Distributed Leadership in Policy Formulation: A Sociomaterial Perspective

Towards a sociomaterial perspective on distributed leadership

Theories of leadership have emphasized characteristics, values and attitudes held by leaders – such as being charismatic, visionary, adaptable and risk takers (Avolio et al., 2009; Bass & Steidlmeier, 1999). In addition to explicating a number of leadership functions (see Vera & Crossan, 2004), scholars of leadership have also highlighted the importance of context (Martin et al., 2008), followers (Heifetz, 1994) and the relationship between leaders, their context and followers (Buchanan, Addicott, Fitzgerald, Ferlie & Baeza 2007; Spillane, 2005). In public service organizations the need for a distributed and collaborative style of leadership has been emphasized (Denis, Lamothe & Langley, 2001; Hartley & Allison 2000; Martin et al., 2008) as stakeholder groups are complex and historically based, and professional jurisdiction limits the ability of formal authority to influence practice (Abbott, 1988; Ferlie, Fitzgerald, Wood & Hawkins, 2005). Institutional structures stymie and shape leadership processes, forming ‘soft regulation’ (Levay & Waks, 2009; Currie, Lockett & Suhomlinova, 2009).

Distributed leadership has been described in diverse ways (Ancona, Malone, Orlikowski & Senge, 2007; Currie et al., 2011; Gronn, 2002; Martin et al., 2008). While definitional consensus does not exist (Avolio et al., 2009; Spillane, 2005), there is broad agreement that distributed leadership emphasizes inclusivity, collectiveness and collaboration. Ancona et al. (2007) describe a key characteristic of distributed leadership as the ability to acknowledge the strengths of others and identify where others are more suited or able to fulfil a role. Drawing on Weick (1995), Ancona et al. (2007) discuss the importance of sensemaking as a capability for leaders, highlighting how creativity and data-informed analyses enable an understanding of organizations and industry contexts. Other scholars (Bate, 2000; Currie & Suhomlinova, 2006; Levay & Waks, 2009; McGivern & Dopson, 2010) emphasize that, for professionals, the informal collective structures strongly influence how they behave and make sense of practice. A key tenet of conduct among medical professions is to respect the autonomy of an individual’s practice (Abbott, 1988; Freidson, 2001), which precludes new practices and policies from being promulgated simply through formal authority structures (Castel & Friedberg, 2010; Levay & Waks, 2009). This suggests that in contexts such as healthcare and professional service organizations, leaders need to recognize and appeal to the informal schemas, theories in use and hierarchies prevalent among professionals in order to facilitate sensemaking processes that will subtly engage colleagues. Rather than ‘followers’ being dependent on ‘leaders’, they can become genuine participants in the leadership process (Heifetz, 1994; Spillane, 2005) and help to (re)form policy.

Research to date has focused on examining the nature of distributed leadership; for example through characterizing ‘what’ is required. Yet there remains a poor understanding of ‘how’ and ‘why’ collaborative styles are enacted (Currie et al., 2011; Feldman et al., 2006; Martin et al., 2008). Spillane and colleagues (2001) provide some clues. They underscore the importance of considering ‘leadership practice’ as being distributed across an interconnected and interactive web of individuals, artefacts and situations. Building on Lave’s (1988) understanding of distributed cognition, they suggest that tools, language and social context are not merely ‘aids’ to leadership practice, but are fundamentally constitutive of it (Spillane et al., 2001). In education, they reveal how leadership ‘is stretched over’ numerous individuals, beyond the formal leader, including curriculum specialists, classroom teachers, protocols and vice-principals, and they go some way in acknowledging the materiality inherent in the enactment of leadership practice in everyday activities. They emphasize that leadership functions with, and in, multiple social and material entities; examining interactions as they unfold in situ in relation to specific tasks, such as curriculum design (Spillane et al., 2001). We build on this perspective, described by Mayrowetz (2008) as unique, to develop an understanding of distributed leadership in health policy formulation within a relational entanglement of assemblages of sociomaterial practices (Mol, 2002; Orlikowski, 2007; Orlikowski & Scott, 2008; Suchman, 2007).

Our perspective emphasizes the relationality of individuals, objects and structures which co-produce and enact leadership. Leadership enactment entails engaging with materiality – for example, offices, meeting rooms, desks, computers, reports, email distribution lists. These materials are not passive mediators or neutral channels for leadership but are consequential. It makes a difference if a leader hands out a written document of a policy vision or produces a digital animation of futuristic care on a big screen with surround sound; it matters if the pertinent databases on the internet are scanned for analytic materials, or if the email distribution list includes the heads of funding bodies and policy committees. We argue that multiple actors, data sheets, structures of accountability, specialized knowledge, technological resources, protocols and workshop rooms come together to enact leadership in formulating policy.

The mutual entanglement of diverse forms of materiality with the social in sociomaterial assemblages are performative of leadership and are the fundamental unit of agency. Agency is achieved through the relationality between multiple (social and material) entities as enacted in practices. We focus on ‘enactment’ to highlight how material objects achieve stability and coherence through use, rather than as discrete entities with inherent properties (Mol, 2002). This performative approach distinguishes material entities as being distributed – knowledge being propagated between social and material entities (Flor & Hutchins, 1991) – rather than focusing on leaders’ discrete strategic resources. Entities do not stand alone with predefined inherent properties, as their capabilities are only manifest in relation to specific practices and structured contexts (Orlikowski & Scott, forthcoming). Rather than emphasizing what resources are used, we highlight how action is resourced and knowing is achieved through strengthening associations (Feldman et al., 2006). We avoid defining a tool or resource as something that ‘is’ and instead foreground the practices by which tools and resources are enacted (Feldman & Orlikowski, 2011) as they achieve their accomplishments (sometimes unintended) only in use.

