Physical coercion

States practically monopolise physical coercion as a source of power, but in liberal democracies the ways it is used are codified as a body of law and para-legal rules (regulations, contracts, judicial interpretations and so on). Direct applications of physical force are exceptional. Rather, the state always retains the option to legislate on any given question, and then a permanent threat of physical coercion is introduced. Courpasson describes governance by tacit appeal to the threat to legislate as ‘soft coercion’ (pouvoir doux) (Courpasson, Reference Courpasson1997, Reference Courpasson2000).

Economic constraint

As a market agent, the state is a major provider and purchaser of goods, services and labour; and for welfare services, the largest. Within the public sector, financial incentives are the mundane source of managerial power.

Ideological ex-communication

Ideologies, says Therborn, define what exists; what is possible (hence, what range of political action is feasible); and what is right or wrong. While legitimation is indeed one way in which ideologies sustain power (Therborn, Reference Therborn1980), Therborn like others (Habermas, Reference Habermas1976; Lukes, Reference Lukes1974) allows that power is also exercised by persuading others to do what they would not do if they were more widely informed, more confident or had a different self-identity. Citing Foucault (Foucault, Reference Foucault1980), Therborn describes three procedures by which power-holders shape discourse so as to promote obedience and ‘ex-communicate’ the disobedient. ‘Restrictions’ stipulate who is qualified (or not) to pronounce upon a given subject and what may be talked about (or not). ‘Delimited appropriation’ means that a discourse is only used in approved settings. A preferred discourse is also ‘shielded’ against alternatives in various ways, including organising it as a ‘discipline’. Standardised discursive procedures (methodological rules, working practices and so on) which professionals internalise during their training (Flynn, Reference Flynn2002) subsequently regulate both the technical and the social components of their work, functioning as a ‘discipline’ in both an intellectual and a micro-political sense.

While economic and physical sanctions often reinforce them, the exclusionary aspect of these processes is itself a sanction. Earlier theorists argued that the simplest form of agenda control is to exclude specific topics from policy and managerial negotiations and decision-making, even though the negotiators are aware of them (Bachrach and Baratz, Reference Bachrach and Baratz1970). ‘Non-decision-making’ occurs when one agent can impose the use of a discourse or ideology which does not even allow the formulation of certain topics (for example, the merits of nationalising profitable firms, in New Labour discourse before the 2008 credit crash), which become not only undiscussed but practically undiscussable (Lukes, Reference Lukes1974). Fairclough (Reference Fairclough2005) describes a sequence of emergence, hegemony, recontextualisation and operationalisation by which a new discursive order replaces alternatives. Therborn adds that by foreclosing (framing) the range of decisions and actions from which to choose, what issues count as politically ‘relevant’, and what demands and forms of expression are admissible, these devices perform an ‘ideological ex-communication’ of dissenters from institutional debate, negotiation and decision-making. By implication, the same applies at occupational level; besides its positive stipulations a professional ‘discipline’ also ex-communicates ‘unprofessional’ practitioners.

The state applies these sanctions in two main ways. One is through the selection, recruitment and ideological formation of those who hold official positions in public bodies. Policy-makers frequently delegate part of the state's powers to particular occupational groups, unwittingly enabling the latter at times to ‘capture’ or co-opt those powers for themselves. Nevertheless, staff recruited from one occupational group may at times implement policies which benefit another. (For example, medical managers might implement policies which reduce doctors' power in favour of some other group.) Such events symptomatise shifts in power from one social group to another.

