly behind the policy, implementation also required complex social processes to mobilize behavioural norms and cultural practices. Also in the renewable energy field, Lenhart (2017) focuses upon the creation and creative use of different sources of authority.

In their description of the SAFF, Moulton and Sandfort (2017) suggest three main roles that the framework might serve (1) to explore and explain the variation inherent in policy implementation, elucidating why the same policy might have different impacts in apparently similar contexts; (2) to support exploration of system capacity to change; and (3) to provide a structure within which to interrogate the concept of ‘social skill’. Whilst previous research has shown the SAFF to be a useful approach to the first of these topics (for example, Chen 2018; Stafford and Wilson 2016; R. Taylor, Rees, and Damm 2016), less has been written about its value with regard to a proactive exploration of system capacity to change.

The Vanguard New Care Models programme

The National Health Service (NHS) in England provides fertile ground for public administration research; it is a mature service, characterized by hierarchical relationships as well as a multiplicity of social orders or fields operating at different levels (Fligstein and McAdam 2011). The Vanguard programme in England, like many such integration programmes that seek to better-align health and social care without merging the two systems (Morciano et al. 2021; Stokes, Checkland, and Kristensen 2016), arose out of concerns that the NHS was ill-designed to meet the needs of an ageing population. Lack of integration between sectors was identified as inhibiting the development of high-quality care, and it was argued that rather than structural reorganization or mandated change:

… the national leadership of the NHS will need to act coherently together, and provide meaningful local flexibility in the way payment rules, regulatory requirements and other mechanisms are applied. We will back diverse solutions and local leadership… (NHS England 2014, 4)

Local statutory organizations were encouraged to work together and volunteer as pilots to test and refine ‘new models of care’ which would identify ‘standard approaches and products’ (NHS England 2015) which could then be rolled out more widely (Checkland et al. 2019, 2021). NHS England, as primary policymaker, identified five types of ‘new care models’ (Hammond et al. 2019). Groups of organizations within local health economies were invited to apply to become pilots, and 50 successful applicants were designated as ‘Vanguards’. We focus upon three of the models: Primary and Acute Care systems (PACS), Multi-specialty Community Providers (MCPs) and Enhanced health in Care Homes (ECH) – a total of 29 sites, all of which focused upon improved population health management and increasing cross-sectoral collaboration to improve care and keep people out of hospital. Table 2 sets out their initial descriptions and funding received.

Table 2. Vanguard types and funding.

Vanguard Type	Date	Number of Vanguard sites	Description	Funding (£million)
Primary and Acute Care systems (PACS)	March 2015	9	Joining up GP, hospital, community and mental health services to improve the physical, mental, social health and wellbeing of the local population. Population-based care model based on the GP registered list.	130
Multi-specialty Community Providers (MCPs)	March 2015	14	Moving specialist care out of hospitals into the community. Working to develop population based health and social care. Population-based care model based on the GP registered list.	124
Enhanced health in Care Homes (ECH)	March 2015	6	Offering older people better, joined up health, care and rehabilitation services. Care homes working closely with the NHS, Local authorities, the voluntary sector, carers and families to optimize health of their residents.	18
Urgent and Emergency Care networks (UECs)	July 2015	8	New approaches to improve the coordination of services and reduce pressure on A&E departments	13
Acute Care Collaboratives (ACCs).	Sept 2015	13	Linking local hospitals together to improve their clinical and financial viability, reducing variation in care and efficiency	72

The programme was well funded, with individual sites receiving a share in £329 million additional funding. A national support and evaluation programme received a further £60 million to establish a broad support programme to engage with sites as they tackled many of the issues known to be associated with integration initiatives requiring action across organizational and sector boundaries. This included such topics as follows: information sharing and governance; contractual models to support cross-sector working; leadership; governance and accountability; and stakeholder engagement. Volunteer sites were required to detail their plans for new models of integrated care, with a focus on breaking down barriers between sectors. Cutting across the market-based modes of governance embodied within the existing legislative framework (Checkland et al. 2021), the programme encouraged ambitious new approaches to care delivery with local leaders required to work together in new ways and mobilize resources to change service design.

In this paper, we draw upon case studies of these initiatives to explore how this was accomplished, using Moulton and Sandfort’s (2017) SAFF to frame our findings. We focus specifically upon the ‘sources of authority’ drawn upon in our case study sites to initiate and sustain change. Before our typology is introduced, we first expand the number of sources of authority and amend their definitions in line with the specific nature of a bottom-up, cross-sector collaboration programme.