We extend the sociomaterial perspective by including a more explicit focus on historical institutional structures, which are prevalent in polity. While sociomaterial scholarship has been examined in the context of human robot interfaces (Suchman, 2007) and the use of specific tools and technologies (Orlikowski, 2007), the focus has been on situated activities with minimal incorporation of wider social structures. Our analysis takes a broader ‘cut’ across network elements to include the diverse structural interests and historical framing of policy groups and their constitutive role in sociomaterial entanglement and leadership.

We join others (Feldman & Orlikowski, 2011, Østerlund & Carlile, 2005) in taking a relational perspective in order to emphasize that structures, subjects, social groups and tools develop significance and enable phenomena through relations rather than independently; mutual constitution as opposed to discrete, isolated entities. We build on previous work examining leadership in public sector healthcare (Martin et al., 2008) and seek to explain how and why clinical professionals are able to lead in situ across a dispersed organizational field and how materiality is entangled in these leadership practices.

Leadership in health policy

Leadership in formulating policy is a unique challenge involving the engagement of different stakeholder groups, each of which is entangled in a unique historical and institutional context. Policy scholars emphasize the importance of developing coalitions as a way of mediating diverse interests (Feldman et al., 2006; Kingdon, 1995; Mechanic, 1996). A primary means of establishing coalitions is through engagement to develop commitment and alignment. Coalitions can facilitate an integrated approach to service goals that extend beyond current institutional boundaries (Bate, 2000; Martin et al., 2008) and their formation has been found to enable an emotional climate of hope (Wallis & Dollery, 2002) which fosters readiness for change (Elster, 1998). Furthermore, enrolling more people personally into the policy process implicates those who volunteer to participate (Smythe, 2005), which may further secure support for change.

Three critical stakeholder groups outside the immediate politics of government have been identified as critical to engage in the health policy coalitions: clinicians, lay public and executives. Gaining consensus from clinical groups is difficult even in matters clinical or scientific (Dopson, Miller, Dawson & Sutherland, 1999), and doubly difficult where political, managerial and financial issues intertwine with clinical and scientific judgements (Ferlie et al., 2005; Levay & Waks, 2009; Oborn, 2008). Important relational dynamics enacted across the medical community include competing emphases on strong collegiality, jurisdictional independence (Abbott, 1988; Freidson, 2001) and increasing requirements for adherence to protocols, published research studies and evolving technologies; these underpin interactions among peers. Coalition formation is further influenced by regional administrative arrangements, such as cancer networks, as well as by national-level bodies, such as the Royal Colleges, which offer collective and dissenting clinical visibility and voice.

Members of the public have a unique and increasing role in shaping policy. Accountability for services in welfare democracies has recast the public’s role from the right to services as determined by the state to the right to contribute in the definition of services provided through the state (Martin, 2008). Increasingly accountability to citizens is being seen as a direct and local dynamic, rather than a ‘vertical’ and bureaucratic process (North & Werko, 2002), even though in practice this ideal is difficult to achieve (Martin, 2008). Yet, ordinary public members are often disenfranchised from being able to contribute meaningfully to health reform (Mechanic, 1998; Taylor Gooby, 2006) given the inevitable power dynamics associated with healthcare systems and policy processes (Kingdon, 1995; Mechanic, 1996; Pollitt & Bouckaert, 2004). Clinicians who continue to hold high levels of public credibility (Calnan & Rowe, 2008) can have an important role in shaping and revealing public views. Furthermore, in the UK, previous disappointments with change processes that have been sidelined, stymied, or abandoned (Klein, 2006; Oborn, 2008) have contributed to a climate of cynicism and erosion of system-based trust in healthcare reform (Giddens, 1990) which clinicians, who continue to hold high levels of public credibility (Calnan & Rowe, 2008), can mediate.

A third stakeholder group of chief executives and senior managerial colleagues are identified as critical in the policy process in view of their authority, essential local organizational knowledge, control of resources and relationships and particular interest in local impact, for example, on demand and resources (Hill, 2003). Leadership engagement with this stakeholder group can productively focus on issues of organization design, incentives and competing policy as well as the integration of care across organizational structures. Leaders can be challenged to find better ways of articulating policy options and implications in order to facilitate the sensemaking process of the managerial cadre (Ancona et al., 2007; Weick, 1995).

Research methods

Data was collected as policy ethnography (Flynn, Williams & Pickard, 1996; Griffiths & Hughes, 2000). One of the authors gained extensive access to study the policy process of the London review of health provision which was led by an internationally renowned cancer surgeon based in a leading UK university medical school, the pseudonymized Professor Sir Dias Arego (hereinafter called Arego). She had access as an observer to all events in Arego’s diary, and regular contact within the policy team and stakeholder groups as policy change unfolded over an eight-month period. She took detailed notes at meetings, presentations, conferences and stakeholder engagement events and participated in informal encounters which elicited discussion, views and experiences of diverse people, including government ministers, London-based Members of Parliament, chief executives from a range of health organizations, medical directors, senior clinicians, hospital frontline staff and members of the public. Data was also collected at public engagement events which were run as a series of focus groups by a contracted company on behalf of the policy team. In addition, 250 members of the public contributed their views on ideas for reforming the health services in day-long workshops. As a method used in previous policy research (Bowie, Richardson & Sykes, 1995), public members in small focus groups of eight to ten were asked to comment on written and video scenarios of possible service delivery. In addition, 52 prearranged interviews were conducted, as summarized in Table 1.