Also, the state mediates and attempts to manage conflicts between different social groups. This element of Therborn's model thus accommodates Strauss' idea that the loci of control between occupational groups, who will exercise that control and by what means, typically constitute a ‘negotiated order’ (Strauss et al., Reference Strauss, Schatzman, Ehrlich, Bucher, Sabshin and Friedson1963): that is, an agreed re-allocation of roles, rights and resources. In health systems the most important such negotiations usually concern ‘industrial relations’ between other occupational groups and managers; and the jurisdictional claims of occupational groups which claim overlapping domains of work (for example, psychiatrists and psychologists) (Strauss et al., Reference Strauss, Schatzman, Ehrlich, Bucher, Sabshin and Friedson1963; Light, Reference Light and Albrecht2000; Beck, Reference Beck2000) or seek to monopolise an activity. How far non-managerial occupational groups resist managerial power, by what means and to what effect, is a contingent (‘conjunctural’) matter. The post-1950 history of British industrial relations suggests that the likelihood of resistance depends partly upon how confident each party feels in the light of past attempts to apply sanctions; how severely threatened they feel their interests to be; how they anticipate any powerful third-party might respond; and how far they regard it as legitimate to apply sanctions. A common negotiating tactic is for each party to exaggerate the sanctions it possesses, including potential harm to others (for example, by ‘shroud-waving’).

However, the most important sources of power are, according to Therborn, ‘impersonal’: a group's place in the economic structures of a society. De facto ownership of productive resources (physical goods, services, labour, knowledge, money) confers on their owners the sanction of being able to withdraw resources on which the other party depends but does not own. In liberal capitalist societies these impersonal sources of power exist more at the organisational than state level, because by preference the state refrains from acquiring them. Nevertheless, even in the public sector, control of physical goods, knowledge and money is the main sanction underpinning managerial power.

For non-managerial groups, this kind of power depends on how vulnerable managers are to the group disrupting production of the organisation's main activity or output. The more alternative providers exist, and the more the managers of a given organisation risk permanent economic losses if an occupational group disrupts production, the greater is that group's power. The smaller the excess supply of labour in that occupation, the harder it is for managers to replace non-compliant members and the weaker a sanction is the threat of dismissal. The limiting case is an organised monopoly in labour supply (‘closed shop’, ‘occupational closure’) from which managers cannot get individual members to defect.

Technical constraints also make a whole occupational group, rather than just its individual members or a particular provider, substitutable or not. An occupational group's economic power depends above all on its relationship to the technology (physical equipment and processes, and the corresponding technical knowledge) used for conducting the core activities of the organisations for which they work (Abell, Reference Abell1975). Here the concept of ‘technology’ is defined widely to include, for instance, ‘organisational technology’ (Therborn, Reference Therborn1978: 39), which could consistently include ‘technologies of repression’ (Foucault, Reference Foucault1977). Such power depends on how far substitutes exist for:

• • The occupational group which operates a given technology. The fewer alternative occupational groups who can operate a given technology, the more powerful is the occupational group operating that techology.

• • The technology itself. The fewer the technologies which can replace it, the more powerful are the occupational groups who can operate that technology.

Since this power derives from production technologies, it is, Therborn implies, typically exercised not through hierarchical control, but through informal organisation (for example, ‘direct’ action) or cross-organisational institutions (for example, professional bodies, trade unions). Nevertheless, insofar as an occupational group's technical knowledge and working practices can be made transparent to other group members, hierarchical superiors or other outsiders, those who scrutinise the occupational group members' activities can apply sanctions to promote practical compliance with that discipline and with their own interests.

Through all the above structures, processes and institutions, the exercise of power is a zero-sum game, where the prizes are the allocation of activities, economic rewards, formal position, status and other perquisites, and of the future means of exercising power.

Law, regulation, contract

In the period studied, the basic legal framework for professionals' clinical accountability changed little. The most important legal development was to make GP contracts more complete and contestable. The 1967 GP contract was little more than a reimbursement mechanism, but from 1990 each revision specified GPs' clinical work more closely. The 1990 contract introduced vaccination and immunisation targets and delegated, ‘indicative’ budgets for pharmaceutical spending, and medical audit. GP fundholding added responsibility for budgets for much of secondary and some primary care − though not for GP medical care itself, except prescribing. GP co-operatives providing out-of-hours services were first permitted under the 1990 GMS contract: 98 existed by 1996, and the 68 per cent of them who responded to a survey (Jessopp et al., Reference Jessopp, Beck and Hollins1997) had 11.4 million patients. From 1997, they provided wider ranges of services until the 2004 GMS contract produced a sharp retrenchment.

