Changing the Ties That Bind? The Emerging Roles and Identities of General Practitioners and Managers in the New Clinical Commissioning Groups in the English NHS

The English National Health Service (NHS) is undergoing significant reorganization following the 2012 Health and Social Care Act. Key to these changes is the shift of responsibility for commissioning services from Primary Care Trusts (PCTs) to general practitioners (GPs) working together in Clinical Commissioning Groups (CCGs). This article is based on an empirical study that examined the development of emerging CCGs in eight case studies across England between September 2011 and June 2012. The findings are based on interviews with GPs and managers, observations of meetings, and reading of related documents. Scott’s notion that institutions are constituted by three pillars—the regulative, normative, and cognitive–cultural—is explored here. This approach helps to understand the changing roles and identities of doctors and managers implicated by the present reforms. This article notes the far reaching changes in the regulative pillar and questions how these changes will affect the normative and cultural–cognitive pillars.


Introduction
The Institution of the NHS In July 2012, London hosted the Olympic Games and television viewers across the world tuned in to watch the opening ceremony. Amid depictions of a selection of key moments in British history and cultural life was a tribute to the National Health Service (NHS). A newspaper headline later proclaimed, "Americans baffled by 'left-wing tribute' to free healthcare during Opening Ceremonies," with a Los Angeles Time sports reporter, Diane Pucin commenting, "For the life of me, though, I am still baffled by NHS tribute at opening ceremonies. Like a tribute to United Health Care or something in US" (Press Association, 2012). Danny Boyle, director of the ceremony, said, "He chose to feature it because 'everyone is aware of how important the NHS is to everybody in this country'" (Stebner, 2012). The prominence of the sequence in the ceremony alongside references to important moments in the nation's social history certainly suggests that the NHS is an institution sufficiently deeply embedded within Britain to contribute significant meaning to its society The NHS is a relatively young institution that came into being in 1948 and was for some time inextricably linked with the restructuring and nation-building that characterized the post-war period. In charting the history of the NHS, Baggott (2004) points out its popularity among post-war generations who have grown up with the welfare state. For those with no pre-1948 experience, a health service that is free at the point of delivery and paid for out of taxation still constitutes the "natural" order of things, although the waves of crisis around management, finance, and reorganization that have beset the NHS from the 1970s onward together with scandals concerning patient care have shaken employee and public confidence in aspects of the institution. Nevertheless, as Appleby (2012) points out, recent surveys of public satisfaction with the NHS fluctuate but remain fairly high. Moreover, public support for the principle that it is "the government's responsibility to provide health care for the sick" is undiminished, while backing for government support for the unemployed has fallen (Park, Bryson, Clery, Curtice, & Phillips, 2013, Table  2.1).
It is not surprising, then, that Danny Boyle should choose the NHS as a symbol of "Britishness" and that outsiders (Americans) would find attachment to this institution baffling. The suggestion of a "left wing tribute" also highlights the idea that the NHS is emblematic of the welfare state where health care is provided to all irrespective of wealth or status. Recent work focusing on developments within the NHS (Checkland, Harrison, Snow, McDermott, & Coleman, 2012;Macfarlane, Exworthy, Wilmott, & Greenhalgh, 2011) has suggested that Scott's analytic approach to institutions (Scott, 2008;Scott & Christensen, 1995) is particularly pertinent. Located within the tradition of sociological institutionalism, Scott argues that institutions can be understood in terms of three constitutive elements or pillars: the regulative, normative, and cognitive-cultural pillars. Institutions are shaped by sets of rules and regulations (the regulative pillar) enforced by sanctions and sometimes coercion. Scott argues that the regulative aspects of institutions are those of which we may be most conscious. The normative pillar of institutions refers to the norms and values held by individuals and the behavior that ensues from their efforts to uphold these values. Institutions, Scott (2008) asserts, ". . . are widely viewed as having moral roots" (p. 56), which shape the conduct and behavior of individuals. The cognitive-cultural pillar of institutions refers to the taken-for-granted and unremarked aspects of institutional life. These are the characteristics that an outsider may remark on but which the "native" intuits (as explicated by Geertz, 1973). We attach meaning to how various institutions work; Scott & Christensen (1995) use the term "cultural template" (p. xviii) and assert we will understand what constitutes a commercial corporation, a school, or a hospital. The cultural template we have for each will be different and we can see how the meanings we attach to these kinds of institutions will be determined by context and history. Macfarlane et al. (2011) and  explain how Scott's institutional theory might help in understanding the NHS. The NHS is made up of many different organizations including hospitals, public health, dental services, GP surgeries, and so on. These organizations overlap and are staffed by individuals with a range of professional interests, for example, managers, nurses, and doctors. However, these writers argue that despite organizational and professional differences, the NHS as a whole has a particular ethos and set of values with which the workforce can and does identify. So the managers interviewed by Macfarlane et al. strongly identified with the "socialist origins and egalitarian, no-frills values" of the NHS (Macfarlane et al., 2011, p. 919), while the doctors and managers in Checkland et al.'s studies saw themselves as being part of an "NHS family." Furthermore, there are a range of embedded cognitive-cultural assumptions that NHS "insiders" take for granted, such as an assumption that those working in the NHS, regardless of their employing organization, are members of a shared common enterprise .