OPEN IN VIEWER

Table 1.

Semi-structured interviews.

No.	Position
7	London CEOs
2	Public Health Director
5	Pathway chairs
2	Former NHS CEOs
1	Former NHS senior executive
2	Department of Health directors
6	SHA CXOs & senior managers
2	House of Lords member
3	Senior doctors
3	Members of voluntary sector
5	Patients/carers
10	Members of core policy team
4	Independent management consultants

Data was analysed using both inductive and deductive techniques in three stages. During the first stage of analysis, thematic summaries were developed to summarize the perspectives of the key stakeholder groups, namely, the core policy team working with Arego, clinicians, patient and public representatives and health managers. These narratives (Golden-Biddle & Locke, 1997) focused the analysis on how stakeholders groups were involved and the key themes of both policy process and content they identified as important. Legitimacy and trust emerged as cross-cutting themes among stakeholder groups, and were observed to be central in leading the engagement process. Interview transcripts, detailed ethnographic notes and transcripts of public engagement focus groups were then re-examined to clarify the role of legitimacy and trust (Denzin & Lincoln, 1998). In the final stage of analysis, insights from our relational perspective of distributed leadership were examined as critical events were scrutinized as sociomaterial assemblages; in particular, we were sensitized to how legitimacy and trust were sociomaterially performed in leadership practices. We acknowledge the limitation of our single case study approach; we do not aim to suggest that a similar leadership process would unfold in the same way in a different context or that the assemblages we reveal are stable. Our goal is for the rich insights enabled by the case to allow a form of theoretical/analytical generalization rather than statistical generalization.

The context: health services in London

Our setting is health care reform in England, where in recent decades there has been radical policy shifts including centralization of acute services, decentralization and networking of community services and moves towards integrated health and social care delivery (Pollitt & Bouckaert, 2004). Health policy development and its realization in changed working practices is particularly challenged by the complexity of stakeholder groups (Currie & Suhomlinova, 2006), inherited patterns of service delivery (McGivern & Dopson, 2010) and distribution of jurisdictional power (Abbott, 1988; Ferlie et al., 2005). These conditions have been exacerbated by the frequency of new policy initiatives (Klein, 2006), the erosion of public trust in politicians and the policy process (Strathern, 2000) and, perhaps in response, an increasing emphasis on audit and outcome monitoring (Power, 1997).

In July 2006 the Department of Health reorganized London’s NHS (National Health Service) structure by combining its five regional health authorities (with separate budgets) into one large regional unit (Strategic Health Authority – SHA), consuming one-ninth of the total NHS budget and governed by one executive team. London has a long history of autonomous hospitals and medical schools (Weisz, 2006) which had avoided coordinated city-wide integration. Hospitals located only a few miles apart, previously under different regional authorities, now had to work together to complement – rather than duplicate – services. Furthermore there were concerns about quality and productivity. The average length of hospital stay in London was considerably higher than the rest of the country and analyses of productivity and statistical outcomes compared to the rest of England showed London had the highest emergency department attendance rates, with many attendees presenting with needs which would be normally be met in primary care.

In autumn 2006, London’s health administrators commissioned A Framework for Action, a strategic policy review to provide directions of change over the next decade. It examined the number and types of services needed; for example, for the delivery of cancer services, the distribution of specialist centres and the role and composition of accident and emergency facilities. At stake was the range of services that would be provided in any one area within London. For example, should three hospitals in close proximity provide full cancer services, and how might these be coordinated with specialist cancer centres? Should mid-sized hospitals lose services and full emergency facilities, and some services (e.g. trauma) be centralized to specialist hospitals? How could access to services within ‘the community’ be improved?

Studies on the English NHS have highlighted significant managerial and government-based reforms since the 1980s, aimed at changing the work practices of clinicians and providing a backdrop of tension and strain between management and clinical communities. A unique feature in the current case was the sense that clinicians could regain a sense of control over the restructuring process; rather than having executives and government units dictating change, they came to feel they had an opportunity to be at the forefront of decision making, and that decisions would really result in change.

Research case

In 2006 the newly created London Strategic Health Authority asked Arego to develop a strategic policy for the region under financial assumptions for cost neutrality. London policy reviews in the preceding two decades (Tomlinson, 1992; Turnberg, 1998) had been led by medically trained individuals; but they were different to Arego in two important respects. They were not actively practising medicine at the time they were involved in policy formulation and they were directly accountable to the Minister of Health, who had commissioned the policy reforms. When appointed, Arego was known for robotic innovation and had no formal connection with the Department of Health or the civil service; he was fully employed as a celebrated surgeon and researcher within a prestigious academic health organization.

The development of the policy was characterized by an unprecedented level of engagement with three groups of stakeholders, namely, the clinical professions, healthcare management and lay members of the public. Six multidisciplinary clinical groups were formed, each of between twelve and fifteen members, chosen for their ‘track record in clinical leadership and outstanding patient service’. Each group examined care delivery in broad pathways, namely Staying Healthy (public health), Maternity and Birth, Acute Care, Planned Care, Chronic Disease Management and End of Life. The majority of members from each pathway had medical backgrounds, but other groups were represented, including healthcare managers, social services managers, allied health professionals and voluntary sector representatives. Groups met face-to-face for two or three half-day workshops in addition to exchanging documents and emails.