Methods

In summer 2018, the research team selected six case study sites from the 50 New Care Models Vanguards that operated in England from 2015 to 2018. The six sites were comprised of two Multispecialty Community Providers (MCP) Vanguards (with Vanguard partner organizations covering populations of 170k to 320k people), two Integrated Primary and Acute Care Systems (PACS) Vanguards (covering populations of 200k and 370k each) and two Enhanced health in Care Homes (ECH) Vanguards (covering populations of 190k and 360k each). Two of the selected case study sites were in the North of England, two in the Midlands region, and two in the South of England to capture a geographical spread. The distribution of case study sites (two per each Vanguard type) allowed an in-depth study of each site while also enabling the exploration of distinct approaches, opportunities and challenges faced by different Vanguard sites of the same type.

A qualitative case study approach was adopted, involving 52 semi-structured interviews (1–2 participants each) and 4 focus groups (3–7 participants each). Interviews were conducted face-to-face or by telephone, while focus groups were conducted face-to-face and facilitated by at least one researcher. A total of 80 respondents participated at these Vanguard sites, including current and past representatives from Clinical Commissioning Groups (CCG) provider organizations, local authorities, voluntary sector organizations, Vanguard programme leads, frontline staff and patient and public contributors (see Table 3 for further breakdown).

Table 3. Case study respondents.

Respondent type	Numbers interviewed	Number in focus group	Total
NHS employees (current or past)	48	14	61*
Local Authority employees	4	1	5
Private/Community/Charity sector employees	9	-	9
Public contributor	1	4	5

*1 NHS employee participated in both an interview and a focus group.

Semi-structured interview and focus group schedules were developed separately for each respondent group. Broadly, the topics focussed on factors that facilitated or hindered Vanguard development, examples of Vanguard ‘success’, Vanguard interactions with the national support programme, other Vanguard sites and other integration initiatives, as well as scale and spread (MacInnes et al. 2023).

Interviews and focus groups were audio-recorded and transcribed verbatim, followed by a thematic analysis using a coding schedule based on the SAFF and findings from our previous work on Vanguards (Checkland et al. 2021). Themes included Vanguard established, their substance (services, change over time, scale etc.); process of implementation; funding; performance; next steps and the SAF framework (sources of authority, social skills, and exogenous shocks). This paper relates solely to the ‘Sources of Authority’ coding theme. Initially, transcripts were coded using sources of authority described by Moulton and Sandfort (2017) as sub-codes (see Table 1). A subsequent cross-case analysis highlighted the dimensionality of sources of authority, with different sources impacting in different ways. Further analysis proceeded via sub-coding of sources of authority, using the four quadrants of these intersecting dimensions (see Figure 1).

Figure 1. Dimensionality of sources of authority with code frequency.

To preserve anonymity, the name of the case study site and the respondent was given a unique identification number. For example, S6R05 related to site 6, respondent 5. Ethics approval was obtained from the University of Manchester (Approval Number: 2018–4359–6573).

Results

In line with Moulton and Sandfort (2017), our analysis found that socially skilled actors in case study sites mobilized political, economic, normative and belief-based or value-based sources of authority. In addition to these, we found a new category, another distinct source of authority, namely, ‘data and evidence’. Table 4 sets out how the five identified sources of authority operated in case study Vanguard sites and suggests what adjustment may be needed to accommodate the specific nature of a bottom-up, cross-sector collaboration programme.

Table 4. Sources of authority observed in case study sites.

