The Inverse Care Law
The Lancet: Saturday 27 February 1971
Glyncorrwg Health Centre, Port Talbot, Glamorgan, Wales
The availability of good medical care tends to vary inversely with the need for the population served. This inverse care law operates more completely where medical care is most exposed to market forces, and less so where such exposure is reduced. The market distribution of medical care is a primitive and historically outdated social form, and any return to it would further exaggerate the maldistribution of medical resources.
Interpreting the Evidence
The existence of large social and geographical inequalities in mortality and morbidity in Britain is known and not all of them are diminishing. Between 1934 and 1968, weighted mean standardised mortality from all causes in the Glamorgan and Monmouthshire valleys rose from 128% of England and Wales rates to 131%. Their weighted mean infant mortality rose from 115% of England and Wales rates to 124% between 1921 and 1968.1 The Registrar General's last Decennial Supplement on Occupational Mortality for 1949-53 showed combined social classes I and II (wholly non-manual), with a standardised mortality from all causes 18% below the mean, and combined social classes IV and v (wholly manual) 5% above it. Infant mortality was 37% below the mean for social class I (professional) and 38% above it for social class V unskilled manual).
A just and rational distribution of the resources of medical care should show parallel social and geographical differences, or at least a uniform distribution. The common experience was described by Titmuss in 1968: "We have learnt from 15 years' experience of the Health Service that the higher income groups know how to make better use of the service; they tend to receive more specialist attention; occupy more of the beds in better equipped and staffed hospitals; receive more elective surgery have better maternal care, and are more likely to get psychiatric help and psychotherapy than low-income groups- particularly the unskilled."
These generalisations are not easily proved statistically, because most of the statistics are either not available (for instance, outpatient waiting-lists by area and social class, age and cause specific hospital mortality rates by area and social class the relation between ante-mortem and postmortem diagnosis by area and social class an hospital staff shortage by area) or else described essentially use-rates. Use-rates may be interpreted either as evidence of high morbidity among high users, or of disproportionate benefit drawn by them from the National Health Service. By piling up the valid evidence that poor people in Britain have higher consultation and referral rates at all levels of the N.H.S., and by denying that these reflect actual differences in morbidity Rein , has tried to show that Titmuss's opinion is incorrect, and that there are no significant gradients in the quality or accessibility of medical care in the N.H.S. between social classes.
Class gradients in mortality are an obvious obstacle to this view. Of these Rein says: "One conclusion reached … is that since the lower classes have higher death rates, then they must be both sicker or less likely to secure treatment than other classes…it is useful to examine selected diseases in which- there is a clear mortality class gradient and then compare these rates with the proportion of patients in each class that consulted their physician for treatment of these diseases… "
He cites figures to show that high death-rates may be associated with low consultation-rates for some diseases, and with high rates for others, but, since the pattern of each holds good through all social classes, he concludes that a reasonable inference to be drawn from these findings is not that class mortality is an index of class morbidity but that for certain diseases treatment is unrelated to outcome. Thus both high and low consultation rates can yield high mortality rates for specific diseases. These data do not appear to lead to the compelling conclusion that mortality votes can be easily used as an area of class related morbidity."
This is the only argument mounted by Rein against the evidence of mortality differences, and the reasonable assumption that these probably represent the final outcome of larger differences in morbidity. Assuming that "votes" is a misprint for "rates", I still find that the more one examines this argument the less it means. To be fair, it is only used to support the central thesis that "the availability of universal free-on-demand, comprehensive services would appear to be a crucial factor in reducing class inequalities in the use of medical care services". It certainly would, but reduction is not abolition, as Rein would have quickly found if his stay in Britain had included more basic fieldwork in the general practitioner's surgery or the outpatient department.