From 1998, GPs had the option of taking a locally negotiated Personal Medical Services (PMS) contract rather than the standard national General Medical Services (GMS) contract. A content analysis of 71 (84 per cent) of the first wave of PMS contracts showed that PCGs set practice-specific targets, though usually broadly defined (for example, in terms of hours of service, not protocols defining quality of clinical care) and weakly monitored (for instance, not one stated how ‘exception reporting’ of problems would occur) (Sheaff and Lloyd-Kendall, Reference Sheaff and Lloyd-Kendall2000). Both types of contract now defined the contractor as the general practice (organisation), not named individual doctors as before. Such a contract therefore survives changes in medical personnel including, in theory, the ousting of individual GPs for disciplinary or other reasons. The 1997 NHS Act relaxed the conditions under which PCTs and general practices were permitted to employ salaried GPs. Above all, it made GP contracts contestable by permitting alternatives to the existing model of general practice, so ending independent general medical practices' legal monopoly of primary medical care.

Spending on primary care

At national level, state spending on the NHS also rose by about 7 per cent per year during 2002–06 (Maynard and Street, Reference Maynard and Street2006), almost certainly faster than non-NHS sources of GP income. In the 1990s, NHS primary care managers did little more than reimburse general practices through a complex, rigid payments schedule. The aforementioned legal changes greatly enlarged their fiscal capacity to manage primary medical care.

A Quality and Outcomes Framework (QOF) supplemented the 2004 ‘new GMS’ contract for GPs, coupling about 20 per cent of general practice incomes to compliance with 136 clinical indicators. A large survey of GPs just before contract implementation showed 80.2 per cent anticipating (it was too soon for them to report actual experience) that the new contract would reduce their professional autonomy, 93.8 per cent expecting increased administrative workloads, 87.6 per cent expecting increased clinical workloads and 54.4 per cent expecting a reduced quality of working life (Roland et al., Reference Roland, Campbell, Bailey, Whalley and Sibbald2006). There were also reports that some (but not all) GPs felt these incentives rewarded what the doctors anyway regarded as good clinical practice (McDonald et al., Reference McDonald, Harrison, Checkland, Campbell and Roland2007; Roland, Reference Roland2007). So complete was reported GP compliance that during 2006–7 general practices collectively earned 95.5 per cent of the maximum possible QOF payments (http://www.precedent.co.uk) of approximately £42,000 per GP gross of any additional operating costs.

Under the 1997 Act, a succession of schemes (Primary Care Act Pilot Schemes; Primary Medical Services (PMS); Additional Provider Medical Services (APMS); Specialised PMS, Innovations in Primary Care Contracting) permitted NHS managers to commission non-general practice primary care services. Probably fewer than ten nurse-led general practices (in which the owning partner is a nurse not a doctor) were commissioned, but by 2005–06, 2,743 (8 per cent of) GPs were partly or wholly employed on a salaried basis (The Information Centre, 2007). Two-thirds of the first-wave PMS contracts left clinical services unaltered (Sheaff and Lloyd-Kendall, Reference Sheaff and Lloyd-Kendall2000), but managerial appetites for local service enhancement under these schemes were greater at national level. In 2006, a first attempt was made to replace a general practice with provision by a US for-profit HMO. With varying success, ten pilot site PCTs have been seeking and evaluating bids from non-general practice primary care providers. Lord Darzi's proposal (Darzi, Reference Darzi2008) to set up ‘polyclinics’ became controversial because it too proposed allowing organisations other than independent general practices to provide these services. The other discretionary area was investment. The Local Improvement Finance Trust (LIFT) scheme instituted joint ventures between the NHS and the private sector as the main channel for investment in primary care, creating another route for PCTs to commission organisations other than general practices to provide primary medical care. Nevertheless, general practices often participated within the ‘private’ element of LIFT schemes and within the commercial organisations which PCTs began to commission. One might speculate (published evidence is lacking) that for GPs one attraction of doing so was the apparent independence of PCTs, but only as the entrepreneur would a GP so benefit. Commercial providers appear, if anything, to manage their medical employees more closely than PCTs do. While independent general medical practice remains the predominant organisational model of NHS primary care provision, NHS managers now have discretion to use others.