Source of Authority	Operation in Vanguard case study sites	Illustrative quotes	Proposed adjustments to definition
Political authority	The Vanguard programme itself carried with it some formal rules and responsibilities (e.g. requirement to meet performance targets in order to receive funding past the first year, and to engage with national processes of learning and engagement. Political authority was generally less vested in formal processes, and more about the status associated with the Vanguard ‘badge’ or label. ‘Being a Vanguard’ at times provided licence to override or supersede existing formal statutory and regulatory requirements (e.g. mandated competition between providers which requires competition and competitive tendering in the NHS (Osipovic et al. 2019) and individual organizational regulation. Local areas felt themselves, at times, to be outside the ‘constraints’ of normal rules.	‘It forces change. You can’t be a Vanguard and say, right, great, we’ve got the cash, now we’re going to [sit around]. It doesn’t work like that’ (S2 R12) ‘So if we talk about [the] Care Quality Commission, then [CQC chief inspector] was interested, and [local leader] would particularly meet with [CQC Chief inspector] and he’d come down and he’d use it as a test case for, how would you develop CQC for multi-organisation developments’ (S1 R06)	As well as being vested in formal processes, political authority can also be drawn from the status or label of being part of a high profile and well-funded policy initiative. Political authority mobilized via the status of the programme can then be used to challenge or seek to change formal (legal and regulatory) processes.
Economic authority	Current and impending financial pressures were used across our sites as a source of authority to galvanize action. Local leaders used financial projections to demonstrate to stakeholders that lack of change would result in a difficult financial situation. Not engaging with change was presented as unsustainable and therefore unthinkable, rendering opposition to the programme difficult. Vanguards were also the recipients of significant amounts of money at the time of very little discretionary spending possible in the NHS. In this context, the amount of money allocated to the programme provided a powerful signal which generated high-level engagement.	‘[T]he trigger was really – it’s the burning platform idea – […] we knew that within a few years the demand problem, plus the amount of funding, was going to cause us actually quite a major issue’ (S2R016) ‘So what we found in the beginning we had the very senior people come along, I think everybody thought, oh, there’s lots of money, so we even had the chief exec of the local hospital came to the first meetings and to be fair, [they have] still played a part’ (S3 R01)	Economic authority can be both positive (additional funding provided) or negative (dramatic vision of a ‘burning platform’ used to galvanize activity).
Values and beliefs as authority	The Vanguard programme was very successful in inculcating in its participants a moral sense of doing good. Across all our sites those involved told us that the programme chimed with their sense of public service, and this was, itself, a strong rationale to act. The programme was (initially) permissive, with minimal national direction. Many of our sites used this freedom to develop a bottom-up, locally driven programme, and this provided a sense of values-driven change, with a shared vision. Interviewees also identified that a shared vision (often predating the Vanguard initiative) allowed them to set Vanguard scope and test the fit between initiatives at the local/micro level and national policy/macro level.	‘So I think you need to believe in it yourself, but you also need to have those good relationships, so when it is difficult people stick with it. And I think you also need to be in a team where people support each other, so when it’s difficult, when you all believe in it together and it’s doing that up and down that keeps you going as well when you believe it’s the right thing’ (S1 R04) ‘[W]hile there was some funding, there was also a lot of goodwill, and that people drawn in, who became part of it, everyone had this shared vision and focus and wanting to make a difference and make a change’ (S3 R01) ‘If you’ve got a vision, you can always come back to it, if you’re having a bit of a wobble, or if you’re testing an initiative, you know, is this initiative appropriate, is it within scope? Well, if you come back to that vision, if it satisfied those three criteria, then it probably is in scope’ (S1R01)	To aid policy implementation, authority can be mobilized both from existing values, beliefs and a shared vision that go beyond the new initiative, and by generating new beliefs and values which provide a moral imperative to act.
Norms as authority	The essence of the Vanguard programme was in integration across professional and sector boundaries. Professional/organizational norms were therefore a potential source of resistance rather than being facilitative. To overcome this, our sites developed a new norm of inter-disciplinary collaboration, in which individual professionals were pushed to move beyond their usual ways of thinking. Normative and economic sources of authority often intersected or were in conflict in the way collaborative working and inter-organizational partnership was achieved. In many cases, existing economic factors created barriers to collaboration and encouraged inter-organization competition. Where the new norms of collaboration were particularly strong, this helped to deal with economic barriers.	‘MCP is much broader […], you’re looking at the voluntary, and the community sector, you’re looking at social care. Which, as a doctor, whether you’re a hospital doctor, or a GP, that’s not what you’re actually trained in. So you’re actually retraining yourself to think holistically about a person, a family, a society. Where’s, you were trained, actually, in looking at diseases’ (S1 R01) ‘A lot of that comes down to the way we’ve been commissioned in the past, because there’s lots of brick walls been put up between organisations. They don’t want to share their intellectual property with one another because of the tendering process and probably being destabilised as an organisation because they’ll lose tenders’ (S1R03) ‘When we didn’t get all the money we wanted, we built such a rapport and such a vision that [named charity] funded our [local initiative] out of their money’ (S1R06)	Norms as a source of authority go beyond professional ones, such as ‘best practice’ to make changes. New norms, that move beyond existing professional ones (such as a new norm of inter-professional collaboration) can also be mobilized.
Data/evidence as authority	Data/evidence provided a rationale for people to engage and were a powerful tool for those ‘selling’ the programme. Respondents described presentations and documents in which the case for change was made on the basis of what were sometimes acknowledged to be ‘unsophisticated’ analyses of activity and outcomes. Despite few sites being able to provide robust evidence on savings or improved quality care, this was rarely questioned or detracted from the authority of being able to recite figures which demonstrated progress. Since in a networked and decentred system, resources do not necessarily flow according to pre-determined allowances or plans and instead must be competed for, the ability to show apparently concrete improvements using data – and do so in a short timescale – was a useful source of authority in this context.	‘I wouldn’t suggest for a minute that it was sophisticated population health analytics, but we said, this is our population, these are the diseases they’re suffering from and if we apply the current model of care to that population taking account of the demographics, then the gap […] it was a bit of a straight line, which I’m sure it never is in real life, but we said we will have a gap of £150,000,000’ (S5 R02) ‘So, one of the pieces of work that I was involved in was to look at the impact of the multidisciplinary teams […]I think the business case that came out of the back-end of it was a little bit questionable but it claimed a certain investment per person that I think was speculative at best, and the cynic in me could almost imagine the audience in which those numbers were designed. I think there was a bit of sales going on there, you know?’ (S2 R12)	Ability to generate and refer to data/evidence (even if it lacks robustness) as a way of demonstrating progress and/or impact can operate as a source of authority and aid implementation.