There is massive but mostly non-statistical evidence in favour of Titmuss's generalisations. First of all there is the evidence of social history. James described the origins of the general-practitioner service in industrial and coalmining areas, from which the present has grown:
"The general practitioner in working-class areas discovered the well-tried business principle of small profits with a big turnover where the population was large and growing rapidly; it paid to treat a great many people for a small fee. A waiting-room crammed with patients, each representing 2s. 6d. for a consultation … not only gave a satisfactory income but also reduced the inclination to practise clinical medicine with skilful care, to attend clinical meetings, or to seek refreshment from the scientific literature. Particularly in coalmining areas, workers formed themselves into clubs to which they contributed a few pence a week, and thus secured free treatment from the club doctor for illness or accident. The club system was the forerunner of health insurance and was a humane and desirable social development. But, like the 'cash surgery', it encouraged the doctor to undertake the treatment of more patients than he could deal with efficiently. It also created a difference between the club patients and those who could afford to pay for medical attention … in these circumstances it is a tribute to the profession that its standards in industrial practices were as high as they were. If criticism is necessary, it should not be of the doctors who developed large industrial practices but of the leaders of all branches of the profession, who did not see the trend of general practice, or, having seen it, did nothing to influence it. It is particularly regrettable that the revolutionary conception of a National Health Service, which has transformed the hospitals of the United Kingdom to the great benefit of the community, should not have brought about an equally radical change in general practice. Instead, because of the shortsightedness of the profession, the N.H.S. has preserved and intensified the worst features of general practice.
This preservation and intensification was described by Collings in his study of the work of 104 general practitioners in 55 English practices outside London, including 9 completely and 7 partly industrial practices, six months after the start of the N.H.S. Though not randomly sampled, the selection of practices was structured in a reasonably representative manner. The very bad situation he described was the one I found when I entered a slum practice in Notting Hill in 1953, rediscovered in all but one of five industrial practices where I acted as locum tenens in 1961, and found again when I resumed practice in the South Wales valleys. Collings said:
"the working environment of general practitioners in industrial areas was so limiting that their individual capacity as doctors counted very little. In the circumstances prevailing, the most essential qualification for the industrial G.P. . . . is ability as a snap diagnostician-an ability to reach an accurate diagnosis on a minimum of evidence . . . the worst elements of general practice are to be found in those places where there is the greatest and most urgent demand for good medical service…. Some conditions of general practice are bad enough to change a good doctor into a bad doctor in a very short time. These very bad conditions are to be found chiefly in industrial areas."
In a counter-report promoted by the British Medical Association, Hadfield contested all of Collings' conclusions, but, though his sampling was much better designed, his criticism was guarded to the point of complacency, and most vaguely defined. One of Collings' main criticisms – that purpose-built premises and ancillary staff were essential for any serious up-grading of general practice – is only now being taken seriously; and even the present wave of health-centre construction shows signs of finishing almost as soon as it has begun, because of the present climate of political and economic opinion at the level of effective decision.
Certainly in industrial and mining areas health centres exist as yet only on a token basis and the number of new projects is declining. Aneurin Bevan described health centres as the cornerstone of the general practitioner service under the N.H.S., before the long retreat began from the conceptions born in the 1930s and apparently victorious in 1945. Health-centre construction was scrapped by ministerial circular in January, 1948, in the last months of gestation of the new service; we have had to do without them for 22 years, during which a generation of primary care was stunted.
Despite this unpromising beginning, the N.H.S brought about a massive improvement in the delivery of medical care to previously deprived sections of the people and areas of the country. Former Poor Law hospitals were upgraded and many acquired fully trained specialist and ancillary staff and supporting diagnostic departments for the first time. The backlog of untreated disease dealt with in the first years of the service was immense, particularly in surgery and gynecology. A study of 734 randomly sampled families in London and Northampton in 1961 showed that in 99% of the families someone had attended hospital as an outpatient, and in 82% someone had been admitted to hospital. The study concluded:
"When thinking of the Health Service mothers are mainly conscious of the extent to which services have become available in recent years. They were more aware of recent changes in health services than of changes in any other service. Nearly one third thought that more money should be spend on health services, not because they thought them bad but because 'they are so important', because 'doctors and nurses should be paid more' or because ' there shouldn't be charges for treatment'. Doctors came second to relatives and friends in the list of those who had been helpful in times of trouble."