Four day-long public engagement workshops were organized by a private firm, where clinical pathway chairs and Arego sought to elicit public views and priorities and gain their incorporation into policy content. The Chancellor of the Exchequer, Gordon Brown, responsible for fiscal policy at the time, also joined one such event just prior to becoming the UK’s Prime Minister. Five half-day workshops were organized for health executives, in addition to numerous evening workshops. These included members from primary, secondary and specialist healthcare provider organizations. Senior managers of volunteer organizations were invited to a day-long engagement workshop and were also members in several multidisciplinary groups.

Engaging clinicians

Important in the policy process was the ability to develop a narrative of ‘real change’ where actual healthcare delivery would be influenced. This was both necessary in order to motivate substantial clinical engagement and only possible to deliver if clinical communities are centrally involved. The historical and institutional structure of medical work shapes leadership; highly autonomous individuals are responsible for the delivery of services and are accountable to registered supervisory clinical bodies for codes of ethics. Rather than being hurdles external to leadership activities, clinicians apprehend the importance of autonomy, collegiality and professional jurisdiction and become orientated to one another in relation to these structures.

You can’t run a hospital like an assembly line. … To get real change you have got to engage the professions. The professions make it work on the ground. … The professional groups will deliver if you put them in the right position and give them the capacity to create the changes. (NHS Clinician, affiliated with medical Deanery)

Given the numerous clinical subgroups involved, multidisciplinary pathway groups enabled dialogue between specialized medical communities. Rather than conceiving of individual specialist practices in isolation (for example, radiology and vascular surgery), they developed a sense of accountability to each other, and a need to relate one set of disease profiles, departmental protocols and work technologies with another. Sociomaterial entanglement across accepted disciplinary boundaries orientated the assemblage. For example, a member of the core team programmed an animated sketch that traced the pathway of a typical patient journey when consulting a doctor about severe stomach pains; the animation visualized the complexity of multiple trips to various locations (e.g. general practitioner (GP) offices, radiology outpatients, surgery clinics) on successive days that is needed to receive a simple diagnosis of gallstones. The technology – shown to pathway groups – externalized and concretized the representations of ideas and the co-dependence the clinical communities have on each other, organizational protocols, scheduling and jurisdictional boundaries. By relating specific protocols and technologies embedded within disciplinary practices to each other, this enabled trust to develop in terms of the policy suggestions being actionable and relevant; though trained to be sceptical and ‘dubious’, trust was nurtured as leaders drew on collegial norms that privileged autonomy yet inter-related details of specialist practices.

For example we [discussed] vascular surgery and intensive care procedure … a radiologist brought in stuff on how you do radiology reporting … so that it becomes a real time process […] we had a sticker session at the end of our first meeting. I was dubious about this at first. But actually it worked out quite well. (Pathway group chair, GP)

The process enabled the ownership of the policy content, and changes implied, by a broad and local base while (re)shaping the context. An important aspect of how leadership was enacted was to distribute connectivity beyond the current boundary norms, both in terms of practices constructed (such as new stroke pathways with increased integration between radiology, intensive care and neurology) and alignment of new coalitions. The Chief Executive of the King’s Fund ¹ highlighted the high level of clinical engagement in one of his speeches:

We know the status quo is unsustainable … but [change] will mean that we [may need to] abandon our little piece of the system … [Previously] so little effort has been made to involve clinicians … So I think [with this current review] there is a real chance to move things forward (July 2007)

The groups were asked to examine evidence from local, national and international information sources to support their proposals. Members of Arego’s core team provided their own independent data analytics; they worked late into the night in their hospital offices, scanning websites and contacting other researchers they knew who had access to datasets on local outcomes. The content of websites and outcome data mattered. Different data, different web-links and different search engines would produce different results. In the first few months of the review process, Arego and his core group met once or twice weekly – often early in the morning before hospital clinics – to secure accurate and convincing analyses, summarized on PowerPoint slides and short reports. Certain datasets were regarded as particularly important; for example, stroke performance data indicated the vast majority of current providers were not meeting accepted international standards. After trying diverse bar chart and pie chart formats, a slide was developed that vividly highlighted in red those breaching standards and in green those complying – red and green being symbolic of ‘stop’ and ‘go’. Though data was anonymized, the slide enabled knowledgeable clinicians to identify a majority of the organizations, stirring a sense of competition and shock at the low standards, thus further legitimizing the policy process. The production and use of these numbers, slides and reports appealed to medical science and tacit norms of technical rationality as a basis for ordering services. The slides, data and technologies were not merely functioning as ‘tools’– which implies stable inherent features controlled by medical leaders who possess the agency. Instead we emphasize how leadership was constituted among multiple actors: Arego, norms of autonomy and collegiality, technical rationality, local outcome data, research evidence, red and green PowerPoint slides; the whole assemblage was important in performing leadership.

I had never heard all the issues laid out so clearly. All the clinicians – we had never seen all this information gathered together and … so clearly laid out. (Pathway Chair)

The policy process enabled multiple communities – that were ordinarily not closely aligned in their priorities, practices and interests – to connect to each other, to new data and diverse protocols in new ways. New ways of knowing – argued rationally around ‘quality’ – were shaped as the context was reconfigured.

