Britain's National Health Service (NHS) is a universal, single-payer health system in which the central state has been instrumental in ensuring equity. This article investigates why from the 1970s a policy to achieve equal access for equal need was implemented. Despite the founding principle that the NHS should “universalize the best,” this was a controversial policy goal, implying substantial redistribution from London and the South and threatening established medical, political, and bureaucratic interests. Our conceptual approach draws on the advocacy coalition framework (ACF), which foregrounds the influence of research and ideas in the policy process. We first outline the spatial inequities that the NHS inherited, the work of the Resource Allocation Working Party (RAWP), and its new redistributive formula. We then introduce the ACF approach, analyzing the RAWP's prehistory and formation in advocacy coalition terms, focusing particularly on the rise of health economics. Our explanation emphasizes the consensual commitment to equity, which relegated conflict to more technical questions of application. The “buy-in” of midlevel bureaucrats was central to the RAWP's successful alignment of equity with allocative efficiency. We contrast this with the failure of advocacy for equity of health outcomes: here consensus over core beliefs and technical solutions proved elusive.
The British National Health Service (NHS) is one of the oldest examples among the liberal democracies of a single-payer, publicly funded health system. It was launched in 1948, following legislation in 1946–47, with three core principles: it would be universal in coverage, furnish a comprehensive range of services, and be free at the point of use, with funding coming principally from general taxation levied centrally. According to its founder, Britain would become “more wholesome, more serene, and spiritually healthier, if its citizens have the knowledge that they and their fellows, have access, when ill, to the best that medical skill can provide” (Bevan 1952: 75). Over its seventy years these principles have frayed at the edges, with the status of non-British citizens compromising universalism, the porous boundary with social care challenging comprehensiveness, and the introduction of charges for prescriptions, dentistry, and ophthalmics undermining free treatment. Nonetheless, the NHS's principles and values remain essentially intact and politically popular (Gorsky 2008).
As momentum in global health builds behind the agenda of universal health coverage, this “Beveridge” model health system is enjoying a revival of interest among analysts. Pluralist developmental models incorporating substantial user fees have proven to impose barriers to access, while empirical evidence that public systems deliver better outcomes in low-income settings has emerged (Yates 2009; Moreno-Serra and Smith 2015). In these circumstances, the NHS can reasonably be proffered as “a highly applicable . . . means for effectively financing a universal coverage system providing access to cost-effective care,” while scoring “consistently high on international benchmarking comparisons . . . especially on equity” (Chalkidou and Vega 2013). The NHS funding model and related devices through which a centrally administered system achieves efficiency and equity therefore merit scrutiny.
The subject of this article is one key device, commonly dubbed the RAWP. This awkward acronym refers to the Resource Allocation Working Party, a committee of the government's Department of Health and Social Security (DHSS), established in 1975 and reporting in 1976. RAWP is also synonymous with a novel formula introduced by the committee, through which the state disbursed funding to the NHS regions. The formula's guiding principle was that need for medical care in a given population could be systematically calculated, allowing resources to be allocated in a fair and transparent manner.
There are two reasons that the RAWP's history has larger relevance to present-day debates about health policy. The first is as a case study of equity of access in the policy arena. At the international level the idea that the furtherance of equity is a legitimate health system function emerged in the 1970s and subsequently became commonly recognized (Anderson 1972: 81, 93, 161–65; World Bank 1993: 54–55, 69–71). This acceptance may be understood as a political expression of ethical principles, whether rights driven or paternalistic, or of individual self-interest in managing risk. Simply defined, equity implies that all should have access to health services regardless of income or residence, yet in practice it has been understood mostly in terms of completeness of coverage or fairness of contributions (Anderson 1972: 5; WHO 2000: 35–39). However, the historical experience of the British NHS shows that, even when universal coverage through progressive taxation is established, other dimensions of equity, such as utilization, health outcomes, or geographical access, may remain unsolved (Hollingsworth, Hage, and Hanneman 1990). The RAWP episode illustrates how the problem of equity of access by residence came to be articulated and addressed within a single-payer system.
The RAWP's history is important, second, as a historical case study of innovation and success in health policy making. Politically, the terrain of equity is highly contentious, even for a country like Britain where “socialized medicine” seems a settled aspect of public life. On the one hand, there have always been critics of the NHS who have objected from libertarian or economic liberal perspectives (Seaton 2015). On the other hand, redistribution of resources for health, albeit in the name of social justice, creates both winners and losers. As in any health system, interest groups such as medical professionals and hospital administrators may be expected to object if their power is threatened (Alford 1977). There are also formidable technical challenges to successful policy making in this area: how exactly is need for medical care to be defined and measured, when both biological and social factors are in play? Given these political and practical impediments then, it is surprising that the RAWP succeeded at all, and this makes it a particularly intriguing case study.
The discussion begins with an introductory description of Britain's NHS and then provides essential details of the RAWP and its context. Next, it reviews existing literature on the episode, arguing that this has given insufficient weight to the importance of ideas and of actors below the level of political leadership. The following section outlines a conceptual approach drawn from political science that adopts just such an analysis: the advocacy coalition framework. This is a generic model of policy change whose explanatory power helpfully illuminates the RAWP case study. The argument that follows emphasizes three main factors of change: the importance over the medium term of policy learning driven by the research community; the key role of midlevel bureaucrats in supporting implementation and embedding of the initiative; and the framing of the RAWP debate as essentially technical, even though it touched on core values and was potentially controversial.
The British NHS: History and Structure
Among the myriad classificatory schemes used in health systems analysis, Britain's NHS has historically been viewed as an ideal type for its universal service pattern of free care as a public benefit and its polar organizational model, in which the state is the dominant payer and provider (Roemer 1960: 158; Anderson 1963: 842). It was established after World War II as a key element of the welfare state, inspired by the universalist blueprint of the Beveridge Report, and put into place by the social democratic Labour government following its 1945 election victory. The arrangements that the NHS replaced were characterized by localism and diversity (Webster 1988). Medicine for the middle class had been predominantly private, while payroll-based health insurance covered blue-collar workers for primary care. Voluntary hospitals funded by philanthropy or mutualism dominated acute care for the working class, with psychiatric and long-stay institutions provided by local government or the Poor Law. Municipal public health departments oversaw infectious diseases, clinical services for women and children, and preventive care. All this was swept aside in 1948, with hospitals taken into national ownership and staffed by salaried doctors and nurses, and primary care physicians (general practitioners) employed under contract by the NHS. In place of pluralist funding sources, income now came principally from progressive national taxation dispensed annually by the UK Treasury, with private medicine permitted but marginal.
The administrative base of medicine in charitable and local political structures was also replaced under the NHS by a hierarchical system (Webster 1988, (1998) 2002; Klein 2005). At its apex was the national government, whose Ministry of Health had prime responsibility for the service. Executive power notionally lay with the minister of health appointed by the prime minister from among the senior politicians of the governing party, supported by junior ministers and advisers. In 1968 the ministry was restructured and renamed the DHSS, while the minister became secretary of state for social services, with a remit including both health and social security. Day-to-day control over the NHS was exercised by the ministry's civil servants. Such government bureaucrats held permanent appointments, were nominally nonpartisan, and, in addition to implementing policy, participated in its development (Hennessy 1989). Democratic accountability for the NHS resided principally with the national parliament, for only limited public health duties now remained with elected local authorities. New regional hospital boards (RHBs) were created, run by appointees of the minister, while separate executive councils, on which local doctors sat alongside appointees, oversaw primary care. This arrangement persisted until 1974, when this tripartite structure was replaced by tiered Regional, Area, and District Health Authorities, which mapped onto local government boundaries, the better to integrate preventive, primary and hospital care.
Very crudely, the political economy of the NHS in its first forty years can be summarized as follows. Though technically a “command and control system” there were initially “rather few commands and precious little control” (Le Grand 2003: 49). Government proffered advice and allocated financing, but managerial responsibility was delegated to the local level, remaining largely in the hands of pre-NHS medical and political elites. Nationally, a broad consensus over the service obtained between the governing Conservative and Labour parties, though Conservatives were more inclined to constrain expenditure, particularly in the 1950s and 1980s (Webster 2002; Appleby 1999). Growth in the 1960s fueled quality improvements in general practice and a hospital renewal program, before economic difficulties slowed welfare state expansion in the 1970s. Henceforth, political conflict over the NHS intensified as governments sought to contain costs and extend managerial authority, culminating (as readers of this journal will know) in more radical structural reform in 1989, when the Thatcher government launched its “internal market” (Klein 2013).