Among those with experience of pre-war services, appreciation for the N.H.S., often uncritical appreciation, is almost universal -so much so that, although most London teaching-hospital consultants made their opposition to the new service crudely evident to their students in 1948 and the early years, and only a courageous few openly supported it, few of them appear to recall this today. The moral defeat of the very part-time, multi-hospital consultant, nipping in here and there between private consultations to see how his registrar was coping with his public work, was total and permanent; lip-service to the N.H.S. is now mandatory. At primary-care level, private practice ceased to be relevant to the immense majority of general practitioners, and has failed to produce evidence of the special functions of leadership and quality claimed for it, in the form of serious research material. On the other hand, despite the massive economic disincentives to good work, equipment, and staffing in the N.H.S. until a few years ago, an important expansion of well-organised, community-oriented, and self-critical primary care has taken place, mainly through the efforts of the Royal College of General Practitioners, the main source of this vigour is the democratic nature of the service – the fact that it is comprehensive and accessible to all, and the fact that clinical decisions are therefore made more freely than ever before. The service at least permits, if it does nor yet really encourage, general practitioners to think and act in terms of the cure of a whole defined community, as well as of whole persons rather than diseases. Collings seems very greatly to have underestimated the importance of these changes and the extent to which they were to over shadow the serious faults of the service – and these were faults of too little change, rather than too much. There have in fact been very big improvements in the quality and accessibility of care both at hospital and primary care level, for all classes and in all areas.
Selective Redistribution of Care
Given the large social inequalities of mortality and morbidity that undoubtedly existed before the 1939-45 war and the equally large differences in the quality and accessibility of medical resources to deal with them, it was clearly not enough simply to improve care for everyone:some selective redistribution was necessary, and some has taken place. But how much, and is the redistribution accelerating, stagnating, or even going into reverse?
Ann Cartwright's study of 1370 randomly sampled adults in representative areas of England, and their 552 doctors gave some evidence on what had and what had not been achieved. She confirmed a big improvement in the quality of primary care in 1961 compared with 1948 but also found just the sort of class differences suggested by Titmuss. The consultation-rate of middle class patients at ages under 45 was 53% less than that of working-class patients, but at ages over 75 they had a consultation-rate 62% higher; and between these two age-groups there was stepwise progression. I think it is reasonable to interpret this as evidence that middle-class consultations had a higher clinical content at all ages, that working-class consultations below retirement age had a higher administrative content, and that the middle-class was indeed able to make more effective use of primary care. Twice as many middle-class patients were critical of consulting-rooms and of their doctors, and three times as many of waiting-rooms, as were working-class patients yet Cartwright and Marshall  in another study found that in predominantly working-class areas 80% of the doctors' surgeries were built before 1900 and only 5% since 1945; in middle-class areas less than 50% were built before 1900, and 25% since 1945. Middle-class patients were both more critical and better served. Three times as many middle class patients were critical of the fullness of explanations to them about their illnesses; it is very unlikely that this was because they actually received less explanation than working-class patients, and very likely that they expected, sometimes demanded, and usually received much more. Cartwright's study of hospital care showed the same social trend for explanations by hospital staff. The same study looked at hospital patients' general practitioners, and compared those working in middle-class and in working-class areas: more middle-class area G.P.s had lists under 2000 than did working-class area G.P.s, and fewer had lists over 2500; nearly twice as many had higher qualifications, more had access to contrast-Media X-rays, nearly five times as many had access to physiotherapy, four times as many had been to Oxford or Cambridge, five times as many had been to a London medical school, twice as many held hospital appointments or hospital beds in which they could care for their own patients, and nearly three times as many sometimes visited their patients when they were in hospital under a specialist. Not all of these differences are clinically significant; so far the record of Oxbridge and the London teaching hospitals compares unfavourably with provincial medical schools for training oriented to the community. But the general conclusion must be that those most able to choose where they will work tend to go to middle-class areas, and that the areas with highest mortality and morbidity tend to get those doctors who are least able to choose where they will work. Such a system is not likely to distribute the doctors with highest morale to the places where that morale is most needed. Of those doctors who positively choose working-class areas, a few will be attracted by large lists with a big income and an uncritical clientele; many more by social and family ties of their own. Effective measures of redistribution would need to take into account the importance of increasing the proportion of medical students from working-class families in areas of this sort; the report of the Royal Commission on Medical Educationshowed that social class I (professional and higher managerial), which is 2.8% of the population, contributed 34.5% of the final-year medical students in 1961, and 39.6% of the first-year students in 1966, whereas social class iii (skilled workers, manual and non-manual), which is 49.9% of the population, contributed 27.9% of the final-year students in 1961 and 21.7% of the first-year in 1966. The proportion who had received State education was 43.4% in both years, compared with 70.9% of all school-leavers with 3 or more A-levels. In other words, despite an increasing supply of wellqualified State-educated school-leavers, the overrepresentation of professional families among medical students is increasing. Unless this trend is reversed, the difficulties of recruitment in industrial areas will increase from this cause as well, not to speak of the support it will give to the officers/other ranks' tradition in medical care and education.