Health Authorities had from 1991 the power to delegate fundholding budgets for the purchase of hospital care to general practices. Non-fundholding GPs' referral rights were limited to hospitals with which their Health Authority had made a service contract, but a system for ‘extra-contractual referrals’ allowed exceptions where clinical grounds could be shown. The 50 per cent or so of GPs who eventually became fundholders received greater budgetary discretion, including discretion to buy their patients care at non-NHS hospitals. The corollary was financial accountability to (then) Health Authorities for the use of these budgets. Practice-based commissioning brings the same discretion and concomitant accountability to all GPs, although in reality general practices appear to differ in how far they actually take up this delegation of financial discretion and responsibility (Checkland et al., Reference Checkland, Coleman, Harrison and Hiroeh2008; Coleman et al., 2007).

Ideological ex-communication

With decreasing lags, the ideological climate in NHS management tended to follow that of the current government. Despite the change of party in government, the deepest policy difference between them – whether to contain or radically expand the level of NHS spending – was not strongly reflected in NHS managerial ideologies. Verbal changes in health policy documents accentuated the political differences, but the substantive ideological continuities are more apparent. Labour policy reiterated a claim likely to be persuasive to doctors: that health professionals, especially doctors, represent patients' interests. One legitimation for each successive NHS reform has been that it will allow clinicians to practise as they judge best and to influence service management. Another continuity is the tendency to legitimate change for change's sake: ‘change’, ‘new’, ‘reform’, ‘modern’ and the like are words of approbation. New Labour especially emphasised its ideology of ‘no ideology’, encapsulated in the slogan ‘what matters is what works’ with its implications of normalising ‘what works’ and promoting evidence-based practice.

Both GPs and managers promoted evidence-based-medicine (EBM). A study of 61 primary care innovations during 2002–04 found that EBM had already become a ubiquitous discipline; not one informant spoke against it. Moreover, about two-thirds of these innovations actually were evidence-based, albeit on a generous definition of ‘evidence’ (Sheaff et al., forthcoming). GPs often chose which aspects of EBM to adopt (Checkland, Reference Checkland2004; McDonald et al., Reference McDonald, Harrison, Checkland, Campbell and Roland2007). Another multi-site study of PCTs suggested that EBM was more readily applied to, and hence legitimated, the biologically oriented aspects of clinical practice (Rogers et al., Reference Rogers, Campbell and Gask2002). Mental health services tended to elude clinical governance (Gask et al., Reference Gask, Rogers, Campbell and Sheaff2008).

The discipline of evidence-basing spread – indeed, partly on medical instigation (Muir Gray, Reference Muir Gray1997) – from clinical work to the legitimation of organisational models of care (Harrison, Reference Harrison2002). To varying degrees, National Service Framework (NSF) standards derived from evidence-based medicine and the national bodies (above all, NICE) through which the NHS implemented it. Besides adopting evidence-basing, NHS primary care management experienced a disciplinary shift towards the ‘new public management’ practices that NHS hospitals started a decade earlier (Flynn, Reference Flynn1992; Ferlie et al., Reference Ferlie, Ashburner, Fitzgerald and Pettigrew1996). Primary Care Trusts were organisations intended, unlike their predecessors (FHSAs, DHAs, HAs), to manage rather than just reimburse GP activity, in particular to implement national standards for services. Clinical quality and access targets were now applied to general practices, although ‘advanced access’ methods were usually implemented through network-based primary care ‘Collaboratives’ rather than by PCTs alone.

At PCT level, constant organisational change allowed GPs' leaders to exercise ‘soft’ coercion within their profession. A 12-PCT case study found that GP leaders perceived that if doctors did not strengthen their management of the quality of clinical work, government and NHS managers might intervene instead (Sheaff et al., Reference Sheaff, Sibbald and Campbell2004). On that basis, local GP leaderships (see below) were able to legitimate their own new role as one of buffering rank-and-file GPs against such managerial ‘encroachment’. The argument that (say) NSFs are ‘national policy’ also helped legitimate them to GPs (Sheaff et al., forthcoming), although this mode of legitimation is more attuned to a managerial than to ‘free professional’ ideology. Thus, GPs appear to assume that it is legitimate for NHS management to influence general practice, provided this influence is mediated through GP leaderships at national and local levels. GPs reconciled this view with their preference for professional autonomy by saying that the policy coincides with local health needs (Sheaff et al., Reference Sheaff, Rogers and Pickard2003).