Cross-case analysis led us to reflect upon the different ways in which each could be mobilized, and the differing extents to which they were amenable to local control or change. We propose that sources of authority can be organized along two intersecting dimensions (or four quadrants resulting from this intersection): their internal versus external origin; and the extent to which they are ready to be mobilized (see Figure 1).

‘Internally-driven’ sources of authority are present within the strategic action field in which they operate. ‘Externally driven’ sources of authority are not necessarily ones imposed top-down; they come from any level of strategic action field – some driven by national policy, some driven by regional actors and some arising directly from local implementers. The shared characteristic of externally driven sources of authority is that they come from outside the strategic action field within which the socially skilled actor is operating, and are present whether socially skilled actors wish or not.

The second dimension of sources of authority is their availability. Readily available sources of authority can be mobilized right away, for example, by employing a rhetoric that refers to existing norms that can facilitate policy change. On the other end of this continuum is authority which cannot be mobilized without first creating something new or dismantling something that already exists (e.g. beliefs that the socially skilled actor deems unhelpful). Both building and dismantling may be needed simultaneously, and the amount of resource required to achieve this can vary significantly.

Our proposed typology should not be interpreted as a set of discrete categories; instead, both intersecting dimensions are continuums along which sources of authority can be mapped (see Figure 1). Likewise, the inside/outside, internal/external or local/national position of specific sources of authority can and should be contested, along with appreciation for their temporally dynamic nature and potential to shift during the implementation process.

We now present examples of these different categories. As shown in Figure 1, each type of authority (political, economic, values and beliefs, norms and data/evidence) was present in all four quadrants of the proposed typology. The distribution of codes per source of authority varied somewhat between the quadrants; to ascertain whether these patterns are characteristic of the typology (e.g. that economic authority is more likely to be externally driven and readily available) requires testing in further studies.

Internally driven and readily available sources of authority

Internally driven, readily available sources of authority are the most advantageous to the skilled actors. Values and beliefs often formed and maintained over a long period of time, often provide sources of authority in this grouping/quadrant. For example, respondents in most case study sites described that a wide-spread existing belief that health and social care services are facing a crisis acted to delegitimise dissent – doing nothing was presented as not being an option. This facilitated fast-pace change and encouraged rapid implementation of Vanguard initiatives:

I don’t mean it flippantly, but I almost think ‘hallelujah for a crisis’. Because if you don’t get to the stage where there is no other way out, it drives you to change. So we’d got to that burning platform with, you know, lack of workforce, GPs, some of the practices falling over, closing, to do something. So sometimes you need that. And it actually has been for the good (S1R03; Values and Beliefs as authority).