The upgrading of provincial hospitals in the first few years after the Act certainly had a geographical redistributive effect, and, because some of the wealthiest areas of the country are concentrated in and around London, it also had a socially redistributive effect. There was a period in which the large formerly local authority hospitals were accelerating faster than the former voluntary hospitals in their own areas, and some catching-up took place that was socially redistributive. But the better-endowed, better-equipped, better-staffed areas of the service draw to themselves more and better staff, and more and better equipment, and their superiority is compounded. While a technical lead in teaching hospitals is necessary and justified, these advantages do not apply only to teaching hospitals, and even these can be dangerous if they encourage complacency about the periphery, which is all too common. As we enter an era of scarcity in medical staffing and austerity in Treasury control, this gap will widen, and any social redistribution that has taken place is likely to be reversed.
Redistribution of general practitioners also took place at a fairly rapid pace in the early years of the N.H.S., for two reasons. First, and least important, were the inducement payments and area classifications with restricted entry to over-doctored areas. These may have been of value in discouraging further accumulation of doctors in the Home Counties and on the coast, but Collings was right in saying that "any hope that financial reward alone will attract good senior practitioners back to these bad conditions is illusory; the good doctor will only be attracted into industrial practice by providing conditions which will enable him to do good work". The second and more important reason is that in the early years of the N.H.S. it was difficult for the increased number of young doctors trained during and just after the 1939-45 war to get posts either in hospital or in general practice, and many took the only positions open to them, bringing with them new standards of care. Few of those doctors today would choose to work in industrial areas, now that there is real choice; we know that they are not doing so. Of 169 new general practitioners who entered practice in under-doctored areas between October, 1968, and October, 1969, 164 came from abroad.  The process of redistribution of general practitioners ceased by 1956, and by 1961 had gone into reverse; between 1961 and 1967, the proportion of people in England and Wales in under-doctored areas rose from 17% to 34%.
Increasing List Size
The quality and traditions of primary medical care in industrial and particularly in mining areas are, I think, central to the problem of persistent inequality in morbidity and mortality and the mismatched distribution of medical resources in relation to them. If doctors in industrial areas are to reach take-off speed in reorganising their work and giving it more clinical content, they must be free enough from pressure of work to stand back and look at it critically. With expanding lists this will be for the most part impossible; there is a limit to what can be expected of doctors in these circumstances, and the alcoholism that is an evident if unrecorded occupational hazard among those doctors who have spent their professional lives in industrial practice is one result of exceeding that limit. Yet list sizes are going up, and will probably do so most where a reduction is most urgent. Fry and Last have criticised the proposals of the Royal Commission on Medical Education for an average annual increase of 100 doctors in training over the next 25 years, which would raise the number of economically active doctors per million population from 1181 in 1965 to 1801 in 1995. They claim that there are potential increases in productivity in primary care, by devolution of work to ancillary and paramedical workers and by rationalisation of administrative work, that would permit much larger average list sizes without loss of intimacy in personal care, or decline in clinical quality. Of course, much devolution and rationalisation of this sort is necessary, not to cope with rising numbers, but to make general practice more clinically effective and satisfying, so that people can be seen less often but examined in greater depth. If clinically irrelevant work can be devoluted or abolished, it is possible to expand into new and valuable fields of work such as those opened up by Balint and his school, and the imminent if not actual possibilities of presymptomatic diagnosis and screening, which can best be done at primary-care level and is possible within the present resources of general practice.But within the real political context of 1971 the views of Last, and of Fry from his experience of London suburban practice which is very different from the industrial areas discussed here, are dangerously complacent.
Progressive change in these industrial areas depends first of all on two things, which must go hand in hand: accelerated construction of health centres, and the reduction of list sizes by a significant influx of the type of young doctor described by Barber in 1950
"so prepared for general practice, and for the difference between what he is taught to expect and what he actually finds, that he will adopt a fighting attitude against poor medicine-that is to say, against hopeless conditions for the practice of good medicine. The young man must be taught to be sufficiently courageous, so that when he arrives at the converted shop with the drab battered furniture, the couch littered with dusty bottles, and the few rusty antiquated instruments, he will make a firm stand and say "Iwill not practise under these conditions; I will have more room, more light, more ancillary help, and better equipment.'"