One of the key concerns of Scott's institutionalism is the idea of stability and continuity: Something is said to have become "institutionalised" when it has become self-perpetuating (Scott, 2008, p. 22). This is not to say that institutions cannot change, and understanding when and how such change will occur is one of the key empirical puzzles for institutional scholars. The NHS is clearly a different institution today from the one created in 1948 (Klein, 2010). Scott's institutional theory can help us to think about how that change might have occurred and become embedded over time. Sustained change requires that the three pillars remain aligned, with regulative changes backed up by complementary evolution in both norms and cognitive-cultural mental scripts (Scott, 2008). Indeed, Caronna (2004) persuasively argues that regulatory changes unmatched by changes to norms or cognitive scripts generated dysfunctional change in the American health system. The past two decades alone have witnessed profound changes affecting the English NHS. For many commentators, the NHS and Community Care Act of 1990 was a watershed regulatory change, as it marked the beginning of the purchaser-provider split and quasi-market. With these reforms came challenges to medical professional control and autonomy, increasing self-regulation and surveillance, and the entry of private providers into the NHS marketplace (Harrison & Dowswell, 2002;McDonald, 2009;McDonald, Harrison, & Checkland, 2008). Commentators have noted that competition, regulation, and the quasi-market are stronger features of the English NHS in comparison with those of Wales and Scotland, whose governance is devolved from British central government (Greer, 2008). Harrison (2009) focuses on the introduction of self-regulation and commodification of health care through mechanisms such as the Quality Outcomes Framework (an incentive scheme for GPs with financial rewards going to practices accomplishing tasks on a menu of patient care indicators). He shows that these regulations challenge normative views concerning the professional roles of doctors, but suggests that this "commodification" can, in time, become "naturalised" as the normal way doctors think about their practice. Thus, it seems that the profound regulatory changes experienced by the NHS in the past have been accompanied by some shifts in the cognitive-cultural taken-for-granted understandings among institutional members.
In the next section, we explore the possibilities of such shifts through an examination of the process of doctors becoming managers, a process that is certainly not new but has recently become more pronounced.

Changing Roles and Identities: Doctors Becoming Managers
One of the shifts in institutional cognitive-cultural assumptions identified by Harrison (2009) was the way in which managerial ways of thinking became normalized among members of the medical profession. This normalization has both fed and been fed by the significant increase in formal managerial roles for both doctors and nurses within the NHS, initially mainly in hospitals but more recently also in primary care (Bolton, 2005;Sheaff et al., 2003;Thorne, 2002). Indeed, although Harrison (1988) noted that some clinicians, including GPs, have been involved in NHS management since 1974, present reforms set the scene for GPs to be involved in management in unprecedented numbers.
The subject of "doctors as managers" has received a great deal of attention and is particularly relevant here because it touches on both the normative elements of what it means to be a professional clinician and the deeper cognitive-cultural assumptions, which underpin identity. What are the responsibilities and duties of a doctor and of a manager and what happens when these roles are combined? A recurrent theme in this literature is the way in which doctors who take on managerial roles are confronted with questions about their identity and in so doing often choose to reassert the primacy of their identities as clinicians. This was nicely illustrated by Lord Darzi, a London surgeon, who was asked by the then Labor government to undertake a "review" of the NHS in 2008. This was, in essence, a managerial task, but the final report (Department of Health, 2008) is illustrated by a photograph of Lord Darzi wearing operating theater clothes. He chose to present himself as primarily a clinician, although he was acting in a managerial capacity.
In this context, we take "roles" to be the formal and informal expectations held by individuals and their colleagues about the duties and responsibilities associated with holding particular positions (Katz & Kahn, 1978). "Identities" are "the individual's own notion of who and what they are" (Watson, 2008, p. 131), shaped by their experiences and by the social situation in which they find themselves. Forbes, Hallier, and Kelly (2004) and Forbes and Hallier (2006) suggest that doctors in hospitals who take on managerial roles undergo a recategorization of the social self. They regard themselves as doctors first, primarily responsible to their patients, and some engage in "organisational misbehaviour," which obstructs the work of senior managers. Both Mo (2008) and Llewellyn (2001) report that management work is regarded as something that takes doctors away from their clinical work where their prime loyalties and allegiances lie. They also make the point, as does Doolin (2001), that management work has lesser status than clinical work and that those doctors taking on managerial roles adversely affect their reputation in the eyes of their peers. Working in these "hybrid" roles entails a blurring of boundaries and hence gives rise to anxieties about identity. Similarly, nurses in managerial roles are keen to prioritize their clinical qualifications over new leadership titles (Martin & Learmonth, 2012), and to make the point that they are nurses first and managers second (Bolton, 2005). McDermott, Checkland, Harrison, Snow, and Coleman (2013) make a similar observation in respect of GPs who took on managerial roles under the scheme known as Practice Based Commissioning (PBC). PBC was a voluntary scheme that enabled groups of GP practices known as consortia to become involved in the commissioning of services. While participating GPs were enthusiastic about the scheme itself, GPs who took on managerial roles as GP commissioners identified themselves as GPs rather than as managers and displayed low levels of certainty about their roles as managers. It could be argued that these uncertainties and tensions have an origin, at least in part, in some deep-seated and only partially conscious cognitivecultural assumptions about what a GP "is" and what they "do." Thus, clinicians taking on managerial roles continue to claim that their prime allegiance remains to the ideal of patient care. Similarly, research among managers working in the NHS suggests that they too have a strong commitment to patient care and to public service (Currie & Brown, 2003;Merali, 2005). Hewison (2002) highlights NHS middle managers' concern for patient welfare and points out that they share a similar "commitment to the values and ethos of the NHS" (p. 564) as their clinical counterparts, contesting the stereotype of the gray suited manager concerned solely with the "bottom line" Since the 1980s, managers in the NHS have borne the brunt of politically inspired, top-down reorganization and change. This has been referred to by McMurray (2010) as a "pandemic of organisational reforms" (p. 56), which he suggests results in "response fatigue" among affected managers. Nevertheless, Greener's (2008) study concluded that while NHS managers' work is constantly changing, they keep sight of their long-term goal to serve the best interests of the public. Likewise, Macfarlane et al. (2011) found that senior NHS managers held fast to their ideals of an NHS culture rooted in a belief of public service.