The RAWP: Key Features and Research Questions
The RAWP episode therefore manifested the mid-1970s moment when central government began to pursue a more interventionist policy. However, it also sprang from a contradiction present from the start of the NHS. The service's Labour architect, Aneurin Bevan, had promised it would “universalize the best” for all citizens, addressing spatial inequities rooted in the “caprice of charity” and the patchiness of local government financing (Bevan 1946: 46, 49). For example, the interwar distribution of voluntary hospital capacity was so uneven that in-patient admission rates varied fivefold across the major cities (Mohan 2006; Gorsky, Mohan, and Powell 1999). London, where about 25 percent of English and Welsh voluntary beds were located, was particularly privileged as the location of venerable teaching hospitals and many specialist institutions attractive to philanthropy (Pinker 1966: 57). Municipal hospitals partially ameliorated the geographic unevenness inherent to the distribution of charitable foundations. However, public expenditure varied significantly according to the local wealth base, and these institutions had worse staffing ratios, fewer technical facilities, and less outpatient capacity (Levene, Powell, and Stewart 2004; Powell 1992; Hollingsworth and Hollingsworth 1985). Nor had statutory health insurance overturned the market incentives that determined the geography of primary care. A sixfold difference in doctor-population ratios existed between major cities, with mining and industrial locations the least favored, while in small towns and rural areas with at least 5,000 residents, the range was between 0.07 and 3.26 doctors per 1,000 population (Powell 2005).
Despite these problems, and Bevan's rhetoric, the founding legislation contained no program for geographic redistribution, assuming instead that the new regional authorities would resolve these issues. However, there was no local enthusiasm for reforms that might disturb existing medical power structures, and resources continued to be apportioned largely on the basis of pre-1948 expenditure patterns. This meant that twenty years into the life of the NHS the existing distribution of facilities was little changed, thanks to “the inertia built into the system by history” (DHSS 1976a: 7). The first challenge for RAWP historians will therefore be to explain how an apparently marginal concern rose to prominence in the mid-1970s.
Before this, however, some preliminary details of the committee and the solution it proposed are needed. The RAWP was set up in May 1975 by the Labour government led by Harold Wilson. The politician formally responsible was Barbara Castle, secretary of state at the DHSS, on the left of the party, and remembered for championing not only egalitarianism in the NHS but also disability rights and equal pay for women (Perkins 2003). However, it was her minister of health, the more centrist David Owen, who led the initiative. The committee's brief was to review and improve the process through which central funding was allocated geographically. The RAWP's response was a novel formula that aligned funding with population health needs. It began with the principle that the needs to which a health service should respond were not the same as public demand, which tended to be “always one jump ahead” of what a nation's limited resources could deliver. Instead it proposed that the needs of a given place could be systematically calibrated with reference to its demographic features, adjusted to account for its specific “morbidity characteristics.” Funding could then be dispensed in response to need, rather than existing supply or incalculable consumer demand (DHSS 1976a: 7–9).
What was the solution that the RAWP devised? Figure 1 illustrates the working of the formula, whose main principle was to allocate resources on the basis of geographic population levels (row 1), weighted to reflect various considerations. The first modifying effect (row 2) was anticipated variations in usage of hospital and community services. This was established by separately weighting seven main fields of activity, principally by regional sex and age structure (using national utilization data); in the case of psychiatric hospitals the known epidemiological link between marital status and utilization was also incorporated. Next a further adjustment was added to account for variations in morbidity (row 3), for which the chosen tool was standardized mortality ratios (SMRs). The RAWP argued that these provided the best available proxy for morbidity and thus need for nonpsychiatric care. Moreover, because they provided a direct measure of health care need, SMRs obviated the requirement to include in the formula factors such as “occupation, poverty, social class and pollution,” with which they already overlapped (DHSS 1976a: 14–15). Standardized fertility rates were also incorporated to calibrate demand for maternity services. Final adjustments (row 4) were made to account for cross-boundary patient flows related to hospital location or tourism, existing numbers of long stay patients, extra costs of teaching hospitals (the Service Increment for Teaching), and a London weighting.
The pattern of implementation from the late 1970s, when the formula was well established, to the late 1980s, when it underwent adjustment in response to the internal market, is shown in figure 2. Immediate transition to the new dispensation was rejected: losers would struggle to maintain services and would probably close hospitals, while winners might make inefficient use of major increases due to inexperience. Instead, the approach was one of gradual advance to the RAWP target. The graph illustrates the privileged position of the four metropolitan regions—London and the home counties—and the relative disadvantage of the North, with an initial range of budgetary excess or shortfall around the RAWP target of 22 percent. Nonetheless, by 1988–89 this had narrowed to about 11 percent, or, if the two outlying North Thames regions are excluded, to only 6 percent.
Here, then, is evidence for the success of the RAWP, though obviously success must be carefully qualified. After ten years, the targets for fair financial allocation had still not been fully achieved. Nor do the gains in spatial distribution represented in figure 2 reveal anything about other dimensions of health system equity, such as the division of resources between acute care and “Cinderella services” like mental health and geriatrics, or the enduring gulf between rich and poor in health outcomes. Nonetheless, judged on its own terms, as a policy concerned with geographic allocation of resources, the RAWP had clearly succeeded in bringing the regions substantially closer to equalization. The case for success also rests on the claim that the RAWP instilled two enduring principles into health policy. One was that equality of access for citizens with equal need was a desirable, popular, and attainable goal. The other was that this should be achieved by empirical formulation rather than by the informed judgment of civil servants. The second research question the RAWP provokes, then, is to account for this accomplishment.
RAWP by the Historians
The RAWP episode is briefly treated in the major NHS histories, as well as in three more detailed studies, with several actors in the events among its scholars. The official historian of the NHS, Charles Webster, attributes action principally to ministerial leadership. He emphasizes Barbara Castle's role, regarding RAWP as part of her larger program of priority setting (Webster 1996: 606–13). An avowed admirer of an egalitarian NHS, Webster is critical of the RAWP's “limited progress,” considering this to have been slowed by Treasury hesitancy and Thatcherite disinterest (Webster (1998) 2002: 84–87). Geoffrey Rivett, author of another major survey text, was a DHSS civil servant in the period. He similarly attributes the RAWP initiative to Labour politicians, with Richard Crossman (secretary of state, 1968–70) as the progenitor, and Owen and Castle responding to a “deep-seated political imperative to redress the inequalities in provision”; he sees this as a consequence of the 1974 reorganization, which exacerbated inequities when the costly teaching hospitals were integrated into regional authorities (Rivett n.d.). He is more accepting of its effectiveness, noting particularly its effect on London.
Rudolf Klein treats the RAWP only briefly, though he himself was involved, initially as a researcher with the ear of David Owen and later as a health policy expert. He depicts the RAWP as a creature of technocratic planning within broad political consensus and thus emblematic of the era that Thatcherism later swept aside (Klein 2005). Its prevailing ideology of efficiency allowed “paternalist rationalisers” to dominate the field, with the RAWP an exercise in “rationing” (the pejorative term preferable to the euphemistic resource allocation). Eschewing evaluation, Klein notes comparatively slow progress toward equalization while acknowledging its work in addressing London's overprovision.
Turning to more specialist accounts, Walter Holland includes the RAWP, of which he was a member, in his history of health services research (HSR). Holland was the University of London's first professor of social medicine, based at St. Thomas's Hospital, and had come to HSR from epidemiology (Holland 2013). He explains RAWP as politically inspired by pressure from northern members of parliament who observed that their constituencies were underserved by new medical facilities. His own contribution figures prominently, highlighting the influential St. Thomas's research agenda on resource distribution and epidemiological modeling of needs. Holland trenchantly defends the RAWP, arguing that it successfully reduced funding gaps and functioned with simplicity and transparency, in contrast to later “fiddles” (Holland 2013: 161–66).