[Arego] has achieved the remarkable feat of rallying almost all the various groups in health to the cause of quality as the organising principle for the NHS (CEO NHS confederation)

Arego himself did not have expertise or legitimacy in all pathway areas; as a cancer surgeon he was not an expert in setting maternity or stroke service policy, for example. Though formally responsible for the policy content, he told the group: ‘I know nothing about maternity care; … you are the experts.’ Arego’s readiness to emphasize his own lack of expertise and his reliance on the expertise of the multidisciplinary groups aided the development of internal cohesiveness within new policy coalitions and enhanced the legitimacy of the process.

On numerous occasions the multidisciplinary pathway groups were unable to frame a consensus on controversial issues. For example, in the maternity and birth pathway, obstetricians and midwives could not easily agree on a role for midwife-led birthing centres and the relationship of these centres to physician-led birth units in acute hospitals. In this case Arego brought in external expert authorities (e.g. members from the Royal Colleges) as part of the assemblage needed to shift historical jurisdictional boundaries. He visited a number of birthing centres across London with midwives and obstetricians and discussed with them the details of scientific publications, and possible blueprints for service configuration. Arego’s preferred service blueprint gave emphasis to the midwifery position, yet retained local flexibility dependent on available resources; though a compromise was established, tension remained throughout the policy process.

Another area of strong discord involved the set-up of GP-led healthcare centres (‘polyclinics’) in the community. An area of historical tension, health executives demanded increased accountability from GP offices, and improved patient access to diagnostics. Yet these changes implied loss of institutionalized specialist services from the hospitals (e.g. X-rays) and the need for decentralized GPs to collaborate in order to make efficient use of expensive equipment. Arego delegated ongoing communication with GP audiences to a leading London GP, while he focused on hospital specialists and health executives. This enhanced the legitimacy of the views presented as each was able to draw on tacit norms and theories in use within their peer groups. In Arego’s view, discussions with GPs fundamentally reshaped the original form of these centres, though tension persisted.

Conflict and disagreement on the policy direction surfaced in other ways. Of particular note was the reframing of the initial six pathway groups into seven pathways as areas of medical specialization fought to have more sovereignty. Mental health specialists (psychiatrists) felt it was inappropriate to fit into current pathways – for example ‘acute care’ and ‘chronic care’. While Arego initially argued that the management of ‘acute’ illness (whether cardiac or psychiatric) should be the organizing principle – lest each specialty claim unique status – those delivering mental healthcare argued for distinctiveness in terms of historical context, organizational management and community links. As explained by Arego, ‘the great and the good’ with expertise in mental health met with him to advocate an exception. Coincidently, the government had recently released documents pointing out inadequacies in current mental health services and the need for community integration. Their historical isolation and unique data on mental health practices worked to reconfigure the psychiatrists’ concerns into a separate pathway group. This pathway group was less integrated into the overall policy process (for example, being absent from public engagement events). Yet the change served to appease psychiatrists’ concern for more autonomy and to minimize conflict from other specialists regarding the unique status of this group by linking mental health services to their unique institutional history, relationship to community services and current outcome data.

Leadership was not enacted by one or another individual accomplishing certain functions; agency was enacted through a collective assemblage that was multiply distributed and contingently enacted. The context of historical isolation (and rivalry) between clinical groups as well as current service structures were constituent in shaping leadership; these structures were particularly important in constituting and reconfiguring conflict. The distributed framing of leadership reveals how the context is (re)constituted; new coalitions are formed, new stroke performance data collected, interpretive schema and international best practice defined and numerous individuals feel ownership.

Leadership practice in engaging the public

In the White Paper, ‘The NHS Plan’, the Secretary of State for Health (2000) noted:

At its heart the problem for today’s NHS is that it is not sufficiently designed around the convenience and concerns of the patient.

In response to this central policy concern, Arego initiated extensive patient and public engagement. At the suggestion of the regional health authority, an initial poll was conducted of 7,000 Londoners to provide a quantitative baseline. Producing numerical data to capture public views and ratings of healthcare services provided a visible basis for the public to trust that Arego was interested in configuring new policy with citizens’ concerns, as the figures were reproduced on PowerPoint slides and glossy documents. It was also understood that this baseline data could subsequently be used by healthcare executives to evaluate performance improvement by use of repeat polls, which would arguably enhance positive relations of trust between clinical and managerial leaders and members of the general public. Thus polls and information were drawn from the policy context to configure leadership, which iteratively shaped the context by defining ‘a baseline’.

A key strength of the review process was having the MORI survey. That gave a lot of good quantitative data to support the change argument. And it can also be used as a good baseline for how people think about healthcare. (Policy team member)

The polls facilitated public systems-based trust in the wider policy process as being reliable. As a representative of the system, Arego enabled repair and built trust through face-to-face contact during ensuing public events; the polls constituted and specified the summative public views, producing a standard against which change was shaped. Rather than acting as external tools, polls as published evidence were central in legitimizing the policy process in the first instance, constituting part of the vision for change and for making sense of the current health service problems.