Unfortunately, the medical ethic transmitted by most of our medical schools, at least the majority that do not have serious departments of general practice and community medicine, leads to the present fact that the young man just does not arrive at the converted shop; he has more room, more light, more ancillary help, and better equipment by going where these already exist, and no act of courage is required. The career structure and traditions of our medical schools make it clear that time spent at the periphery in the hospital service, or at the bottom of the heap in industrial general practice, is almost certain disqualification for any further advancement. Our best hope of obtaining the young men and women we need lies in the small but significant extent to which medical students are beginning to reject this ethic, influenced by the much greater critical awareness of students in other disciplines. Some are beginning to question which is the top and which the bottom of the ladder, or even whether there should be a ladder at all; and in the promise of the Todd report, of teaching oriented to the patient and the community rather than toward the doctor and the disease, there is hope that this mood in a minority of medical students may become incorporated into a new and better teaching tradition. It is possible that we may get a cohort of young men and women with the sort of ambitions Barber described, and with a realistic attitude to the battles they will have to fight to get the conditions of work and the buildings and equipment they need, in the places that need them; but we have few of them now. The prospect for primary care in industrial areas for the next ten years is bad; list sizes will probably continueand the pace of improvement in quality of primary care is likely to fall.
Recruitment to General Practice in South Wales
Although the most under-doctored areas are mainly of the older industrial type, the South Wales valleys have relatively good doctor/patient ratios, partly because of the declining populations, and partly because the area produces an unusually high proportion of its own doctors, who often have kinship ties nearby and may be less mobile on that account. (In Williams' survey of general practice in South Wales 72% of the 68 doctors were born in Wales and 43% had qualified at the Welsh National School of Medicine ) On Jan. 1, 1970, of 36 South Wales valley areas listed, only 4 were designated as under-doctored. However, this situation is unstable; as our future becomes more apparently precarious, as pits close without alternative local employment, as unemployment rises, and out-migration that is selective for the young and healthy increases, doctors become subject to the same pressures and uncertainties as their patients. Recruitment of new young doctors is becoming more and more difficult, and dependent on doctors from abroad. Many of the industrial villages are separated from one another by several miles, and public transport is withering while as yet comparatively few have cars, so that centralisation of primary care is difficult and could accelerate the decay of communities. These communities will not disappear, because most people with kinship ties are more stubborn than the planners, and because they have houses here and cannot get them where the work is; the danger is not the disappearance of these communities, but their persistence below the threshold of viability, with accumulating sickness and a loss of the people to deal with it.
What Should Be Done ?
Medical services are not the main determinant of mortality or morbidity; these depend most upon of standards of nutrition, housing, working environment, and education, and the presence or absence of war. The high mortality and morbidity of the South Wales valleys arise mainly from lower standards in most of these variables now and in the recent past, rather than from lower standards of medical care. But that is no excuse for failure to match the greatest need with the highest standards of care. The bleak future now facing mining communities, and others that may suffer similar social dislocation as technical, change blunders on without agreed social objectives, cannot be altered by doctors alone; but we do have a duty to draw attention to the need for global costing when it comes to policy decisions on redevelopment or decay of established industrial communities. Such costing would take into account the full social costs and not only those elements of profit and loss traditionally recognised in industry.