Nicholas Mays and Gwyn Bevan, who worked on the RAWP as researchers in Holland's department in the 1980s, take a similar approach. Their historical survey of earlier policy makers' attempts to address the issue highlights both the degree of continuity informing the initiative and the flurry of literature in the early 1970s that triggered action (Mays and Bevan 1987). Finally, John Welshman's study of the RAWP from the perspective of the Sheffield region broadens the reading of the intellectual precursors. Contra Mays and Bevan's case for intellectual continuity since the 1950s, he asserts a step change in thinking began in the late 1960s. Welshman also flags the policy role of health economists as an issue that “deserves more study than it has received hitherto” (Welshman 2006: 232).
Sources, Methods, and Concepts
The argument here carries forward these latter approaches, asserting the role of ideas and actors as a critical variable. The analytical focus lies with policy elites in government and academia, not with pressure groups or patient activists, who seem not to have been important in this episode. It shows particularly that the intellectual impact of HSR and health economics mattered and that their proponents' influence, applied from key positions in the policy-making architecture, was instrumental to the genesis of the RAWP, to its recommendations, and to the embedding of its findings. It builds on recent work on disciplinary developments in postwar public health research that has deepened understanding of the research-policy relationship (Sheard 2013; Holland 2013; Shergold and Grant 2008). It also draws on the archival record of the RAWP period, which is now mostly in the public domain (with the exception of material relating to the late-1980s RAWP review, for which our freedom-of-information requests remained unsuccessful). It relies too on recent oral and written memoirs of participants and key civil servants, who, with the benefit of distance, offer candid and illuminating reflections.
To conceptualize these issues, a theoretical resource from political science, the advocacy coalition framework (ACF), is employed. Associated principally with Paul Sabatier, this has gained traction as a useful heuristic for understanding policy processes in liberal democracies (Sabatier 1988). It emerged to address inadequacies of existing models. For example, pluralist approaches that treated policy as balancing the demands of competing interest groups neglected the power of changing ideas to shape outcomes. Similarly, institutional approaches that set out programmatic stages—issue recognition, agenda setting, solution finding, political action—were too focused on temporal processes than on causal mechanisms operating over the long run (Heintz and Jenkins-Smith 1988; Jenkins-Smith and Sabatier 1994).
The ACF instead depicts change as the outcome of struggle between groups within a given subsystem, or field, of policy. An advocacy coalition includes all actors whose beliefs and ideas shape a shared goal; thus, it can include bureaucrats, legislators, and formal interest groups, but also academics, journalists, and others. The assumption is that research evidence matters because it furnishes resources to advocacy coalitions as they seek to influence policy brokers—ministers and senior civil servants. However, such policy-oriented learning does not translate swiftly or rationally into action: politics is too much determined by core beliefs about the world, anchored in emotion or instinct, for this to occur (Schlesinger 1968: 285; Sabatier 1988: 143–47). Indeed, the ACF holds that all public policy change is essentially a “translation of belief” into action (Weible, Sabatier, and McQueen 2009: 122–23).
Thus, members of advocacy coalitions will themselves be motivated partly by their deep core beliefs (e.g., about the desirability of an egalitarian or a libertarian approach to health systems), partly by policy core beliefs (e.g., about equal access for equal health needs), and only partly by secondary narrower beliefs on technical aspects of policy, which are open to modification (Jenkins-Smith and Sabatier 1994: 180–82). This does not mean, as some claim, that research only serves to legitimize decisions taken for other reasons (Klein 1990: 503–6, 513; Schlesinger 1968: 283–84). Instead, it is assumed to exercise a longer-term “enlightenment” function, reshaping debate more gradually and strengthening cumulatively the advantage of one or other coalition (Weiss 1977). Hence, a time frame of at least a decade is required to observe the effects of research on policy.
While proponents of the thesis emphasize the play of ideas, they also explain policy change by reference to the parameters in which these are debated (Sabatier 1988: 134–39, 155–57; Jenkins-Smith and Sabatier 1994: 183–84; Weible, Sabatier, and McQueen 2009: 130–31). Some of these can be relatively fixed. For example, is the constitutional structure in which debates are held conducive or inimical to reaching solutions? Is the nature of the problem essentially practical, or does it encompass value-laden and potentially divisive social factors? The ACF hypothesizes that, where debates are technical and tractable, nonpartisan cross-coalition learning can occur. Other factors are more short-term, including changes of governing party or of the socioeconomic environment, which reframe the policy context and usher hitherto background issues to the fore. A final factor of change can be the composition of the advocacy coalition itself, as new members and intellectual resources are incorporated.
The argument advanced here is that an advocacy coalition around spatial redistribution of health resources formed in the early 1970s. Drawing on the emergent disciplines of HSR and health economics, it crystallized concerns hitherto expressed by disparate voices, which had kept alive Bevan's original ideal. An oppositional coalition existed favoring a market-driven alternative but remained marginal despite some support from health economists. Against both groupings was the tendency of the policy brokers in the DHSS to maintain the status quo, with any redistribution incremental at best. However, once the RAWP was implemented, a new coalition developed, uniting those who stood to lose. The next section examines the first manifestations of this debate under the newly established NHS.
Protocoalitions and Tendencies
Although it is unrealistic to talk of a “redistribution coalition” emerging in the 1950s and 1960s, it is possible to distinguish early protagonists. The first were located in the Department of Social Administration at the London School of Economics. Brian Abel-Smith, a newly qualified Cambridge economist, was key, as initially was Richard Titmuss, who led a group of social policy experts whose pronounced Fabian perspective made them favored advisers to Labour politicians (Sheard 2013; Halsey 2004). In 1953–56 they worked on the Guillebaud Report, an investigation into the cost of the NHS that demonstrated the good value it offered to taxpayers (Abel-Smith and Titmuss 1956). This also included discussion of underinvestment in new hospitals since 1938, the lack of incentive for regions to use allocations efficiently, and variations in local authority health funding, with attendant impacts (National Health Service 1956). Though not explicitly challenging spatial inequities, the report instead argued for better statistical data to support policy making (National Health Service 1956: 250, 267).
A second interested party was the UK Treasury, which expressed early concerns about the method of calculating regional allocations. Its Select Committee on Estimates found that these simply perpetuated existing expenditure patterns, with marginal adjustments for salary increases and inflation. Such an approach meant that “lack of proper economy can go unchecked and variations in cost between Region and Region may tend to become entrenched” (Select Committee on Estimates 1956: para. 19).
The third early advocate was the Acton Society Trust, a nonpartisan research charity. Between 1955 and 1959 it published six pamphlets on hospitals and the state, written by its director, Teddy Chester, later chair of social administration at the University of Manchester (Acton Society Trust 1958; Snow 2013). Chester described the Ministry of Health's ability to alter established patterns through bidding to the Treasury for extra discretionary funds for extensions or improvements (Acton Society Trust 1958: 28–29). Though noting some shifts in overall distributions away from the metropolitan RHBs, Chester argued that better empirical evidence was vital if “dangerous” allocative mistakes were to be avoided. For individual hospitals, this should set accurate costing against performance expectations based on national utilization indicators (Acton Society Trust 1959: 7–8). Chester's call for the ministry to accelerate its external research program (from which his own university department stood to benefit) was echoed by the trust's chair, Sir George Schuster, also an RHB chairman (Acton Society Trust 1959: 49–55).
Despite the political consensus and public approval that the NHS enjoyed, a strand of opinion existed favoring private medicine in its stead. Narrowly based, principally among reactionary medics and academic economists, this marketeers' coalition remained politically marginal (Seaton 2015). However, it articulated a vision of medicine in which resource allocation was best expressed through demand in the marketplace. In their efforts to demonstrate the failings of the NHS by the late 1960s, it was the marketeers who brought forth compelling data showing the persistence of uneven distribution.