Which viewpoints were thereby ‘restricted’? Apart from exceptional events such as gross malpractice or criminality, the topic of how doctors exercised their clinical autonomy was closed to NHS managers until the late 1980s. The relaxation of this restriction, especially at local level, is the most striking change in patterns of ‘ex-communication’ since 1991. Individualistic conceptions of professional autonomy have gradually become ‘restricted’, including the beliefs that the clinical quality of general practice is sufficiently maintained by the exception management of clinical errors only, and that, gross malpractice or criminality aside, the quality of a GP's routine clinical practice was just her concern. One gap in existing research is a detailed analysis of how often, and in what ways, the discussion of clinical practice has entered the agendas of NHS decision-making bodies, although the development of management information systems (see below) is some corroboration. GPs' uptake of it suggests, too, that salaried employment is no longer an excluded option for mainstream general practice.

Restrictions upon managerial influence over entry and exit to medical careers also relaxed. During 2006–08, control of postgraduate medical education was transferred to a national board, answerable to the General Medical Council (GMC; whose own composition was also changed to include more managers) and having a large minority (9/20) of managerial members. Medical opposition (Independent Enquiry into Modernising Medical Careers, 2008) changed the educational proposals, but not the new management regime. Earlier (2002) arrangements for identifying and ‘helping’ GPs whose ‘performance gives cause for concern’ (Department of Health, 2000) were managerially instigated, although the systems remained professionally operated and predominantly educational.

Generally, medical ideology has posed little substantial challenge to the NHS reforms since the middle 1990s. The most coherent counter-ideology contrasts evidence-based medicine (see below) with the ‘art’ and ‘holism’ of general medical practice. Certain topics, however, remain off managerial agendas. According to the Professional Executive Committee chair in a large provincial city:

The taboo appears little changed since this remark in 2002, although global data on GPs' earnings are now somewhat more freely available. English GPs have remained sensitive that new working methods should be adopted voluntarily, not imposed by non-doctors, regarding closer regulation by fellow GPs as a lesser evil than direct managerial control.

Official positions

At national level, the medical profession's capture of decision-making bodies has weakened. Before 1991, the Department of Health merged its former distinct parallel apparatus of medical managers into its general management structure and, as noted, strengthened lay influence over medical careers.

At local level, NHS restructuring created new medical-managerial roles. In 1998, GPs were given a majority on Primary Care Group (PCG) boards and the chair. By 2002, however, Primary Care Trusts (PCTs) had replaced them. GPs were guaranteed representation on PCT Boards, but most Board members were centrally appointed non-GPs. Former PCG Boards became Professional Executive Committees (PEC), which, although still medically dominated, were now subordinate to the new PCT Board. PCTs typically nominated GPs as ‘leads’ for monitoring or developing particular areas of clinical work (for example, mental health), frequently buying out that time from the doctor's own practice. These links between PCTs and general practices were supplemented with others such as locality groups, pharmaceutical and other advisory structures, typically but not always with GP leads. In many localities, a stratum of about 10–20 per cent of leading GPs appear consistently to have been early adopters of the clinical practices and models of care, and of organisational innovations (fundholding, PMS contracts) and roles (for example, PCT committee membership). For some years, there has been an increasing tendency for English hospital doctors to occupy managerial positions, but a study of four PCGs in 1998–2001 described a stratum of doctors emerging who mediated between the profession and NHS management (Sheaff et al., Reference Sheaff, Smith and Dickson2002), and thus acquired an interest in reforms which enhanced their status, influence over other doctors and voice in NHS management. The universal but relatively weak, predominantly educational form of medical audit introduced in 1991 was from 1998 supplanted. Medical audit and local professional networks were reconstituted as clinical governance networks with GP ‘leads’. GP membership was compulsory. These changes shifted the emphasis of medical self-regulation from private and individual towards transparent and collegial scrutiny of medical work (McDonald et al., Reference McDonald, Harrison, Checkland, Campbell and Roland2007) based upon EBM. Nevertheless, clinical governance was often initiated by a core of ‘early adopter’ doctors, gradually involving other GPs through peer influence later and, as noted, GPs' uptake of EBM was selective. Other clinical professions, especially nursing, either had their own more fragmentary clinical governance networks or participated ad hoc as semi-detached members of the medical networks on an issue-by-issue basis. Clinical governance activities thus remained medically dominated.