Evidence and locally produced data formed an important associated source of authority in this category. Socially skilled managers were able to use available data analyses as a means of underlining the sense of crisis by, for example, starkly demonstrating future shortfalls in funding. Audiences unable to understand the detail underlying the figures would nevertheless be impressed and motivated by the story being told.

Skilled actor ability to identify locally-existing sources of authority and mobilize these was also crucial in helping embed Vanguards within organizational cultures and aid sustainability beyond the funding period. One such example is skilled actors drawing on the existing, readily available normative authority of collaborative working and ‘doing things differently’ within their Vanguard. The start of the quote aptly captures that this source of authority is internal:

[D]espite the Vanguard, there is always a [local] realisation that we needed to do different things, there was always a willingness to change our approach and our culture in [area] to move away from organisational working to system working and everyone coming together to deliver the best benefits for [county] (S5R10; Norms as authority).

Nonetheless, while existing or readily available sources of authority can be advantageous if identified early, they can also pose a challenge. As discussed earlier, sources of authority often overlap and can clash. Where a clash occurs, existing sources of authority are notably harder to shift or counteract than sources of authority generated through new initiatives. For example, existing economic rules required individual organizations to break even, whilst becoming a Vanguard meant letting go of individual organizational self-interest. As described by a respondent, in Site 2 in this case, the economic authority inherent in current structures clashed with the norm of collaboration:

So if you’re the acute provider and you can see that the CCG is proposing to split your £500 million budget in half and transfer half of that budget into a new organisation called a multi-specialist care provider, if your accountant is not in the ear of the chief exec saying ‘woah, don’t we want a piece of that?’, then they’re not doing their job properly, are they? So of course, the acute wants to win that contract, irrespective of whether it will deliver fundamental change or not, because if it doesn’t, it’ll affect viability, it’ll affect jobs, it’ll affect the hospital as a going concern (S2R12; Economic authority).

Since internally driven and readily available sources of authority are easier and quicker to mobilize than other types, it is important that local skilled actors identify these sources early. Early identification of sources in this quadrant is also crucial when they may be detrimental to intended policy implementation, allowing the pre-emptive creation of new forms of authority to mitigate the issue.

Internally driven sources of authority, which require practical changes before they can be mobilized

Relying on existing sources of authority is rarely sufficient to ensure successful implementation of new policy. Our analysis has demonstrated that skilled actors often need to create practical changes in existing systems before a source of authority can be mobilized. The most common example of this across our case study sites was creation of cross-organization groups and meetings in order to demonstrate and create greater buy in, before mobilizing that buy in as authority:

[M]eetings, when they occurred, were a couple of hours, however […] the biggest benefit was […] in demonstrating that [CCG] had buy in from a wide selection of stakeholders […] These stakeholders are always attending meetings, you know, they come along, so this is the buy in we’ve got (S3R05; Political authority).

Similarly, respondents in Site 1 described a strategic selection of GP practices to lead on Vanguard initiatives. The prominence and power of these strategically selected GP practices then exerted political authority, placing other practices under pressure to agree and follow suit:

[Y]ou’re obviously going to want to put [Vanguard initiatives] into the practice where it’s going to benefit your partners the most […] and politically where shall we put this, otherwise we’re going to alienate everyone else who’s working with us, and where do patients need it the most […] when your driving force is a large, successful practice […] they can then just do it, because everyone will agree (S1R05; Political authority).

Meanwhile, at the start of Vanguard implementation, Site 2 created various opportunities for public engagement in order to mobilize value-based authority, aligning Vanguard initiatives with ‘the will of the people’:

What communities and patients have told us has really shaped what we’re doing. And I think that’s incredibly important because that comes back to your hearts and minds, doesn’t it? If this is what communities and patients are telling us they want it to be like, it’s hard to argue with. I think that’s really important (S2R15, Values and Beliefs as authority).

In contrast to the readily available sources of authority, creating new structures or adaptations in order to mobilize sources of authority was time-consuming and required a considerable amount of effort or resource by socially skilled local actors:

To get people together to, first of all, buy into the vision, then buy into the strategy, and then operationalise it, was, you know, it’s not a fast process (S1R01; Values and Beliefs as authority).