A prominent early figure was the economist Dennis Lees, who argued in 1961 that health could reasonably be treated like any other good in the marketplace (Williams 1998). Lees had imbibed the ideas of Milton Friedman while studying in Chicago and subsequently championed neoliberal policies (Anon. 2008). His perspective aligned with that of the Institute of Economic Affairs (IEA), a free-market think tank established in 1955 to promote Friedrich Hayek's thought (Stedman Jones 2012). Not only did the IEA's leading light, Arthur Seldon, write on health care, but the IEA introduced American critiques to the NHS, such as James Buchanan's application of public choice theory: unbounded desire to consume health services conflicted fatally with resistance to commensurate levels of taxation, so better to let markets adjudicate supply and demand (Jackson, forthcoming). A celebrated joust between the IEA and Titmuss occurred over the latter's study of the economics of blood donation, The Gift Relationship (Titmuss 1970). Titmuss used this case both to demonstrate market failure in health and to argue for the motivating force of altruism. Various economists such as Tony Culyer held an opposing view, and a vigorous dispute ensued (Fontaine 2002).
The misallocation of resources by state inaction was therefore already a theme of the marketeers' coalition when it gained a foothold in the British Medical Association (BMA) in 1967. Doctors were angry that their remuneration was lagging and wanted NHS funding to rise to address this. A BMA faction led by Ivor Jones considered establishing a rival insurance system and flirted with economists critical of the service, like Seldon (Mencher 1968). Their report, Health Services Financing (Cooper and Culyer 1967), contained a lengthy appendix by two health economists, which suggested the “ideal of equality” was a chimera (Cooper and Culyer 1967). Detailing and correlating indicators of provision, costs, specialist care, and health outcomes, Cooper and Culyer (1967: 208–14) argued that the NHS had failed to deliver “social justice,” that northern “more working class areas” were disadvantaged, and “that discrimination works . . . in favour of the better-off citizens.” They adopted the marketeers' position, that the problem was a lack of managerial incentives to address demand, and floated three possible solutions: reversion to private medicine, better planning by the state (“not really a sensible objective”), or a mixed system that injected demand through vouchers or (their preference) insurance (Cooper and Culyer 1967: 207, 242–49). Thus, the marketeers' coalition had encouraged sophisticated analysis of spatial inequity, in which British health economists took a more libertarian stance.
Incrementalist Policy Brokers
Throughout the pre-RAWP period, then, the problem was identified but remained peripheral in policy circles. In ACF terms, the policy brokers supported a status quo in which only very gradual change was countenanced. This followed the political upheavals of the service's birth, after which a policy of continuity with prior patterns of funding was adopted. As postwar austerity was gradually relinquished, real increases maintained this status quo. Ministry bureaucrats had some latitude to direct extra funds to poorer regions, though through ad hoc assessment rather than statistical principle. This achieved some shifts between 1950 and 1959; for example, the overall share of the distribution to England and Wales of the four metropolitan RHBs had fallen from 41.7 percent to 38.3 percent, while various regions had gained, such as Newcastle, whose share rose from 5.3 percent to 6 percent (Acton Society Trust 1959). In the 1960s a new, potentially redistributive approach emerged with the Hospital Plan, a building program aiming to create a network of district general hospitals. Planning was premised on ideas about optimal bed-to-patient ratios for the major categories of care and regional estimates of new facilities required. Renewal of capital stock, and the consequent adjustments to current funding needs, would therefore bring in its wake a more rational distribution. Until then, incrementalism could continue.
Discussions about replacing this with a formula approach finally began in the late 1960s. It was becoming clear that the Hospital Plan would not be quickly fulfilled, due to lack of building capacity and a deteriorating national economy (Crossman 1976). Instead, policy under Labour's Richard Crossman turned to reconfiguring the NHS's tripartite administrative structure, which divided the RHBs from primary care and public health. Crossman was particularly frustrated by ministerial impotence over the “self-perpetuating oligarchies” that ran the RHBs, which he considered to be “80 per cent non-Labour” (Crossman 1976: 255–56, 804). Their constitution had essentially preserved the pre-NHS status quo in which ex-voluntary hospital and consultant elites dominated, tending to privilege the interests of acute care over mental health and geriatrics. It was reorientation toward these programs, rather than spatial readjustment per se, that Crossman sought, particularly after various scandals exposed the failings of “chronic” care (Crossman 1976: 419, 466). However, a ministry official, Dick Bourton, convinced him of the “great unfairness to Sheffield, Newcastle and Birmingham” that current financing methods maintained. His green paper on NHS restructuring (DHSS 1970) flagged their replacement by a needs-based population formula as the “long-run” aim (Crossman 1976: 569). This put in train the creation of the “Crossman formula” by his adviser Brian Abel-Smith, which was actually implemented under Crossman's Conservative successor, Sir Keith Joseph.
However, the new formula perpetuated existing inequities. The RHBs collectively put up a “tremendous struggle to maintain the status quo,” although the main difficulty lay with the formula itself (Crossman 1976: 876). This proposed that 50 percent of a region's allocation should be determined by population size, 25 percent by its number of existing beds, and 25 percent by its utilization levels. As would soon become clear, provision and utilization rates were faulty indicators of need, because hospital usage tended to follow supply. Nor would population alone help, without some adjustment for anticipated morbidity (Holland 2013). A further complication was introduced by the Revenue Consequences of Capital Schemes (RCCS) portion of funding, which augmented regional allocations to take account of the presumed extra current spending that new building under the Hospital Plan would incur. This tended to favor the southern regions, where more new infrastructure development had occurred. In sum, despite awareness of maldistribution, substantive change was impeded by the policy brokers' acceptance of the status quo maintained by regional NHS leaders, coupled with technical uncertainty about how to achieve readjustment.
Consolidation of an Advocacy Coalition
A Redistribution Coalition Emerges
In the years immediately preceding the establishment of the RAWP, it is possible to discern the emergence of an advocacy coalition to challenge this status quo. There are three senses in which this happened: academic research promulgated the intellectual justification for spatial redistribution, health services researchers and health economists became accepted as technical experts who could offer policy-relevant advice, and individuals conversant with these disciplines and sympathetic to core egalitarian values gained access to policy brokers (Klein 1976: 468–71). Thus, by 1975 a loose advocacy coalition was in place, by no means focused on spatial redistribution as a single urgent issue but with the belief and expertise to drive policy-oriented learning.
The intensification of public discussion began with Julian Tudor Hart's 1971,Lancet paper proposing an “inverse care law.” A socialist general practitioner, epidemiologically trained and based in industrial South Wales, Tudor Hart argued that medical resources tended to be lowest where population needs were greatest (Tudor Hart 1971). Although his call to action lacked empirical justification, this was soon supplied by others (WS 2014: 19). Key contributions were made by health economists from the University of York: Cooper and Culyer (1971), Alan Maynard (1972), and Peter West (1973). They demonstrated the failings of the Crossman formula. John Rickard, originally with the Oxford RHB, produced a study of unevenness across its areas and later extended the analysis nationwide (WS 2014: 75; Rickard 1974). Another affirmation of the inverse care law in the Lancet showed a negative correlation between financial allocations to community health services and the percentage of population in lower socioeconomic groups (Noyce, Snaith, and Trickey 1974). In the year of the RAWP's appointment, the BMJ carried papers by Gentle and Forsythe (1975) and by Buxton and Klein (1975), the latter reporting regional variations from national means of hospital services spending of +41 percent to −23 percent, with intraregional differences even greater.
This growing volume of technical analysis is best understood in light of larger developments in academic public health. The specialty of HSR had emerged in the 1960s initially because epidemiologists became interested in the relationship between service inputs and health outputs (Morris 1957: chap. 3; Berkowitz 1998). Medical sociologists and operational researchers providing academic training for NHS managers also contributed, and the subdisciplinary trappings of a journal (Medical Care) and scholarly meetings soon arrived. The availability of funding from the Medical Research Council and the Ministry of Health meant that, in addition to Holland's group, various other centers became prominent (Bierman et al. 1968). Much effort went into understanding utilization patterns, the better to plan service needs. This included a major survey of Liverpool and Manchester by Robert Logan's cluster at the London School of Hygiene and Tropical Medicine, which included Rudolf Klein and John Ashley (both RAWP actors), and Holland's studies of the St. Thomas's Hospital catchment in London (Logan et al. 1972). These demonstrated that usage rates responded to existing provision rather than to underlying population factors, a finding already established by American investigators Milton Roemer and Kerr White (Shain and Roemer 1959; White, Greenberg, and Williams 1961). Another pivotal moment was the publication in 1972 of Effectiveness and Efficiency by the epidemiologist Archie Cochrane (best known today as progenitor of the Cochrane Collaboration centers for collating systematic reviews). He urged that randomized controlled trials be applied to clinical therapies and procedures to ensure effectiveness (i.e., that they worked in a laboratory setting) and efficiency (i.e., that they were cost-effective in the real world), with the ethical imperative that all effective treatments should be free (Cochrane 1972; Berkowitz 1998).