Following the public poll, four deliberative sessions were held: invited members of the public attended day-long professionally run consultative events. Arego also invited clinical leaders from the multidisciplinary pathway groups and high-profile health managers. Written scenarios across whole pathways of care were developed by the firm running the events. These stories – called ‘pen portraits’ – embodied an innovative design philosophy that integrated the perspectives of participants in pathway groups; views of ambulance, emergency medicine and rehabilitation staff as well as complex performance data and international protocols of best practice could be incorporated into scenarios of stroke care, for example. Computer-animated videos were developed in a virtual world – ‘Second Life’ – by one of Arego’s junior colleagues to illustrate scenarios of stroke or heart attack. These animations were a powerful sensemaking device, helping to shape meaning around complex issues, and particularly appealing to younger audience members. Voting through hand-held computer technology was used during public engagement events, so the range of views – such as preferred treatment scenarios, or opinions regarding the biggest NHS problem in London – was quantified and concretized. The data and numbers also produced the public ‘consensus’; a legitimized summary of public concerns could be claimed.

A recurring theme from elected representatives of the public, such as councillors from the mayor’s office, was the need to localize policy content so as to enable public engagement.

If you base your [argument] on clinical evidence you can convince the public. … We need to convince Londoners that it is their service. Turn it around and get Londoners to think ‘yes, this is my vision’… East London has been waiting for 25 years for [emergency heart attack and stroke] services. (London Assembly Councillor)

Technology was thus designed that portrayed clinical pathways in concrete terms, shown at public events and town hall meetings across London. A camera crew developed a video displaying the London Ambulance call centre as an emergency call comes in regarding a possible stroke victim. ‘Claudette’, the person who answers the call, connects to the ambulance which is filmed driving quickly with flashing lights across London streets to a stroke centre, bypassing a local hospital. During this ‘brain attack, minutes count’, and the narration explains that the clinical standard is to administer critical drugs within three hours of stroke onset. Before the drug can be administered, sophisticated radiology scans must be taken and interpreted. Thus for public members from the East London region, referred to in the above extract, the importance of having adequate access to emergency stroke services localized the story to their own community issue and educates them regarding stroke treatment standards. It also reminds them that the local concerns of their underprivileged borough have been previously ignored, but now given attention. In addition to bridging boundaries, material elements connect with local concerns, moving from a general strategy to contextualized account, building important relational connections between members of the public and possible health services. Arego highlighted:

This [standard] has been achieved in Canada! This has been achieved in Oxford, which is not that far from us. We are not achieving this [in London]

In so doing, enacting leadership entails distributing across the technology, current medical standards and statistics, global events in Canada and more regionally in Oxford and relating them to the current London services and public while legitimizing change. Without the established protocols, achievements elsewhere, big gaps in current services, and affordances enabled by the diverse technologies, a different sociomaterial entanglement would result, with an alternative change process. The transformation arises precisely because leadership is relationally distributed across the analyses, programs, call centres, videos, Canada, local public and Arego.

Technologies often achieved unintended consequences. For example a video exemplifying a possible community health centre (‘polyclinic’) showed the public how GPs could perform simple surgery in their offices. However, numerous specialists sneered when watching this video as the handwashing procedures shown did not meet hospital standards. The video related to public members differently from the way specialists related to it; tacit norms and local practices were intertwined with the video contents yielding different outcomes, a different assemblage.

Within the small focus groups, public members made frequent comments about the ‘good doctor’ (i.e. Arego) and that ‘they [i.e. Arego and colleagues leading the exercise] know what they are talking about’. This indicated high levels of trust towards Arego and other medical leaders at the event that drew on institutional norms of a doctor–patient relationship, rather than someone seeking or representing votes. This highlights how other individuals with similar skills – such as previous reviews executed by former doctors now working for the government – will not be able to relate to the general public in the same way; the institutional structures of medical care and doctor–patient relations mattered in our policy context.

Arego frequently reiterated that the money supporting the health system was ultimately coming from the public, acknowledging their right to participate in priority setting and his accountability to them:

One of the questions you asked was ‘is there political will?’ and I think [there is]. [Everyone is aware that Gordon Brown, prime minister elect, has entered the room.] … I have been getting stacks of papers from stakeholders sending us input and I am very grateful to [colleagues] for pulling this together … We started this process with a Mori poll to see what the public thought and … [lots of activities cited]. We have incorporated your feedback. (Arego at public event)

Health executives and other clinicians at locally organized public events, such as town hall meetings set up across London, reiterated their accountability to the public’s concern. In this way the senior management conferred added legitimacy on the policy process by supporting Arego’s views to the eyes and ears of the public. And vice versa, the strong public participation demonstrated in the review process also legitimated the policy content to others.

Many members of the public were surprised to be invited to give feedback on the content of the strategic vision. Anticipated public scepticism – that their participation would remain as a token involvement and lacking real influence on the policy process – was a key reason for planning the successive public events. While originally only two public engagement events had been planned, Arego was keen to reflect genuine concern, reciprocating the professional expectations of doctors for potential patients. At the final sessions, people were shown how the feedback from the previous events had been used in reshaping the vision for future services.

You feel you have been part of the project. At the end of the day, you feel your message is important and you have the leverage of making this happen. (Public member)

In addition to displaying responsiveness to the current policy context, these sessions also worked to reconfigure the views and attitudes of the public, thereby reconstituting the context. While not all the concerns raised by public members during the numerous consultative events were addressed, participants were able to see where group feedback had been influential. The resistance from the public was in some way mitigated, perhaps not so much because of satisfaction with the content, but because of the sense of empowerment that emerged from having been heard.

Distributing leadership practice with senior managers

The engagement of health executives and senior managers was largely founded by the Strategic Health Authority’s support for policy formulation and the executives’ need for a coherent strategy during a time of significant system wide change. Aware of cynicism related to frequent changes and policy initiatives coming from the cabinet offices, the SHA chief executive highlighted the uniqueness of working with a practising doctor. Arego’s legitimacy was enhanced because he was unusual in his position, informed about the actual health delivery in a specific London context and unconnected with party politics.