In addition, while skilled actors had some control over how and when to mobilize this type of authority, there was also a risk that resource-intensive changes would not result in desired outcomes and would not mobilize authority. For example, in Site 3 this respondent explains how their failure to generate local data demonstrating success killed off one initiative:

My project with primary care and nursing homes did not demonstrate the cost benefits, the good outcomes that were expected. But because they had done it in other areas and come up with that, with something a bit more tangible, the expectation was that we would go to the main part of the CCG and make them fund it. But of course the world doesn’t work like that and although there was national evidence, this is always the challenge isn’t it in the NHS, there was national evidence that that model works but we couldn’t prove it locally and therefore locally we couldn’t fund it. (S3R01; Data/Evidence as authority).

This emphasizes the need to carefully consider if adaptations needed to mobilize authority are achievable and if potential benefits outweigh the time and energy required.

Externally driven and readily available sources of authority

In contrast to internally driven sources of authority, externally driven ones exist outside of the decisions (and often control) of socially skilled actors.

Externally driven sources of authority can also be categorized into those readily available to mobilize, and those that required prior adaptations. In our case study sites, one example of external authority readily operating and available for local skilled actors to co-opt was the Five Year Forward View (FYFV)¹ (NHS England 2014). Our respondents described that FYFV in itself provided an impetus for action, which in turn facilitated joined-up working:

The model of care was actually very much what was in the Five Year Forward View, so the things that people said needed to happen, they predated the forward view, but they were very much aligned to it. (S5R01; Political authority).

Similarly, being awarded the status of a Vanguard provided normative authority to innovate.

It forces change. You can’t be a Vanguard and say, right, great, we’ve got the cash, now we’re going to sit [around]. It doesn’t work like that (S2R12; Political authority).

In fact, our respondents noted a loss of impetus to innovate once the Vanguard programme ended, showing the influence of mobilizing readily available, external authority of being a Vanguard site:

[A]lthough a lot of the stuff was embedded and we’re carrying it on, you notice the difference without the Vanguard being on now. [… W]e’ve embedded everything that was done with the Vanguard, so those things are carried on but, I think, with the Vanguard it’s just that opportunity to be a bit more innovative and that, probably, is not so prominent now. (S3R04; Norms as authority).

Beyond authority to innovate, being awarded a Vanguard status was described by respondents as disallowing dissent and disengagement, as well as ensuring rapid implementation:

[Vanguard status] did exactly what [Chief Accountable Officer] wanted it to do, it set a metronome, it set pace, it set an expectation of change, and it didn’t give anybody the right to turn their back and say, no, I’m not interested, and I don’t want to be part of that (S2R12; Political authority).

However, as the quote above demonstrates, externally driven, readily available sources of authority might not be straightforwardly advantageous to local skilled actors. This was particularly true for the external push of the Vanguard programme to make financial savings. While receiving continued funding provided socially skilled actors with economic authority to implement new initiatives, the conditions to secure funding – namely being asked to evidence reduced emergency admissions – also diverted attention from the main objectives of their New Model of Care including other, non-admission-related patient outcomes:

I mean there was at least some threat […] that if you don’t start reducing emergency admissions, you don’t get your Vanguard money. But, of course, that drives behaviour to do things that lead the line on the graphs to go down, and so some of the behaviour locally, was then focused around that, maybe productively, but also maybe unproductively as well. (S2R11; Economic authority).

Thus, it is important to acknowledge that while local actors can mobilize this type of authority with little effort, they also have little control when external authority clashes with local objectives and local authority.

One such example came from Site 3, where anticipated funding drove cross-organizational representation from the most senior managers at Vanguard meetings. Senior management representation, which acted to legitimize change in a hierarchical setting, waned once the funding obtained proved to be lower than expected. This demonstrates that external sources of authority can undo internal authority, with local socially skilled actors having little control:

So what we found in the beginning we had the very senior people come along, I think everybody thought, oh, there’s lots of money, so we even had the chief exec of the local hospital came to the first meetings and to be fair, he has still played a part. I think as the funding allocation became more aware and we worked that out and it wasn’t going to be millions, we got about half a million, really, then we got down to other people further down the tree. (S3R01; Economic authority).

If a different, locally driven source of authority had been mobilized to facilitate senior representation in conjunction with external economic authority, senior representation may have been retained. This underscores the importance of identifying external, readily available sources of authority early, anticipating unintended consequences, and developing local sources of authority that may, if necessary, outweigh external authority and avoid detrimental outcomes.