The consolidation of health economics came in the wake of these earlier trends. A social policy specialty within economics had a long lineage, concerned principally with explaining trends in public spending. Jack Wiseman had made this a departmental focus at the University of York, and one of his protégés, Alan Williams, had narrowed his interests to health and established a research cluster, in which Culyer and Maynard became major figures (Croxson 1998). An early symposium convened by Wiseman announced the specialty's identity and preoccupations (including three papers on resource allocation), but it was the inaugural meeting of the Health Economists' Study Group (HESG), again led by the York center, which definitively signaled arrival (Williams 1998; Hauser 1972). The HESG went on to become a forum for engaging academics and policy makers, and Williams's stewardship ensured British health economics adopted an advisory posture compatible with the NHS. Contra Dennis Lees, Williams had argued that the economics of the firm did not best suit analysis of health, a position also developed in the United States by Kenneth Arrow, who identified market failures of commoditized health care, arising from information asymmetries between patient and doctor and consequent trust problems (Williams 1998). Later Williams declared himself supportive of Cochrane's egalitarian philosophy, believing that economics brought to it the dispassionate tools of assessment (Williams 1997).
Actors and Relationships
These disciplinary developments bore upon the redistributors' coalition in a practical sense: people espousing new ideas now confronted the policy brokers more closely. In an early placement with the UK Treasury, Williams riled health officials by criticizing the lack of statistical indicators on which to base policy decisions (Williams 1997). Cochrane was among his audience and liked what he heard, subsequently recruiting an economist, David Pole, to his department of epidemiology at the University of Cardiff (WS 2014: 79). Pole was a Cambridge-trained contemporary of Abel-Smith, and after Cardiff moved to join the Economic Adviser's Office (EAO) at the DHSS. He was also a HESG member. York's direct influence came not only from Culyer, Cooper, and Maynard's interventions but also from the careers it fostered. Peter West was a PhD student of Culyer's, who followed a trajectory parallel to Pole's, as an economist joining Holland's Community Medicine unit to work on resource allocation (WS 2014: 21–22, 31). Terri Banks, a DHSS official who later played a major part in implementing RAWP, had learned economics methods from Williams while he was seconded to the Treasury (Terri Banks, personal communication, October 7, 2015). Jeremy Hurst and John Rickard were both early HESG members who worked with Pole at the EAO and were involved with RAWP (WS 2014: 79–80; Croxson 1998; Hurst 1998).
Finally, Brian Abel-Smith, with whom the concern for indicators of health resource allocation had begun in the 1950s, had achieved an influential position. Now an international leader in health systems statistics, he had acted as a special adviser to Crossman and understood the workings of the DHSS (Sheard 2013). With Labour's victory in 1974, he returned to advise Castle and Owen, who valued his expertise and diplomatic skills. They perceived him as “utterly Labour, to his core,” and as someone whose pragmatism counterbalanced idealism (WS 2014: 15). He also remained significant in public policy research at the London School of Economics, where his appointee Bleddyn Davies had analyzed spatial inequities in local government, coining the phrase territorial justice (Davies 1969). He helped shape HSR, too, chairing the advisory committee of Holland's unit at St. Thomas's Hospital (on whose governing board he had sat).
External Factors of Change
Looking beyond individual agency, what aspects of the external environment, in ACF terms, helped facilitate change? First, the RAWP debate began under a broad consensus over the core values of the NHS. It could be positioned as essentially a technical question of means, which assumed the ends of spatial equity were uncontested. The Conservatives had accepted these in principle when they implemented the Crossman formula, so in parliamentary terms this was a neutral issue. A new general practitioner contract had quelled the libertarian rebels in the BMA, removing momentum from the marketeers' coalition. A window opened in which resource allocation policy could be discussed without immediately provoking controversy.
Various other factors made it politically attractive. By 1975 the health care economy was entering a transition. The fiscal crisis of European welfare states was just beginning, as OPEC-induced oil price shocks coincided with the end of the trentes glorieuses (Lowe 2005: 315–27). Although NHS spending remained relatively high under Castle compared to the tighter settlements demanded later, it was clear that the years of expansion were over (Appleby 1999). Ministers now accepted that the challenges of inequality would not be resolved by steadily rising NHS budgets.
David Owen also favored an active policy toward the NHS. Coming from a medical background, he saw it as embodying British values of altruism and citizenship rights but felt strongly that the inequalities agenda demanded attention (Owen 1976: 1, 3, 172). Now that Keith Joseph's 1974 reform had resolved debates over the NHS's administrative structure, that agenda could be revived. Moreover, with financial strictures looming, the case for adjustment could be made on grounds of allocative efficiency, thus spiking the guns of those set to lose from social redistribution (Owen 1976: 49–54). Finally, although the ACF approach minimizes individual actors, Owen's intellectual capacity to master a complex brief, his willingness to confront vested interests, and his impatience with temporizing mandarins should be noted (Webster 1996: 747–49; WS 2014: 74–77). For all these reasons the redistributors' coalition now had its opportunity.
The Redistributors' Coalition in Action
Inception of the RAWP
Two contradictory accounts of the RAWP's establishment are provided by key actors. Walter Holland recalls that Abel-Smith suggested his St. Thomas's unit should conduct research into resource allocation shortly after Labour's return to office. He devised a complex randomized trial of health authorities, selected to represent places with high or low cardiovascular, cancer, and perinatal mortality. Some would receive earmarked financing to address these, while others would receive a general funding uplift, and the health outcomes would then be compared (Holland 2013: 161–62). Owen promptly vetoed this proposal on the grounds that “no way would he be able to sell a randomised control trial to Parliament” (WS 2014: 25). Shortly after though, Holland was invited to join the RAWP, which Owen announced in May 1975, and he believes his draft proposal, coupled with Abel-Smith's urging, sparked Owen's initiative. By this stage Abel-Smith would have been well aware of research showing that the combined effect of the Crossman formula and the RCCS were worsening the problem. Moreover, Castle trusted Abel-Smith and willingly delegated to Owen provided he was involved (WS 2014: 13–14).
David Pole's alternative account begins with a summons to advise Owen on principles of capital allocation. Owen had been asked to approve a new hospital in Conservative-supporting Lincolnshire, and was considering the justification, when other towns were equally deserving, such as Labour-supporting Lancashire (WS 2014: 74–76). Pole's investigation began with the senior DHSS official responsible for capital schemes, who explained that the “imponderable elements” were such “as to make rational planning impossible,” before joking that “one found out where the local MP and the chairman of the hospital board lived, and took it from there” (WS 2014: 74). Such was the confidence of incrementalist mandarins in their existing approach that they tried to dissuade Owen even from reading Pole's subsequent report. “Owen did, of course, read it” and “immediately set up the . . . RAWP” (WS 2014: 75). Pole also credits Abel-Smith's intervention, believing that their personal Cambridge connection explains why the hitherto marginal EAO gained Owen's attention (WS 2014: 78).
Whatever the precise causal factors, the RAWP's establishment placed key advocacy coalition figures in positions of influence. The main committee included Holland, Forsythe, and Pole, who also figured alongside others in the three subgroups where the analytical work was done: RAWP(R), tackling the main revenue expenditure formula (Holland, Forsythe, and Rickard); RAWP(C), addressing capital allocations (Forsythe and Rickard); and RAWP(T&R), responsible for assessing what teaching and research increment would be needed (Holland, Snaith, and Hurst) (National Archives 1975a).
Problem Parameters: Technical Issues or Core Values?