This time it isn’t being written by an uninterested bystander who disappears. It is being written by a practising clinician who has to continue to practise in London. … This is a serious opportunity to build a coalition which could make a good impact on health care and outcomes in London. (CEO of SHA, to group of 90 senior managers)

Arego sought to promote collaborative coalitions and broad participation between and within diverse management teams. For example, Arego and clinical pathway chairs presented their concerns about current health organization and ideas for the future at half-day workshops for all senior management executives in the previous five geographical regions of London, where videos and virtual world animations of pathways were shown along with data highlighting regional variation. The reconfiguration of services, such as the treatment of trauma or stroke, was discussed in terms of where current resources were in place, organizational protocols for referrals and service capacity. Integral to leadership were the historical relationships among the numerous executives, as enabled by the workshop dialogue, videos and PowerPoint slides. As highlighted by a senior executive of the SHA:

It is so important to get [executives] in the room talking to each other … They need to take the initiatives amongst themselves to make this happen … Some of them [have strong ties] with each other … but many … are used to working in isolation, just overseeing their patch – not seeing this as a whole integrated system where each organization has to work with the other.

An important Labour Government objective was to shift health services into communities – as evidenced by the 2006 White Paper entitled ‘Out of Hospital’ – distributing and coordinating leadership across multiple acute and primary care providers. Discussions between acute and primary care executives encouraged joint working and strategic alliances where previously there had been little communication.

You need to get primary and secondary care working together – it is not about complicated medicine. (Management consultant)

Arego’s legitimacy in his role to orchestrate coalitions was acknowledged by the managerial cadres. He had strong endorsement from the CEO of the SHA, who emphasized Arego’s (political and regional) independence. It was also widely known among health managers that Arego had previously conducted smaller-scale reviews for the Department of Health, often arbitrating between medical jurisdictional claims; he communicated his understanding of their position by reference to his experience, orientating to their concerns regarding issues of clinical turf battles and government efficiency targets. As a knighted professor at a prestigious medical school, his capacity to lead an evidence-based policy process held substantive legitimacy founded in links between macro institutions such as university systems and peerage with micro tasks of policy leadership.

Unsurprisingly, management leaders were wary about clinicians gaining too much control over the delivery of health, which they feared would risk undermining their own authority and hard-earned (post-Thatcher) jurisdiction amid creeping costs. Rather than evade engagement with policy formulation they strongly encouraged Arego – through repeated questioning about whether the data provided was relevant to their context – to avoid being prescriptive about how organizational reconfiguration would unfold in detail. Arego was aware of the strong historical local autonomy of London hospitals and medical schools and regional disparities of the London boroughs. Goodwill and trust from health managers was achieved by a loosely specified framework (for example, how blueprints for midwifery centres might be applied locally), providing managers with local autonomy. They were encouraged by Arego’s open stance in being transparent, his appeal to evidence and previous reform decisions.

Need to give clear strategic direction, but beware of doing too much top down detail. Hit the principle and direction, but let the people figure out how to do it. (NHS executive)

A former NHS CEO nicely highlighted how the social and material context of the NHS constitutes and shapes policy leadership practice.

Change in London is rooted in history. Change must be thought of in relation to the many stakeholders, the London colleges, research, nurses, medical education and the growing requirement to ensure clinical translation … all the variables have to go into the mix.

However, engagement with executives of volunteer organizations (the ‘third sector’) was less successful. Senior managers of charity organizations, many of which support health service delivery goals within community settings, were also invited to extended workshop sessions, using a similar format to public events to get their feedback as well as to inform. Historically charitable organizations have operated independently from formal health providers, often in a position of perceived domination and isolation. For example, while NHS organizations received stable government income to manage patients, charities continually compete for scarce funding to provide niche services that are rooted in patient and community concerns rather than providers’ roles. While formal providers might perceive these services as ‘extra’ or non-essential (i.e. in medical terms), charities believe their services to be essential, holistic and empowering of ‘patients’ to reintegrate more fully into productive meaningful life, an important alternative perspective. An unintended consequence of the numerous pen portraits, videos and virtual world technologies was the portrayal of possible service provision from a medical perspective, where doctors define worthy needs, goals and resources, and control relations. In addition, the technologies did not embed the organizational structures of the charity organizations, instead highlighting numerous protocols and structures held by hospitals and GPs. This reinforced an orientation of volunteer organizations’ isolation and marginalization in relation to state provision:

Why have they bothered to consult us? Where is our role in this? We are not even mentioned in any of these [scenarios]! (Executive of charity)

The effects achieved by sociomaterial assemblages emerge relationally, rather than being inherent in properties.

Furthermore, several executives of volunteer organizations questioned why workshops and programmes aimed at enabling dialogue with charitable organizations were not integrated with discussions involving other groups, such as hospital managers and senior medical staff beyond Arego and pathway chairs.

Why are they talking to us alone and not with [other doctors and managers]? … What we have is a dialogue. But what we need is a trialogue … [so that Arego is not mediating] the discussions … we need all parties talking together. (Charity chief executive)

The findings show the significance of going beyond the leadership activity of aligning new coalitions and designing workshops to enable collaboration. While volunteer sector executives were pleased to be included, their positioning within the overall institutional configuration unintentionally served to reinforce their segregation and did not engender relations of trust or legitimacy.