It has been suggested that a public service ethos is characterized by traits such as working for the good of the community, loyalty to one's institution, and accountability to the public (Pratchett & Wingfield, 1996). This implies that ideas about public service are deeply connected to people's sense of identity, incorporating both a normative understanding about roles and duties and a deeper cognitive-cultural conception of what it means to be a public servant. McDonough (2006) attempts to unravel these understandings using Bourdieu's notion of habitus to explain Toronto-based municipal workers' fierce attachment to a sense of public service even when experiencing restructuring, cutbacks, and privatization of services. She argues that their "public service habitus" usually remains subconscious but is forced into consciousness when confronted with the idea that the private sector may be more efficient. She suggests that the idea of what constitutes the public good will be increasingly contested and potentially shifted by the continued restructuring, resulting in a "destabilised habitus" for affected workers. Scott (2008) considers the way in which institutions change, and of interest here, discusses how they may weaken and disappear. He points out that scholarly analysis often highlights one aspect-regulative or normative or cognitivecultural change-as precipitating deinstitutionali-zation but argues that these elements usually interact with one another. For the NHS, it can be argued that successive regulative change has clearly shifted the roles and duties of personnel and may, as noted above, also begin to shift norms and values. Doctors and nurses who take on managerial roles are described as becoming "hybrids" while health care professionals who implement guidelines and algorithms are seen as having been gradually "co-opted" by management (Harrison, 2009;Numerato, Salvatore, & Fattore, 2012). Managers with long careers in the NHS have witnessed periods of empowerment and legitimation followed by a "delayering" of middle management and a clawing back of power (Currie & Brown, 2003;Macfarlane et al., 2011). Thus, roles and associated normative understandings have changed (Greener, 2008;Harrison & Dowswell, 2002). However, it would seem that, through these changes, both clinicians and managers have held fast to a cognitive-cultural sense of identity anchored in a notion of public service and patient welfare.
Currently, the English NHS is adjusting to a major reorganization. As we show in the next section, workers in all parts of the service are taking on new roles and responsibilities as new structures are put in place and new organizations created. In terms of Scott's three institutional pillars, it is clear that the regulative pillar of the NHS institution has undergone profound change. We argue that these changes call into question the normative and cultural scripts held by GPs who must now understand themselves as commissioners (that is contracting with other organizations to provide services) as well as providers of services. Being a commissioner is a challenge to the normative sense of the GP role understood as focused on patient care. To a lesser extent, managers' normative scripts of public service are also being queried, for example, as some are moves to support units destined to separate from the NHS. As McDonough (2006) points out, these levels of change force into consciousness questions about identity, obliging individuals to think about their changing roles and what these mean, and potentially challenging the deeply rooted cognitive-cultural assumptions about the way in which "things are done" in the NHS. The rest of this article briefly outlines some of these recent changes and then reports on a research study that focused on one aspect of the changes: the introduction of Clinical Commissioning Groups (CCGs). Findings from the qualitative part of this study are presented to show the implications of some of these changes for the roles and identities of GPs and managers.

Reorganization-The New Clinical Commissioning Groups
In July 2010, plans for a significant reorganization of the NHS were outlined in the White Paper-"Equity and Excellence: Liberating the NHS" (Department of Health, 2010). Among other changes, responsibility for commissioning care for defined geographical populations was to be passed to groups of GPs working together in CCGs. Thus, in addition to their status as independent contractors to the NHS, some individual GPs are now used by the CCG to carry out management duties. The White Paper envisaged that this would make GPs more accountable to the patients they serve, and that it would provide them with an incentive to act in ways that cut costs. At the same time, the previous (managerially dominated) purchasing organizations, Primary Care Trusts (PCTs), would be abolished and a new body, NHS England (initially called the NHS Commissioning Board), established, with responsibility for overseeing the work of CCGs, allocating budgets, and undertaking some aspects of commissioning themselves (e.g., specialist services). It was argued that GPs' proximity to the frontline of patient care put them in the best position to understand the needs of their patients, making them responsive commissioners: In order to shift decision-making as close as possible to individual patients, the Department will devolve power and responsibility for commissioning services to local consortia of GP practices. This change will build on the pivotal and trusted role that primary care professionals already play in coordinating patient care . . .. (Department of Health, 2010, p. 27) While GPs take on increasing commissioning roles and responsibilities, the White Paper declared that management costs were simultaneously to be cut by 45% and that £20 billion of efficiency savings would be realized by 2014. As Asthana (2011) points out, in this narrative, managers are unambiguously equated with unnecessary bureaucracy and cost. GPs, however, add value to the commissioning process by bringing their "skills, knowledge and standing in local communities" (NHS Commissioning Board, 2012b).