For salaried GPs, the differences in patterns of employment between doctors and other primary care clinicians have narrowed. Salaried GPs have become less of a minority, partly because salaried practice more readily accommodates short-term work as a GP, part-time working and career shifts between general practice and other activities. The economic changes described below have also sharpened the division of clinical labour. Slow, partial moves towards patterns more characteristic of hierarchical governance structures have thus occurred alongside the development of existing contractual and network governance structures at PCT level.

PCT general managers have been recruited increasingly from the general NHS managerial workforce, most with experience of NHS hospitals where, at least nominally, general managers hold authority over doctors (Harrison and Ahmad, Reference Harrison and Ahmad2000). By 2008, managerial recruits from outside the NHS were numerous enough to warrant dedicated training programmes. Their career experience did not include taking a relatively passive, ‘diplomatic’ role towards clinicians.

Conflict management

Many doctors resisted the 1991 NHS reforms. The British Medical Association (BMA) ran a public campaign of advertisements, press conferences, lobbying of MPs and releasing research study findings supporting their standpoint shortly before the main parliamentary debates. Nevertheless, the reforms were implemented, including a revision of the GP contract to include the first measures of clinical activity and indicative prescribing budgets. Medical audit was introduced following the tacit threat that if the medical profession did not cooperate, the government would legislate (NHS chief executive; personal communication). Later a succession of widely publicised medical scandals (in general practice, the Shipman case) created a perception among GPs that if the profession did not adopt more rigorous and transparent quality and safety controls, national-level medical organisations – in particular, the GMC – would be reformed, with the tacit threat of a stronger non-medical voice in them (Klein, Reference Klein1998). By this stage, the stratum of doctors mediating at national level between other doctors and NHS management was on the defensive, lacking confidence to take on the state (Armstrong, Reference Armstrong2002). Initially, the Labour government displayed less negotiating confidence than its predecessor. As the political price for participating in PCG/Ts, GPs were guaranteed a majority on PCG boards and PCT professional executives. On generous terms, the government pre-emptively ‘bought out’ potential GP objections to the closer monitoring of medical work under the 2004 GP contract and QOF. GPs also played a central role in negotiating the QOF contents, partly because first-hand clinical knowledge of evidence-based medicine was required to establish which clinical indicators would be scientifically valid and credible to GPs. To that extent, GPs influenced the choice of technical areas in which they were, and were not, to be managed. In early 2008, however, the government proposed to increase payments to general practices which increased their opening hours and reduce payments to those who did not. Although the basis of payment was re-negotiated, the increased opening hours were introduced from April 2008. As noted, the nature of medical education has been re-negotiated, but not managerial involvement in controlling it.

Local Medical Committees (LMCs) were generally passive or accommodating to PCT formation, clinical governance and successive forms of GP contract. None of three multi-case studies, together covering 25 PCTs during the period 1998–2005, found much evidence of LMC influence. Although a steady trickle of individual jeremiads has appeared in the professional press, and although certain researchers report certain GPs' passive resistance and skepticism towards the ‘reforms’ (McDonald et al., Reference McDonald, Harrison, Checkland, Campbell and Roland2007; Roland et al., Reference Roland, Campbell, Bailey, Whalley and Sibbald2006; Sheaff et al., Reference Sheaff, Sibbald and Campbell2004), at local level sustained opposition has been rare and not on the scale of the 1990s.