Enacting legitimacy in leading policy formulation

Legitimacy in leadership can be mediated by the available statistics related to the policy situation as well as clinically relevant knowledge within the respective disciplinary community. Denis, Langley and Rouleau (2006) have shown how statistics are readily constructed to display transparency, enable reconciliation and generate commitment between diverse values and interests in establishing legitimacy. Numbers can be powerful tools to support decision making and generate consensus. Legitimacy regarding the policy process was sustained by the entanglement of espoused medical values, technologies and health outcome statistics rather than overt political agendas, partisan party stances or single clinical perspectives. As Feldman and colleagues (2006) note, knowledge networks shape policy issues; material entities and social actors and the associations that connect them fluidly shape ways of knowing.

The practice of medicine is characterized by tacit knowledge structures concerning collegial interdependence and consensus seeking (Freidson, 2001; Levay & Waks, 2009), albeit within a context of autonomy and jurisdictional power (Abbott, 1988). These knowledge structures were embedded in Arego’s way of working with clinical colleagues, underpinning a level of taken-for-grantedness, and cognitive level legitimacy (Suchman, 1995) to co-orientate leadership. Further, rather than challenging the organizational protocols and institutional structures of healthcare delivery (e.g. clinician autonomy, peer accountability, codes of ethics) clinical leaders drew on these protocols and structures – along with numerical data and research reports – to find new ways of working. In this manner, we observe how knowledge structures and organizational protocols mediate and shape processes of legitimacy, sustaining leadership processes and enabling change.

Further, we suggest that in relation to the public, policy leaders developed a moral basis for their legitimacy (Suchman, 1995), as mediated by their relational entanglement with pen portraits, videos and workshops. A concern for patients’ best interests is a foundational value espoused by the medical profession (Abbott, 1988; Freidson, 2001) and thus a reified expectation frequently extrapolated onto ideas and intentions of individual clinicians, as well as tools entangled in their practices (Oborn, Barrett & Davidson, 2011). The open and transparent nature of the workshops, egalitarian approach to include views of diverse stakeholders as well as international research reports enforced a moral and pragmatic basis for legitimacy (Suchman, 1995).

Our findings also build on the work of Human and Provan (2000) and others (Bryson, Crosby & Stone, 2006) who have demonstrated that the legitimacy of a collaboration needs to be established in order to support interaction across multiple stakeholders. We additionally highlight that legitimacy is established relationally among medics, their expertise, jurisdictions, work protocols, government legislation and tools such as radiographic images.

Establishing trust relations through sociomaterial practices

Another important dynamic in understanding why leadership was successful in establishing new policy concerned the establishment of relations of trust. The public at large is more likely to trust medical doctors than many other occupational groups involved in policy formulation, such as managers, lawyers or elected politicians, due to the perception of their unique moral commitment (Freidson, 2001; Hunter, 1995; Oborn, 2008). Trust is an important response to uncertainty when one cannot gain access to evidence or information to assess impact adequately (Sztompka, 1998). Given the inherent uncertainty in healthcare and treatments, trust has been suggested as a cornerstone of modern healthcare delivery and reform (Calnan & Rowe 2008; Taylor Gooby, 2006), though there has been little empirical or conceptual focus on the role of trust in institutional health systems and the policy process (Calnan & Rowe, 2008).

Our study reveals how trust has been established between individual policy leaders and the public. While Arego did not go so far as to appear at the public engagement events wearing a white coat, he contributed to the constitution of trust in a number of ways. His entanglement in sociomaterial assemblages included: history and experience in the National Health Service; visible connections with polity and health elite; doctor–patient relationships developed through patient treatment and visits to stroke units and birthing centres; as well as face-to-face contact in workshops. In attending to numerous tools, places and interactions in a seemingly transparent configuration, trust based on rational cognition could be relationally configured as inferred from past actions (Shapiro, Sheppard & Cheraskin, 1992). Polls acknowledged input from the public and including them in the policy process, nurturing instrumentally based trust by explicating concern for their (the public’s) own interests (Barber, 1983). Our study therefore highlights the establishment of trust through sociomaterial practices and in so doing builds on earlier work which has highlighted the development of trust as an ongoing process (Huxham & Vangen, 2005; Ring & Van de Ven, 1994) accomplished by sharing information and demonstrating competency (Bryson et al., 2006).

Trust was also established in the policy process itself. In order to establish trust in abstract systems, Giddens (1990) suggests that individuals can serve as representatives to reinforce trust in structures through rules, procedures and even standards that comprise the system. In our case, we suggest that Arego – himself sociomaterially configured by degrees, robots, honours, hospitals, Royal Colleges, surgical knowledge – configured public trust. As numerous clinicians, as representatives of the health policy system, became involved in the policy formulation process, trust that the recommendations were actionable and relevant increased. In this way, members of the public could build up their trust in the health reform procedures and standards, and in so doing develop co-ownership. This highlights an important relational dynamic whereby entanglements of clinicians, leadership activities, institutional structures and public can enable the public’s trust in the overall policy process. As the formal leader, Arego’s ability to distribute the accountability afforded to him through the complex web of sociomaterial relations could thus be re-appropriated as the polity system being accountable to the tax-paying public, motivating their involvement and support. The role of trust relations in health institutions and the policy process has been relatively neglected (Taylor Gooby, 2006) yet are central to securing commitment and support for health systems, particularly in an era of changing public attitudes towards health.