Externally driven sources of authority, which require adaptations to local context before they can be mobilized

The final quadrant of sources of authority is externally driven, but requires skilled actors to adapt or translate this authority to the local context. Sources of authority in this quadrant were somewhat rarer in our case studies and might be uncommon in general.

One such example of external authority requiring adaptation was present in relation to the aforementioned ‘license to innovate’. As explained by the interviewee below, in order to use external permission for innovation in a way that would benefit the Vanguard programme, it was important to first develop a shared cross-organizational understanding on what the site was aiming to achieve:

Against the background of this permissive environment of the development of the ICP we do need to be absolutely crystal clear about a shared vision because without that we will have anarchy (S1R14; Values and beliefs as authority).

The example above shows that adaptations or new structures may be needed before socially skilled actors can make-use of an identified external source of authority. Similarly, socially skilled actors are often tasked with ‘translating’ external authority in ways that fit local contexts. This may not require practical changes, like creating new groups or processes, but instead call for changes to the external narrative of authority to fit the local context. For example, while the FYFV was a readily available source of authority in most cases (as discussed above), one case study site needed first to control the narrative of privatization around FYFV before it could be used to mobilize authority:

[A]t a time that the media were influencing the Five Year Forward View narrative of being all about privatisation, that rallied people in [area] around how are you going to prove to us that it’s not going to be?. (S2R10; Data/evidence as authority).

Our interviews also demonstrated that adaptations were not merely a vehicle to mobilize a source of authority. In a small number of cases skilled actors identified an emerging external source of authority and were able to shape it as it was forming. Respondents in Site 1 did this by ensuring that their approaches to operating an MCP were visible to the national programme support team when they were drafting guidance for all Vanguard sites:

So there were a number of case studies that were written for NHS England. So [colleague] wrote one, [colleague] wrote one and I wrote one. […] And also as well as people [the national oversight team, as well as representatives of healthcare organisations outside of England] visiting and talking to them about what we were doing and us going to people and us having phone calls or sending things, we used to get quite a lot of email queries asking us for information, and because after a while we built up these packs, we would just send those out and refer people to those. (S1R04; Data/Evidence as authority).

Thus, while most of the time proactively influencing external authority by internally driven adaptations may not be possible, there are exceptional cases where skilled actors are able to anticipate external sources of authority as they are developing and shape them.

Altogether, sources of authority in this quadrant are characterized by local actors ‘translating’ within their field. While the process of mobilizing authority in this quadrant is resource-intensive, in select instances it also offers unique opportunities for local actors to shape external authority sources.

Discussion and conclusions

In this paper, we have explored the sources of authority mobilized within a particular context and shown how they can usefully be classified along two dimensions: their internal/external origin; and the ease with which they can be mobilized. In doing this, we aim to enhance the utility of the framework and suggest how it might be used to support the implementation of policy-driven change. In addition, we have identified an additional source of authority, data and evidence, and illustrated its use.

A number of implications flow from our analysis. Firstly, we have demonstrated that an extended typology of sources of authority enables SAFF to be better-applied to bottom-up, cross-sector collaborations such as the New Care Models Vanguard programme. Secondly, we suggest conceptualizing sources of authority along these intersecting dimensions can illuminate sources of authority that are needed for successful implementation, but which cannot be mobilized at present. For example, if establishing new norms of collaboration and joint working is identified as integral in operationalizing new initiatives, but current governance structures prevent information sharing across organizations, it may be paramount to enable information sharing as early as possible. Thirdly, considering the dimensions of sources of authority can also help anticipate conflict. Local socially skilled actors can consider which externally imposed sources of authority are likely to interplay with or even conflict with other sources of authority and thus require careful management. Lastly, we have shown that the resources required to perform such management of sources of authority will depend on which quadrants sources occupy. Once this is considered, informed choices can be made on whether mobilizing some sources of authority is worth the resources required.

Our work also extends the possibilities of how sources of authority can be used by multiple audiences, including socially skilled actors implementing policy, as well as evaluators or researchers of policy implementation.