Each practice compulsorily became a member of a CCG, with mechanisms in place to elect peers onto the governing body. These representatives are overwhelmingly GPs, although some have also elected practice nurses or practice managers. In addition, each CCG is required to have a nurse member, a consultant, and two lay members. The configuration and organization of CCGs differ from site to site, with no overall template set from the center. Most have some kind of formal members' body, consisting of representatives from each member practice, and many have also set up geographically based locality groups to involve a wider range of local GP members. The choices that they have made so far often derive from their own recent history  and the particularities of the local context .
During the transition period, managerial support for CCGs came from PCT staff, some of whom transferred to the new structures as PCTs disbanded. However, CCG managerial budgets have been set at a much lower level than was the case for PCTs, and the new bodies are expected to "buy in" significant parts of their managerial support from newly formed organizations called Commissioning Support Units (CSUs). These have been set up by former PCT staff and often cover a large geographical area. While, initially, CSU staff will be formally employed by NHS England, it is the stated intention that these organizations will become "autonomous organisations in 2016 and will be fully established, self-sustaining entities in a competitive market" (NHS England, 2014). Overall, this represents a significant contraction in managerial numbers meaning redundancies and job losses, and will also mean that managerial staff who are transferred from former PCTs to CSUs could be working for non-NHS employers after 2016. At the time of writing, there are 211 CCGs in England (NHS Commissioning Board, 2012a) and 17 CSUs (NHS England, 2014).

Method
This article derives from a project examining the early workings of emerging CCGs . The aim of the project was to capture the experiences of "Pathfinder" CCGs (i.e., aspirant and emerging CCGs during their period of early formation) during a period of intense change. The overall research questions for the project explored their experiences during this process. In this article, we focus on the experiences of GPs and managers as they came to terms with their new responsibilities, answering the following broad research question: "How are GP and managers adapting to their new responsibilities, and what issues are arising?" Eight case study sites in England were purposively sampled, to include a range of population sizes, organizational structures, socio-demographic variation, relationships with provider organizations, and configuration in relation to local authority boundaries. Groups had signed up to become Pathfinder CCGs in five different temporal "waves," and our sample included CCGs from Waves 1, 2, 3, and 5. Both urban and rural sites were included. This sampling strategy enabled the capture of the full range of complexity in developing CCGs.
Fieldwork was carried out in these sites from September 2011 to May/June 2012 and included interviews with a range of NHS managers (47), GPs (33), and others (11) associated with the CCGs (e.g., lay members) and the observation of a range of meetings including CCG governing body meetings, operational meetings, locality meetings, general members' meetings, and meetings with a variety of external bodies such as Health and Wellbeing Boards (forums bringing together commissioners of health, public health, and social care). Contemporaneous field notes were made and subsequently written up. Documents associated with these meetings and with the governance and organization of their groups were collected and read. In total, 146 meetings were observed (approx. 439 hr) and 96 interviews undertaken; these were recorded (with consent) and transcribed. Ethics approval for this study was granted by the Greater Manchester West Ethics Committee Research Ethics Committee 11/ NW/0375.
During the field research, we undertook to preserve the anonymity of our participants. This has been a strong and guiding principle throughout the research as we wanted to ensure that all involved could speak freely. Care has been taken that the anonymity of research participants has also been maintained in the presentation of this article.
All of these data were analyzed with the help of Atlas ti™ software. Data collection and analysis for each case within the study have been undertaken in parallel, allowing the research team to modify and develop topic guides as appropriate, following up significant findings and seeking contradictory or confirmatory examples. Field notes taken during meetings were coded alongside interview transcripts and relevant documents, and initially emerging coding definitions, analytical themes, and theoretical ideas were discussed and refined at regular face-to-face team meetings and Skype conferences. Nineteen such meetings/conferences were held between September 2011 and May 2012. Transcripts and field notes were read repeatedly for familiarization and coded according to an initial framework based on our research questions, our reading of relevant literature, and our understanding of the policy context. Furthermore, "second level" coding was undertaken to capture any unexpected themes, which emerged from the data, and the analysis continually refined in written memos and team discussions. Emerging analytical ideas were tested among the research team members and refined. Coded data were then further read and analyzed by a number of team members to ensure consistency of approach. Initial thematic coding was refined, revisited, and revised over the course of the fieldwork. Thus, the data have been read and reread several times and unexpected second-order themes included in the analysis. For example, "doing things differently" emerged as a recurring theme, as those involved sought to establish their credentials as "new" organizations, untainted by the perceived failures of their predecessor commissioning organizations.