ACF theorists draw attention to the manner in which an advocacy coalition formulates and tackles a policy problem (Jenkins-Smith and Sabatier 1994: 191–93). Can it be framed as essentially a technical issue, where disagreements hang only on “secondary” scientific criteria? Or, as is often the case with matters of social policy, does it touch on core political values and thus court controversy in the public realm? If the latter, then opponents may question a policy's legitimacy rather than restricting debate to its detail, thus increasing the likelihood of failure. Although the RAWP enjoyed cross-party consensus at its launch, this was by no means guaranteed to last. London regions that stood to lose accommodated powerful interests in medicine and academia who potentially might mobilize dissent. How, then, did the RAWP coalition succeed in coalescing support for core policy beliefs?
One answer is that, although its initial terms of reference were technocratic, the committee skillfully reworked these to consolidate nonpartisan ethical credentials (WS 2014: 41). Its brief from Owen had been to devise “a pattern of distribution responsive objectively, equitably, and efficiently to relative need” (DHSS 1976a: 5). The report, however, reinterpreted “the underlying objective” as “to secure, through resource allocation, that there would eventually be equal opportunity of access to health care for people at equal risk” (DHSS 1976a: 7).
The power of this formulation lay in its simple affirmation of equal access for equal need. Further moral high ground was staked by the report's title: Sharing Resources for Health in England. This discursive positioning of centralized rationing as fairness and mutuality inhibited potential opposition, for who in the British polity could reasonably dispute these principles?
Another key factor was the attention to consensus building in the committee's makeup and working. It combined representation of DHSS and NHS staff, including authorities in both the South and the more deprived regions (Mays and Bevan 1987). It also ensured a gradualist transition by issuing an interim report in August 1975, which set the next year's formula pending final proposals. This signaled the direction of travel, though not the extent of what was planned. It combined in a ratio of 3:1 the first steps in the new population weighting calculation (fig. 1, rows 1 and 2) with the Crossman formula's inclusion of existing utilization; it also maintained the RCCS portion, favoring the South, and applied a new increment to support teaching hospitals. Finally, it gave the DHSS latitude to avert objections by introducing the premise of a floor and ceiling, so that no regional health authority's allocation decreased or increased beyond 2.5 percent (National Archives 1975b: 7, 8, 12; DHSS 1976a: 94). Alongside its statement of key principles, this signaled only a modest departure from incrementalism.
Nonetheless, the aftermath of the interim report was a dangerous juncture at which the issue might have flared into controversy when, for the first time, the debate moved beyond Whitehall and academia. A BMJ (1975) editorial, “Painful Redistribution,” caught the attention of the tabloid press, which spun the story as “Axe to Fall on Hospitals—Ministry's Secret Plan” (National Archives 1975c). The capital's Evening Standard similarly announced that “London Bears Brunt of New NHS Cuts.” Staff interest groups such as the National and Local Government Officers' Trade Union, the Institute of Health Service Administrators, and the Health Visitors' Association were also exercised about threats to jobs and services (National Archives 1975d).
However, the policy core remained intact. Of the regions, only the four London authorities were actively opposed, with objections leveled not at principles but at immediate budgetary implications (National Archives 1975e). For example, it was argued that savings would necessitate service cuts, meaning the poorest districts within the overfunded regions would suffer, too. In addition to southern lobbying against too-rapid adjustment, several regions argued that the formula took insufficient account of social deprivation and that the teaching allowance and London weighting proposals were as yet unconvincing (National Archives 1975f). In sum, the inflammatory aspect of the debate was articulated as a “cuts scare,” not as a controversy about promoting equity over localism. Castle was able to neutralize the former objections when, following annual departmental negotiations with the Treasury, she secured a budgetary settlement large enough to ensure a floor that protected loser regions (National Archives 1976a). Remaining objections were not counterproposals but practical concerns about pace of implementation and about making the formula more redistributive (National Archives 1976b).
The final element positioning the RAWP debate as a technical problem, bounded by agreement over core values, was the buy-in of midlevel bureaucrats. ACF theory does not attend greatly to civil servants as members of an advocacy coalition, though it has noted, for example, that they tend to be more moderate elements within the coalition and may retain powers of clientelism (Weible, Sabatier, and McQueen 2009: 129; Cairney 2012: 213–14). The RAWP example affirms this but reveals something more. Owen had sought a formula that was “readily available at all relevant levels of aggregation,” “would reliably predict . . . variation in health need between localities,” was “unambiguous,” and would “reflect ‘need’ alone and not be influenced by supply” (National Archives 1975i). Evidence suggests that as the process unfolded the DHSS members came to believe that this could be achieved, and that by its end they had a methodology that was transparent, workable, and defensible.
Crucial to garnering this internal support was the formula's most innovative feature, the application of regional SMRs to adjust population allocations for need (as proxies for morbidity) and deprivation (because they correlated closely to poverty indicators). Its adoption is illustrative of the advocacy coalition in action, although, again, there are conflicting accounts. Pole claims that it occurred to him while “pondering the problem in the early hours,” while Holland attributes the idea to his St. Thomas's unit and its comparative analysis of morbidity indicators (WS 2014: 25, 76; Holland 2013: 163). In any event, the documentary record of successive RAWP committee meetings points to a joint endeavor among experts in HSR and health economics (National Archives 1975g, 1975h, 1976c). In January 1976 the strategy was approved and the RAWP(R) subcommittee was tasked with finalizing the formula (National Archives 1976d). Confidence grew following a modeling exercise, which showed that the overbedded but comparatively deprived Mersey region would suffer less than the interim report had implied (National Archives 1976e). By early 1976 the RAWP felt it had an accessible and acceptable formula with which to proceed.
Civil service buy-in to the policy core beliefs was therefore explicable in terms of the science, but there was another individual factor that the ACF does not capture: the role of John Smith, the RAWP's chairman. Now a DHSS under secretary, Smith was an economist who had come to health administration from a background in social security when the two sides of the department merged. He thus epitomized a changing departmental culture, as health policy opened up to “economists, the statisticians, the operational research people” (WS 2014: 32, 34). Smith was also sufficiently senior to be unfazed at upsetting colleagues whom the RAWP disempowered (WS 2014: 79). Less tangibly, his style had the ability to inspire staff, and oral reminiscences of his leadership are fond and admiring. Lis Woods, of the RAWP secretariat, recalls: “He was very clear that we must not aim for perfection; perfection was impossible. What we must and could aim for, and was possible, was less imperfection. I think that principle again helped us to do something practicable that worked and lasted” (WS 2014: 35; National Archives 1975b: 2) Thus, as implementation neared, an esprit de corps was fostered in support of this “least imperfect” solution, within a broad consensus over equity goals.
Implementation and the “Losers” Coalition, 1976–89
Countercoalitions and Changing Policy Brokers
Although the RAWP was launched in propitious circumstances with strong political backing, its embedding was far from certain. A countercoalition emerged, which articulated stronger objections. The outlook of the policy brokers altered, too, particularly when Thatcherite conservatism challenged the ideological and partisan dynamics. Internal review processes also presented opponents with opportunities. Despite this, the advocacy coalition supporting the RAWP formula held firm, sustained by civil service support. The problem parameters remained largely technical, and belief in the policy was sustained even as challenges to the NHS's core values emerged. The result was that cross-coalition learning could take place, with consolidation and refinement of the formula. This final section explains how.
An angry response of loser regions and hospitals followed the interim report (National Archives 1975b), but attacks on core principles quickly gave way to debate about the risks of rapid implementation. Early critics included London consultants Sir Francis Avery-Jones and, from a “marketeer” position, Reginald Murley of the Fellowship for Freedom in Medicine. They argued emotively that the RAWP formula neglected “conurbation factors” and social deprivation, which local clinicians could perceive better than administrators (Avery Jones 1978; Murley 1976). The Royal College of Surgeons and the editors of the BMJ challenged the RAWP methodology in defense of the Southeast (Anon. 1976; Heslop 1977). This attack was short-lived, though, for provincial BMA members were incensed at their national leadership lending support to “London's howl of dismay,” which to underresourced regions seemed like special pleading (Hole 1976; Lockley 1977). However, more compelling arguments emerged from London's primary and community care sectors, experiencing rising demand as hospital services contracted (Jarman 1978). Brian Jarman, a general practitioner and academic from St. Mary's Hospital, developed a new deprivation index to capture excess medical need attributable to poverty in inner-city practices, which implicitly challenged the RAWP formula (Jarman 1983). For the first time, too, the patients' voice was heard, when community health councils, the NHS's newly created public representation bodies, joined the fray in RAWP loser areas of the South (see fig. 2) painting adjustments as cuts (Langton-Lockton 1978).