For socially skilled actors, our analysis suggests the need for an initial ‘diagnostic’ phase in policy implementation situations. While respondent quotes above demonstrate that local socially skilled actors intuitively and often successfully mobilized all types of sources of authority, there was little evidence of overt and proactive assessment of which sources of authority can be best-utilized and how; what authority was the most readily available versus which may require additional resources; and where conflict is likely to arise. In keeping with the multi-level complexity identified by Moulton and Sandfort (2017), this analysis might take place at a number of levels. For example, national policy support teams such as those established within the Vanguard programme (Checkland et al. 2019) could, in the planning phase, proactively consider the external sources of authority that are associated with the programme (financial and other incentives as well as resources, prestige etc.) and ensure that these are both harmonious with one another and likely to act in the required manner at local level. At the same time, local teams bidding to join a pilot programme, or on the receiving end of nationally mandated change programmes, could proactively consider available local sources of authority and make efforts to align these with those arising externally (or, as evidenced above, even attempt to shape external authority). This also allows for the co-ordinated use of sources of authority by, for example, harnessing those which are internal and readily available to mobilize early enthusiasm, whilst awaiting those external ones which may take longer to materialize or conflict with the existing ones. Data and evidence, for example, may be readily available within a local context and could be used to create a receptive context for change. Proactive seeking out of internal sources of authority might also be of benefit if it is anticipated that those arising externally might be perceived as problematic in particular local circumstances. Our proposed way of considering sources of authority can also be of use in the early stages of implementation, when formative evaluation is required. Such early studies might aim to identify and classify available sources of authority in order to adjust their mobilization to optimize implementation. It may also offer an alternative way of capturing ‘success’ in the process of implementation, by evaluating if and how sources of authority were mobilized as intended and by whom. Moreover, a retrospective, summative analysis of implementation using this framework might support wider learning and the potential adjustment of incentives and implementation approaches for the future. It may also support policy adjustment as implementation proceeds. For example, where particular policies run counter to important local norms, policy adjustment may support better implementation.

Other frameworks or typologies have been used successfully to explore and explain how change is created within policy systems (e.g. Derwort, Jager, and Newig 2022; Frisch Aviram, Cohen, and Beeri 2020; Petchey, Williams, and Carter 2008) and numerous other typology-based tools are available to managers and change agents to facilitate strategic management (e.g. Bryson, George, and Seo 2022; Gürel and Tat 2017; Van Poeck, Læssøe, and Block 2017; Walker 2013); we do not propose ours to be necessarily more useful or advantageous. What we offer is a specific focus on the mobilization of sources of authority. With most policy implementation frameworks enabling retrospective analysis (Cairney and Heikkila 2014; Sabatier 2007), our typology can also be used in proactive planning and development of anticipatory implementation strategies. Few other typologies (see Buffardi, Pekkanen, and Smith 2017) offer applications that can create change within policy systems. Our paper does not, however, test this in practice – rather we show how it might be used by socially skilled actors responsible for policy-driven change. Further research is required to explore whether our proposed proactive use of sources of authority in this way is helpful.

Given the nature of the Vanguards and their implementation, most instances of externally driven authority in our data came from the national policy level. External sources of authority should not, however, be equated to ones solely in the policy field. Instead, external sources of authority should be considered as external to the strategic action field in which socially skilled actors operate and over which they have direct control. Ascertaining or even anticipating external sources of authority from other fields then can inform socially skilled actors in how best to mobilize and align internal sources of authority to external ones.

We are not, of course, suggesting that considering these aspects of available sources of authority provides a complete explanation for any given example of implementation, nor that paying attention to sources of authority in this way will alone support successful change. Furthermore, we do not suggest that contexts of policy implementation are static; even if all sources of authority are proactively assessed and anticipated, both the implementation landscape and available sources of authority will continue to change. Much like implementation itself and the wider political and policy contexts within which it occurs, proactive consideration of sources of authority should be considered as a dynamic and ongoing process rather than a one-off occurrence.

Acknowledgments

The authors are very grateful to the participants, who were generous with their time.

Disclosure statement

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

Additional information

Funding

This paper is based on independent research commissioned and funded by the National Institute for Health and Care Research (NIHR) Policy Research Programme “National evaluation of the Vanguard New Care Models Programme”, PR-R16-0516-22001.

Notes

1. FYFV (NHS England 2014) set out a vision for the future development of the NHS and suggested that Vanguards would set out to design, test and deliver a variety of scalable and replicable New Care Models (NCM) for the whole of England, with the expectation that success would be repeated elsewhere).