Two web-based surveys and a telephone survey were carried out as part of the broader project; however, this article is based solely on the qualitative data collected, focusing on evidence about possible changes in what participants take for granted about their roles and identities in the NHS institution. For full details of the broader project, see . Analysis of the data resulted in four broad themes relating to the roles of GPs and managers: shifting roles and identities, time pressures and other difficulties experienced, what it means to be a "good commissioner" in the new system, and the uncertainties facing managers.

GPs With a Foot in Two Worlds: Shifting Roles and New Identities
Clinicians who take on managerial roles are described as being "bridges," "two way windows," "Janus-faced" (Witman, Smid, Meurs, & Willems, 2011) or "wearing two hats" (Checkland, Snow, McDermott, Harrison, & Coleman, 2011). As mentioned above, they often feel that their clinical identity is threatened or compromised by managerial involvement but that clinical identity is paramount. GPs taking on positions in the new CCGs often had some previous experience of commissioning work such as GP Fundholding 1 in the 1990s or PBC (see above) in the 2000s . Miller et al.'s review highlights that many of the GPs who became involved in Fundholding were entrepreneurial by nature and were innovative leaders. The enthusiasm of these individuals was not necessarily shared with rank and file GPs. Membership of CCGs, unlike predecessor primary care commissioning arrangements, is mandatory for all GP practices. GPs with formal roles will, in future, be involved in decision making about commissioning/decommissioning of services and will have statutory responsibility for these decisions in terms of finances and care. Their commissioning responsibilities will be wider than under previous schemes as will be their power to spend budgets. They will also negotiate with secondary service providers, Local Authorities, and other external bodies such as Health and Wellbeing Boards. It will also fall to these GPs to communicate the work of the CCG to its members and to oversee the clinical behavior of members through performance management of practices relating to CCG budgets. It should also be remembered that GPs who sit on CCGs will also remain working GPs with caseloads of patients in their own practices.
Among the GPs that we interviewed, some articulated tensions between their identities as clinicians and as managers. As the literature cited above demonstrates, this is not necessarily new. As in previous GP commissioning arrangements, the enthusiasts have taken on roles in the new CCG governing bodies. The key difference now is that all GPs, irrespective of inclination or enthusiasm, are through their practices, members of CCGs. Many enthusiasts commented that they enjoyed their new roles on CCG governing bodies but that this work differentiated them from fellow GPs: . . .you have to understand that the vast majority of GPs are not in the least bit interested in all this, they just want to do their day job, get on, look after their patients; they don't want to spend their time reading documents and papers and commissioning and all the rest of it. (GP and CCG chair ID 33) Elections onto governing body positions showed the continuing tension between the clinician and manager role, with many of the leadership positions being taken on by previous office bearers and evidence of reluctance among the general membership to take on these roles. Some sites struggled to recruit new leaders and one site had a GP vacancy on their governing board throughout our fieldwork period. On being asked about election as locality chair and member of the CCG governing body, this GP responded, I've been very clinical in my career and I've got lots of clinical special interests. And I've got lots of ideas and energy and enthusiasm to develop those, and this is putting some barriers in the way of that, in terms of time and also conflict of interest. So it's opened some doors and closed others. And I still don't know how I feel about that. (GP, locality lead and CCG board member ID 103) Some CCGs were unable to hold elections for office bearers because of lack of nominations for positions and several GPs explained that they had obtained their positions by "default." In common with some clinician managers in hospitals, GP interviewees voiced misgivings about taking on managerial roles (Thorne, 2002). They were concerned, on a practical level, about not having the skills for these roles and not properly understanding the governance structures of CCGs. On a deeper level, they did not view themselves as really being "proper" managers: I think they're expecting a lot from people that don't have those skills and experience to do it and in all honesty I didn't go into . . . medicine to be a manager and a commissioner. I mean I quite enjoy it I think it's quite interesting in a way . . . but you only know what you've picked up, it's not something you're trained to do. (GP, CCG locality board member ID 229) Anxiety about their role as both commissioners and decommissioners of services is evident in the meeting extract below. Being on the "frontline" of patient care means that GPs will be responsible for commissioning decisions, including decisions that may be unpopular with patients. Here, locality board members need to limit an exercise referral program: [Name] started her introduction by saying that there would be a lot of opposition to what was being proposed . . . to relook at the current gym where GPs could refer patients to if they required exercise programmes. Problem was that once a patient was referred there was no end date so many people were long term attenders. An additional issue was the location of the gymmany people could not access the facility geographically so it was not an equitable service and was currently not cost effective . . . We spend eight times as much on this service as other localities and have an insufficient service.
Proposal is to have a 16 week time limit on the service. For GPs to design a new pathway. We can't decommission the service-this is unrealistic but we need to put a hold on new referrals for 3 months . . . Patients will be unhappy as they like the service. (CCG Locality executive meeting notes M35, emphasis added) This extract seems to suggest that those present share some cognitive-cultural assumptions-it is said to be "unrealistic" to decommission a service, and this statement is not challenged.