Despite these budding objections, in the later 1970s the policy brokers and the external environment remained favorable, even after the new Labour leader James Callaghan elevated Owen to foreign secretary and dismissed Castle. Her successor, David Ennals, nonetheless maintained the inequalities agenda, including both the RAWP and Castle's “programme budgeting” initiative (Webster 1996: 606–9). This was a related planning exercise that sought to redistribute resources across client groups, essentially to shift expenditure away from acute hospitals and toward older people, the physically impaired, and psychiatric patients (DHSS 1976b). Ennals also commissioned an inquiry chaired by Sir Douglas Black into the third dimension of inequality, the relationship between health outcomes and class, income, and occupation (Webster 1996: 612–13).
Despite this continuity, Ennals's tenure contained two flashpoints that might have presented an opportunity to the RAWP's opponents. One was the Royal Commission on the NHS, which the Wilson government had conceded in 1975 to assuage professional anger during a bitter dispute over private practice. In 1979 this produced the “first comprehensive, independent” report on the NHS, including the RAWP (Webster 1996: 725). Some of its evidence critiqued RAWP's centralizing tendencies and crushing of local diversity (National Archives 1976–79). However, it ultimately reaffirmed the policy's core values. Its review of the formula noted some of the underlying “heroic assumptions” and stressed that perfect spatial equity was a chimera, but it accepted the principle of equity and endorsed the mechanism as “rational and equitable” (RCNHS 1978: 3, 25, 27; RCNHS 1979: 344–45). Otherwise, its concerns were methodological, for example, over the proper adjustments to be made for teaching hospitals (RCNHS 1979: 282, 345–46, 374).
The other area of potentially flammable debate was internal. Ennals had established a DHSS Advisory Group on Resource Allocation (AGRA) “to consider minor changes” to the formula, relating to such issues as patient flows, age/sex patterns of utilization, and age-specific mortality weightings (National Archives 1978e). The unspoken motive, however, was concern about RAWP's impact on London. The capital's areas and districts had begun developing different approaches to calculating subregional allocations, and AGRA was urged to intervene quickly, lest this become “extremely damaging” (National Archives 1978b). Accentuating the difficulty was the work of the London Health Planning Consortium, whose remit was health services reconfiguration, and the review of medical education in the capital that was chaired by Lord Brian Flowers (University of London 1980). These both were concerned about oversupply in the acute hospital sector and seemed certain to exacerbate the RAWP squeeze when they eventually reported.
Civil servants had therefore to tread a delicate line, preventing AGRA from becoming a platform for special pleading while also retaining enough latitude to manage the London situation. John Smith managed this by keeping at arm's length the teaching hospital representatives and those with a “radical . . . but dangerous voice” (National Archives 1978a). He also obtained an additional weighting for London, reflecting its supraregional and specialty services, ignoring concerns that this was “protection for the status quo” and a “backdoor method of funding the London teaching hospitals in the style to which they are accustomed” (National Archives 1978b, 1978c, 1978d). This careful balancing act prevented AGRA becoming a forum for dispute, and its final report in early 1980 endorsed the RAWP's core principle of equal access for equal need while urging ongoing research to allow fine-tuning of the formula (DHSS 1980; WS 2014: 56–57).
Thatcherism and the Resilience of Policy Core Beliefs
The external context changed more emphatically after 1979, with the Thatcher government's victory heralding new policy brokers. This threatened the RAWP process for several reasons. First, the Conservative government's willingness to “think the unthinkable” on welfare initially seemed likely to revive the marketeers' coalition (Banks 2015). Free-market think tanks articulated neoliberal critiques of social policy, and the government's Central Policy Review staff actively explored switching to an insurance-based model of health service funding (Lowe 2006). Second, Thatcherite creed held that “inequality is not only just, it is necessary to freedom itself,” both as reflection of innate difference and as reward for wealth generation (Thatcher 1991). The idea of directing public policy to ameliorating inequalities of health outcomes was incompatible with this worldview, as indicated by the rejection in 1980 of the Black Report by the new secretary of state, Patrick Jenkin (Berridge and Blume 2003). Equality of access might be vulnerable, too. Third, the party-political calculus had shifted. Now it was representatives of the loser regions that wielded parliamentary power, for Tory strength was historically rooted in southern England. Finally, Jenkin's early policy direction for the NHS emphasized a revival of localism as an antidote to the bureaucratized central state.
The durability of the RAWP therefore seemed far from assured as the 1980s advanced. Sir Graham Hart, then a leading figure in the NHS management board and later permanent secretary at the DHSS, recalled “voices were being heard from Number 10 and other political directions, quite insistently, through the mid-1980s . . . saying, ‘What is this, this instrument of torture, RAWP, which is inflicting pain on Conservative constituencies and giving money to Labour-voting constituencies in the north of England?’ It was not an obvious policy you could make stick and carry through with” (WS 2014: 52–53). Again, the critical factors in explaining the RAWP's durability are the continued framing of the issue as essentially technical and the belief in the policy held by bureaucrats charged with its implementation.
As the 1980s began, then, work on improving the formula had stalled, and pressures were bearing on the Thatcher government to row back from redistribution. Yet the policy core still held. One reason was that macroeconomic policy dictated ongoing austerity for social programs, with real growth in NHS budgets now much reduced. In this context, the RAWP remained attractive as a driver of allocative efficiency. Jenkin's successor, Norman Fowler, was also sensitive to equity issues, quickly scotching talk of a new funding model as politically unviable, a position eventually accepted by senior Conservatives. His argument was undergirded by a DHSS review setting out the problems and risks of insurance-based approaches, which, ironically, was prepared by Terri Banks, the civil servant later responsible, as director of health authority finance, for managing RAWP (Banks 2015). Thus, revival of the marketeers' coalition was muted, its influence confined to promoting private medical insurance and the contracting of ancillary services. It was also fortuitous that Fowler and his minister of state, Kenneth Clarke, both held midlands seats (Sutton Coldfield and Rushcliffe), so they were not subject themselves to immediate constituency pressures from “losers.”
By the mid-1980s another juncture was reached at which policy change might have occurred: the RAWP Review (NHS Management Board 1988). This arose from Fowler's focus on enhancing NHS productivity through stronger management and better performance indicators. Following a report by a leading industrialist, Sir Roy Griffiths, he set up a new NHS Management Board, conceived on the model of corporate general management, nominally to take responsibility for planning, implementation, and expenditure out of the political arena (Edwards and Fall 2005). In December 1985, shortly after its establishment, Fowler tasked the board to recommend improvements to the RAWP formula in light of experience, new research, and consultation. Though instructed to prepare recommendations within a year, the review team proceeded slowly, issuing an interim report in 1986 and requesting further time for research. A final report appeared in 1988, recommending several changes to the formula.
The RAWP Review illustrates again how those sympathetic to the policy core ensured that debate centered on means, not principles. As with AGRA, members were appointed not for interest representation but for technocratic ability, such as John Ashley, an epidemiologist specializing in morbidity measures (Ashley and McLachlan 1985). The board also stated explicitly that the principle of equal access for equal need was “not in question” (NHS Management Board 1988).
Even so, two issues threatened fundamental change. The first was the question of whether RAWP should be discontinued once it had removed historic inequities, which by now had been “substantially reduced” (NHS Management Board 1988: para. 1.2). This was firmly rejected: demographic change was ongoing, so a national formula should be maintained (NHS Management Board 1986: 8). The second issue was whether SMRs as an indicator of need should be abandoned, and now Jarman proposed an alternative formula that incorporated various social factors (e.g., measures of overcrowding and lone parenthood) alongside existing utilization (NHS Management Board 1986: 12–14, E5–6). Again, this was rejected due to the problems of basing the formula on utilization and the risk of double counting arising from the correlation between SMRs and deprivation indices.