Time Pressures
Those GPs who have accepted new roles in the CCG now face significant pressures on their time. They are usually freed from their clinical duties for two to four sessions a week and their practices are paid for locum cover. However, these are not necessarily straightforward substitutions of time and effort. As respondents pointed out, CCG work entails attending and participating in formal meetings, workshops, and training sessions both internal to the CCG and external (e.g., Health and Wellbeing Boards). Associated with these meetings is a significant volume of documents and emails. These commitments have implications for time spent with patients, in the practice and personal time. As illustrated below, there is no direct equivalence of the time spent on CCG work with the time missed from GP practice work. Time spent with patients, with practice colleagues, and on home life has significance over and above the minutes and hours involved. This resonates with Jones and Green's (2006) findings on the significance of work/life balance and flexible ways of working for the professional identities of contemporary GPs. I think the roles . . . I mean if I look back to where we were . . . you know, in practice-based commissioning from where we are now and the role is just unrecognisable the amount of work that comes through, is huge. And to be honest, it is limitless. So I very much see it as I have to prioritise and I have to do what I can, accept that I cannot go to every meeting, and I cannot be available for everything, and there are some things that at the moment are going to get more attention and for the others maybe less of a priority for the CCG. (GP and CCG clinical lead ID 36) The ideal of GPs who are close to their patients and thus responsive commissioners-as set out in the government's 2010 White Paper-may potentially place a heavy burden on the shoulders of a few GPs. This GP sets out her strong conviction that being a "good commissioner" also requires her to continue to shoulder a reasonable load of clinical work. She reflects negatively on "certain colleagues" whom she feels have already relinquished their "GP identity." I always believed that to be a good clinical commissioner I need to be a good clinician, so I have to study being a GP . . . and I have to deliver a sensible number of sessions so that people still know that I'm a GP and I'm delivering what GPs do. Certain of my colleagues I've already seen doing less and less and less . . . and their ability to continue in meetings already is . . . you know, you could say: "well, what do you think about this? [clinical service]" And they're like: "I've never used that." (GP and CCG chair ID 231) In her view, the commissioning role requires the maintenance of a strong GP clinical identity, but the evidence presented above suggests that this is both practically difficult to achieve as well as carrying with it inherent tensions and conflicts.
In addition, our respondents were aware that their involvement at CCG level had an impact on their practice colleagues: I think my biggest fear about all of the commissioning and the changes that are going on with the reforms, is about the fact that I'm not sure there is capacity . . . you're taking out, you know, experienced GPs out of practice and . . . as a consequence either there's backfill with less experienced people, with locums, with less continuity of care . . .. (GP and CCG board member ID 36) Many also pointed out that a great deal of CCG work had to be done in personal time, in the evenings or at weekends. It is unlikely that this will be sustainable over any length of time, raising issues for the future of GP-led commissioning, especially if those involved are to continue to carry out significant clinical duties. If they relinquish those duties, our respondent quoted above would suggest that they will lose credibility as commissioners.

Being a Good Commissioner and Being Accountable
The idea that GPs are on the "frontline" of care and are therefore responsive to patient need and requirements is one that resonated with our GP interviewees. They agreed that their proximity to patients placed them in a good position to commission services intelligently. There was, however, a growing awareness among some that they will shortly be directly accountable to patients for commissioning decisions (Lind, 2012). With the handover of budgetary responsibility in the context of the target of £20 billion of saving in health care, it is likely that GP commissioners may have to make unpopular decisions about spending and decommissioning services. The responsibility for such decisions will no longer be attributable to PCTs, but to CCGs of which all GPs, via practices, are now members.
GPs are probably the right people to do this, because the beauty of the fact that we have to sit across the table from the individual patient. And yes, we're not the most patient responsive bunch of people, but we still have to meet Mrs Jones, and she still gets to rant at us about the fact that her hip operation isn't being done. And it will be our ears that get bent if we get it wrong. Whereas that's not the case if you ask anybody else to commission. (GP and CCG locality board member ID 221) What does worry me is the very difficult financial situation we're in. Huge quick savings that we have been tasked with, and the potential difficult rationing decisions we're going to have to make and how that's going to sit with the public in the future and the newspapers and that sort of thing. (GP and CCG board member ID 103) By contrast, both GPs and managers voiced more optimism and excitement about increasing opportunities to take part in contract negotiations with secondary care providers.
Best thing I've done, I suppose, in terms of the service redesigns I've . . . discussing clinician to clinician with secondary care colleagues, you . . . agree what makes obvious sense from a patient pathway point of view, and then . . . the managers find a way of making that work. (GP and CCG board member ID 67) As McDonald (2009) points out, GPs enjoy high status relative to the broader community, but this is not the case within the narrow context of medicine. It may be that entering into contract negotiations in this way might mark an increase in status for GPs.