Between the interim and final reports, debate hinged on specialized methodological matters. This was conducive to cross-coalition learning, but not a challenge to policy beliefs. Jarman now argued for better sensitizing the SMR measure to social deprivation, and the review commissioned the accountants Coopers and Lybrand to lead a small-area analysis of the problem. It concluded that need for health services was determined by social factors above and beyond those captured by the SMR and that Jarman's underprivileged area (UPA) index could model these. The weighting of SMR to need ought therefore to be reduced from 1:1 to 0.44, and the UPA index introduced to adjust for social influences (NHS Management Board 1988: paras. 2.1–2.52). The ensuing debate was mostly arcane, centering on the conceptual entangling of utilization and need and the appropriateness of Coopers and Lybrand's regression analysis (Morgan, Mays, and Holland 1987; Carr-Hill 1988; Mays 1989). Occasionally, rancorous core belief language crept into this technocratic arena. One York health economist condemned the UPA measure as “methodologically confused . . . out-of-date . . . and uninterpretable,” suspecting the whole endeavor was designed to favor London at the expense of areas in the North and Northwest, which by the SMR rankings alone were worst off (Carr-Hill 1988: 10–11). The St. Thomas's unit, where the SMR approach had originated, went further, identifying the UPA adjustment lobby as RAWP losers, purveying an “essentially political” strategy driven by “powerful interest groups” (Mays 1987: 46, 58). Despite these critiques, the review endorsed the changes, believing that in practice the effects would be “relatively small” (NHS Management Board 1988: table 1.1). Jarman's concerns seemed sincerely driven by the pressures falling on London services, and civil servants shared this perception (Gorsky and Preston 2013: 24–25, 56–60; WS 2014: 63). The debate had ultimately remained within existing parameters, and if the resulting compromise displeased some in the RAWP coalition, the intention at least was progressive redistribution to the poorest.
Throughout the 1980s, the behavior of those midlevel bureaucrats who had initially endorsed the RAWP remained crucial to its ongoing success. Now in senior positions, civil servants such as Terri Banks, Michael Fairey (director of planning with the management board), and Jeremy Hurst (senior economic adviser, Economic Advisers Office) retained, in ACF terms, both core and secondary policy beliefs. In contrast to the Black report on health inequalities, whose costly agenda for change seemed hopelessly unrealistic to civil servants working under Thatcherite ministers, RAWP's compound of equity and efficiency, its simplicity and transparency, and its underlying logic sustained internal support (Klein 1990: 518–19). When confronted with skepticism, civil servants felt able to defend the policy to Conservative ministers, whose fair dealing ultimately rewarded them (Banks 2015: 12–14). As Graham Hart noted in a recent witness seminar: “I was really surprised . . . and pleased, at the way in which officials and ministers—Norman Fowler and Ken Clarke—stuck to the policy, and they took a lot of stick for it . . . but I can honestly say that there was a real commitment. Terri is a very tough lady, and she reminded them from time-to-time what we were supposed to be doing, and they did accept it in the end . . . you had to talk through it, but it went on. The redistribution went on” (WS 2014: 52).
In this sense, then, the advocacy coalition that emerged in the early 1970s achieved its goal over the long term. Though subject to later changes, such as an adjustment for social inequality introduced by the Blair government, and still at the heart of fierce debate, for example, over the proper weight to be assigned to age as a need indicator, the RAWP approach successfully rode the waves of change to become established in British health policy.
In the taxonomy of comparative health systems, it is customary to classify the postwar British NHS as the emblematic “Beveridge” system, whose universalist aspirations were initially distinct from Bismarkian social insurance or more pluralist arrangements. Its history offers a case study of government and medical care in a tax-funded system, where the state is the main provider and health stewardship entwines with broader economic policy. As welfare costs have grown, government has increasingly sought to maximize efficiency and cost-effectiveness while acknowledging the electorate's visceral commitment to the NHS as a beacon of equity. To achieve this, it has pioneered several influential approaches, one of which was the RAWP, the subject of this article.
To explain the inception and persistence of the RAWP formula, we adopted the advocacy coalition framework (ACF) approach. This approach is attractive because it offers an explanatory model that goes beyond the actions of political elites and narrowly conceived interest groups. In the case of the RAWP, the model yielded helpful insights about the impact on policy of the slow diffusion of ideas, borne by academic experts and midlevel bureaucrats. However, before recommending its utility as a generic approach to the history of health politics, a few caveats should be entered. First, European researchers have raised concerns that the ACF model is overdetermined by the American political system, from which it was developed (Klein 1990; Cairney 2012). There the division of powers and multiple veto points in the legislative process necessitate broad coalitions to sustain change. By contrast, the British polity, with its tendency toward single-party majority government, makes ministers less beholden to lobby politics. In this context, as the RAWP case suggests, advocacy coalitions imply a looser affiliation of actors with a level of shared belief and expertise. Second, the emphasis placed by the ACF on the agency of these actors needs always to be balanced against the importance of structural economic forces within which they operate. Here, the imperatives of furthering cost-effectiveness in the 1970s and preserving allocative efficiency in the 1980s provided the context in which the “enlightenment function” of research ideas could flourish. Finally, while the ACF approach is helpful in reconciling conflicting evidence from personal testimony (as in the contradictory reminiscences of Walter Holland and David Pole), it underplays the importance of the individual and contingent. The intellect and temperament of key figures like David Owen and John Smith, and the fact that Thatcherite health ministers represented RAWP-gaining areas, fall outside the ACF model but matter to a convincing account.
Nonetheless, the ACF approach proved illuminating, particularly in affirming Welshman's speculations about the debut of health economics in policy circles. Indeed, the Health Economists' Study Group has itself historicized the RAWP episode as the first real impact it made on government (Hurst 1998: S48, S51–52, S56–57). The ACF approach also demonstrated how the “libertarian” beginnings of UK health economics gave way, if not to an egalitarian position, then at least to that of sympathizer (Williams 1997: 118; 1998: S3–4). Thus, in a 1981 essay the York triumvirate of Culyer, Maynard, and Williams ((1981) 1991: 339, 340) declared allegiance to the values underlying the NHS, which balanced freedom, social concern, and equality. The economists' job was to help it “perform better according to its own lights”—part neutral adviser, part advocate. That said, the RAWP's history also shows that this budding discipline did not automatically gain influence. Instead, it achieved access to power through the offices of others. An earlier generation of expert advisers paved the way, exemplified by Brian Abel-Smith, whose economics was grounded in social administration. The growing interest of epidemiologists in the effectiveness and efficiency of health services also facilitated their arrival.
Another central theme was the behavior of civil servants as advocacy coalition actors. This idea is not central to ACF theory as originally conceived but was prominent here in the importance of buy-in by midlevel bureaucrats who later became senior officials. They represented a new cadre of administrators open to the management sciences of operational research, statistics, and economics. Their support turned partly upon faith in the technical aspects of the formula, which despite its imperfections they found workable, transparent, and intellectually coherent. And, notwithstanding their retention of some clientelist powers to preserve stable service delivery in London, they also remained committed to the larger principle of equity of access, which extended beyond the transitory leadership of ministers.
There are limits, though, to the assumptions we can make about the instantiation of core social democratic beliefs, notwithstanding the logic of the ACF approach. It is manifestly the case that since the mid-1970s the British political class, Conservative and Labour alike, has loosened its commitment to welfare (Castles 1998). Comparison with other advanced industrial economies shows that across the spectrum of policy the state has retreated from universalism and social rights (Bambra 2006). Even after the “New Labour” era, inequalities of health outcome, as measured both by life expectation and by disability-free life years, have manifested “no narrowing of the gap” over time or space (Marmot Review 2010: 48). The coterminosity of this turn with the RAWP era lays bare some contradictions of the policy goal of equity. Where fairness sits comfortably with allocative efficiency, it is more likely to proceed. Where it does not, the appeal to social justice is harder to meet.
Early versions of this article were presented at the European Association for the History of Medicine and Health Conference, “Economics and Values in the History of Medicine and Health,” September 2015, Cologne, Germany, and at the Modern British History Seminar, Nanjing University, China, April 2016. We thank participants for their feedback and are also grateful to Virginia Berridge and Sally Sheard for their advice. Research for this article was sponsored by the Wellcome Trust (awards 088670/Z/09/Z and WT098457), whose support we gratefully acknowledge.