Managers Facing Uncertainty and Insecurity
Many of our respondents, both clinicians and managers, jokingly enumerated how many NHS reorganizations they had experienced. Managers in particular have borne the brunt of many of these (Light & Connor, 2011;Macfarlane et al., 2011). They have also been portrayed by politicians and the media as a bloated sector whose numbers need to be cut (Merali, 2005). NHS managers have been characterized as being resilient and resourceful in the face of these changes (Checkland et al., 2011;Greener, 2008), and many of our informants were optimistic and enthusiastic about the current reforms, highlighting the value that GPs' clinical expertise adds to the commissioning process. However, they were concerned about GPs taking on managerial roles and the concomitant shift in their own roles. Some articulated this as both GPs and managers having to undergo a "culture change": It's really the cultural challenge I think, for GPs to understand what it's like to be commissioners. The statutory responsibilities you have, you know, there are things that you must do, should do, and them understanding, I think, some of the legal frameworks and constraints. And then for managers I think the challenge has been around the shift of culture, becoming more clinically orientated . . . and for some managers that's quite a challenge because they have little clinical awareness. (Manager ID 9) At a more pragmatic level, there was concern about GP capacity to take on commissioning and managerial functions: . . . they don't always understand the pathways, which is surprising-you realise how little some of them know about hospital-based care. As a manager, sometimes you think hmm, maybe, physiologically I may not understand it, but I understand the system and pathways and I understand what the cardiologist does. (Manager ID 9) There was also a feeling among managers that GPs are first and foremost clinicians and that clinical work is something they can always "fall back on." For managers, though, there is no fall-back position.
Running parallel with these concerns are worries about changing employment structures and associated job losses. As PCTs neared the end of their existence, managers were being assigned to work in CCGs, in Public Health in the Local Authority or in the newly formed CSUs. In addition, some faced redundancy and job loss and many were undergoing processes of having to apply formally for their posts in the changed structures. For those joining one of the CSUs in England, there is the prospect that these organizations will potentially no longer be part of the NHS after 2016. This manager explained how some PCT colleagues had been assigned to the CCG while others had been assigned to a CSU: So they're quite happy [those assigned to the CCG], I think, really. The rest of the people in the PCT probably aren't quite as happy because it's this, are you going to be a commissioning support organisation, and I think the connotations around that are that eventually . . . the NHS Commissioning Board [now NHS England] say that they'll host it till 2016 and then potentially they'll be privatised, or could be, and I think for individuals that presents a challenge to them in terms of their employment status, job security, etcetera . . . Because people want to remain in the NHS. (Manager ID 287) Assignment to a CSU means that managers may lose their connections with their immediate local community. This has implications for both managers and CCGs, with the latter voicing concern about losing local knowledge.
On a personal level, job losses and job insecurity place a heavy burden on individual managers, including those who have to manage these changes.
In spite of upheaval and uncertainty, many managersparticularly those in senior positions-still voiced enthusiasm about the possibilities of future partnership working with GPs. Evident too were high levels of commitment to the ethos of public service and to patient care. I think, some of the clinical engagement that's come out of it: massive buzz for me, because, if you get people with fire in their belly and they get going, oh, it's fantastic . . . at the end of the day, and, you know, I drive people mad about this, but, we come to work, because, there's patients that need help. They need healthcare and it's incumbent on us to make, you know, best use of the tax payer's pound. (Manager ID 117) This suggests that the new opportunities offered to some managers in the new system had in fact strengthened their sense of themselves as public servants, working more closely with clinicians to provide high quality health care.

Discussion
This research was undertaken during a period of intense and major change. These changes and their implications are still unfolding and it is too soon to talk definitively about their results and outcomes. Notwithstanding, it may be possible to discern the "direction of travel" brought about by the current reforms. Periods of fundamental change can force into consciousness notions of identity and relationship to work and peers that are normally taken for granted: the cognitive-cultural mental scripts referred to by Scott (2008). This research has captured some of the issues surrounding role and identity change currently being faced by NHS GPs and managers.
Our findings resonate with previous literature in showing that both GPs and managers (perhaps with different emphases) view patient care as their raison d'être and have a strong public service ethos. These are the ties that bind them to the institution of the NHS. As Scott points out, we cannot presume that institutions simply persist and that the beliefs and practices associated with them just endure. He argues that changes in the three constitutive pillars of institutions interact with one another and that there may be times when the pillars are "misaligned" and may "support and motivate differing choices and behaviors" (Scott, 2008, p. 62). It is at such moments that institutional change occurs. What our data suggest at this point of the reforms is that radical change taking place in one of the organizational pillars may not necessarily be followed by changes in the other two pillars, thus leading to institutional dysfunctionality.
In Scott's terms, the current reforms constitute an obvious instance of regulative change. Two hundred-eleven CCGs in England now have full responsibility for spending £60 billion of public money. It could be argued that these changes have much in common with previous attempts to reform the NHS. We would contend that the current reforms go much further than their antecedents. In contrast to the predecessor commissioning organizational forms of GP Fundholding and PBC, CCGs have wider commissioning responsibilities; and unlike the earlier schemes of Total Purchasing and PBC, CCGs control and spend real money . A further marked contrast with earlier GP commissioning experiments is that CCGs are membership organizations where membership is not voluntary but compulsory. GPs are formally in control of CCGs with no buffer organization replacing PCTs or their various predecessors.
The responsibility for commissioning spending, including taking unpopular or difficult decisions in the context of squeezed resources, will also now fall to CCG leaders. In Scott's terms, this means an inconsistency or instability in the normative organizational pillar. The norms and values that GPs uphold in their roles as "frontline" clinicians are different from their roles as CCG leaders, managers, and budget holders. At its simplest level, GPs in their clinical role view the needs of the individual patient in front of them as paramount. Our data suggest that GPs in managerial positions in CCGs must marry this with a population-level concern with budgets, priorities, and possible service cuts. GPs in our study showed themselves to be acutely aware of this tension. At best, some argued that their very closeness to patients would ensure that they made good decisions as commissioners; others expressed a concern that the need to engage with burgeoning budgetary pressures would undermine their relationship with their patients and, by extension, their sense of identity as clinicians. These role and identity tensions faced by GPs taking on managerial roles are not completely new, but we argue that they are greater than under previous GP commissioning models. These roles have the potential to change their relationships with peers as they will now take responsibility for their performance management in light of devolved budgets and the need to attain savings targets. The use of the term frontline to describe the position of GPs is significant. It has been used as shorthand to refer to GPs' proximity to their patients and to the concerns of their patients. It is clear that GPs identify with the notion that they are patient facing; it is less clear that they fully embrace the norms associated with their new roles as managers and commissioners. They certainly do not take it for granted that all GPs by virtue of CCG membership have a stake and responsibility for commissioning.
The issue of the time and resource commitments required for GPs to take on roles in CCGs is not a trivial one. At the individual level, it also calls to question normative understandings about the role of GPs. It also reaches into general practices drawing on the expensive time of experienced GPs. So while "rank and file" GPs may not understand CCG membership as changing the nature of their clinical role, the absence of practice members is felt. The need to cover these absences might be seen as disrupting the cognitive-cultural assumption underpinning general practices as small businesses. In relation to secondary care, GPs will have increased opportunities to play a major role in contract negotiations also potentially shifting elements of the cognitive-cultural understandings of what it means to be a GP.
The data presented above already point to a distinct uneasiness concerning the roles of GPs as CCG board members who are responsible both for commissioning and decommissioning of services. It also points to a reluctance of many GPs to take on formal commissioning roles. It remains to be seen whether these shifting roles will become normalized, that is, constitute change in Scott's cognitive-cultural pillar; will GPs and those they serve come to understand that they are commissioners and that this is part of their professional identity? Either way, we would argue that such far reaching change obliges those involved not only to consider their changing roles but also what these changing roles mean in terms of identity. Such shifts in the cognitive-cultural pillar of institutions take time to discern, and the data we have presented do not yet present a clear story of cognitive-cultural change. Indeed, while some GP respondents highlight their desire to cling on to a strong clinical identity, and argue that this is essential if they are to fulfill their new role, others voice their discomfort at the tensions they experience, with some managers suggesting that GPs will "need to change." Overall, this suggests that the significant regulative changes that we have seen are in the process of destabilizing longheld cognitive-cultural assumptions about "who we are" and "what we do." It is not possible at present to suggest what the eventual outcome will be, but it seems likely that there will be at least some shift in the nature of the institution that is the English NHS.
The managers supporting CCGs are also in the midst of another round of ebb and flow of their power and authority. Many of the managers that we observed and interviewed were in the process of moving into different organizations set on different pathways. Some were being made redundant and others had chosen to leave ahead of the changes meaning the loss of talented and experienced people to the NHS. Many of our respondents pointed out that managerial teams are being fragmented with concomitant loss of organizational memory and of connection to local communities. For others, their leadership roles have changed as GPs are set to lead the new CCGs. There has clearly been a regulative shift seeking to subordinate managers to doctors; whether this change will spread to the other pillars, thus recreating "diplomatic management" as described by Harrison (1988), will be interesting to watch. A key question to follow will be whether these changes affect managers' identity as public servants who are part of the NHS family.
The question that our data raise is whether the current changes may go further than creating shifting roles for some GP and managers. Might these changes erode what Scott refers to as the cognitive-cultural pillar so that some GPs and managers begin to relate to the NHS in different ways? Many of the issues that our interviewees raised were reflections on changing identity, alongside their changing roles. The previously stated rationale for bringing GPs into commissioning was their proximity to frontline patient care, their standing in the community, and their understanding of local health care needs. Our findings suggest that engagement in managerial work may potentially take GPs away from patient care, may alienate them from their peers, and may, in time, make them unpopular with patients. GPs who have undertaken roles in their CCGs could opt to retreat to their "day jobs" leaving commissioning in the hands of the few. Early evidence appears to support this contention, indicating that GP representation on CCG governing bodies is falling away (Kaffash & Mooney, 2014). It seems likely that those who stay on will see themselves and be seen by others as being different from "ordinary rank and file GPs." These GPs will be taking on new identities as "GP commissioners," and it is not yet clear whether they will be able to maintain the same relationships with their patients and with their colleagues.
The changes that we observed as CCGs are becoming established are still unfolding, and the questions posed here will require further time to be answered definitively. In particular, can a good commissioner be a good GP? How will budgetary accountability affect relationships between GP leaders, their patients, and their colleagues? Will managers be able to retain their public service ethos if they no longer work in their local community or no longer work directly for the NHS? Will, as this initial research suggests, these changes serve to change the ties that bind them to the NHS?