The improved access to medical care for previously deprived sections under the N.H.S. arose chiefly from the decision to remove primary-care services from exposure to market forces. The consequences of distribution of care by the operation of the market were unjust and irrational, despite all sorts of charitable modifications. The improved possibilities for constructive planning and rational distribution of resources because of this decision are immense, and even now are scarcely realised in practice. The losses predicted by opponents of this change have not in fact occurred; consultants who no longer depend on private practice have shown at least as much initiative and responsibility as before, and the standards attained in the best N.H.S. primary care are at least as good as those in private practice. It has been proved that a national health service can run quite well without the profit motive, and that the motivation of the work itself can be more powerful in a decommercialised setting. The gains of the service derive very largely from the simple and clear principles on which it was conceived: a comprehensive national service, available to all, free at the time of use, non-contributory, and financed from taxation. Departures from these principles, both when the service began (the tripartite division and omission of family-planning and chiropody services) and subsequently (dental and prescription charges, rising direct contributions, and relative reductions in financing from taxation), have not strengthened it. The principles themselves seem to me to be worth defending, despite the risk of indulging in unfashionable value-judgments. The accelerating forward movement of general practice today, impressively reviewed in a symposium on group practice held by the Royal College of General Practitioners,is a movement (not always conscious) toward these principles and the ideas that prevailed generally among the minority of doctors who supported them in 1948, including their material corollary, group practice from health centres. The doctor/patient relationship, which was held by opponents of the Act to depend above all on a cash transaction between patient and doctor, has been transformed and improved by abolishing that transaction. A general practitioner can now think in terms of service to a defined community, and plan his work according to rational priorities.
Godber has reviewed this question of medical priorities, which he sees as a new feature arising from the much greater real effectiveness of modern medicine, which provides a wider range of real choices, and the great costliness of certain forms of treatment. While these factors are important, there are others of greater importance which he omits. Even when the content of medicine was overwhelmingly palliative or magical -say, up to the 1914-18 war- the public could not face the intolerable facts any more than doctors could, and both had as great a sense of priorities as we have; matters of life and death arouse the same passions when hope is illusory as when it is real, as the palatial Swiss tuberculosis sanatoria testify. The greatest difference, I think, lies in the transformation in social expectations. In 1914 gross inequality and injustice were regarded as natural by the privileged, irresistible by the unprivileged, and inevitable by nearly everyone. This is no longer true; inequality is now politically dangerous once it is recognised, and its inevitability is believed in only by a minority. Diphtheria became preventable in the early 1930s, yet there were 50,000 cases in England and Wales in 1941 and 2400 of them died. I knew one woman who buried four of her children in five weeks during an outbreak of diphtheria in the late 1930s. No systematic national campaign of immunisation was begun until well into the 1939-45 war years, and, if such a situation is unthinkable today, the difference is political rather than technical. Godber rightly points to the planning of hospital services during the war as one starting-point of the change; but he omits the huge social and political fact of 1945: that a majority of people, having experienced the market distribution of human needs before the war, and the revelation that the market could be overridden during the war for an agreed social purpose, resolved never to return to the old system.
Perhaps reasonable economy in the distribution of medical care is imperilled most of all by the old ethical concept of the isolated one-doctor/one-patient relationship, pushed relentlessly to its conclusion regardless of cost-or to put it differently, of the needs of others.
The pursuit of the very best for each patient who needs it remains an important force in the progress of care; a young person in renal failure may need a doctor who will fight for dialysis, or a grossly handicapped child one who will find the way to exactly the right department, and steer past the defeatists in the wrong ones. But this pursuit must pay some regard to humane priorities, as it may not if the patient is a purchaser of medical care as a commodity. The idealised, isolated doctor/patient relationship, that ignores the needs of other people and their claims on the doctor's time and other scarce resources, is incomplete and distorts our view of medicine. During the formative period of modern medicine this ideal situation could be realised only among the wealthy, or, in the special conditions of teaching hospitals, among those of the unprivileged with "interesting" diseases. The ambition to practise this ideal medicine under ideal conditions still makes doctors all over the world leave those who need them most, and go to those who need them least, and it retards the development of national schools of thought and practice in medicine, genuinely based on the local content of medical care. The ideal isolated doctor/patient relation has the same root as the 19th-century preoccupation with Robinson Crusoe as an economic elementary particle; both arise from a view of society that can perceive only a contractual relation between independent individuals. The new and hopeful dimension in general practice is the recognition that the primary-care doctor interacts with individual members of a defined community. Such a community-oriented doctor is not likely to encourage expensive excursions into the 21st century, since his position makes him aware, as few specialists can be, of the scale of demand at its point of origin, and will therefore be receptive to common-sense priorities. It is this primary-care doctor who in our country initiates nearly every train of causation in the use of sophisticated medical care, and has some degree of control over what is done or not done at every point. The commitment is a great deal less open-ended than many believe; we really do not prolong useless, painful, or demented lives on the scale sometimes imagined. We tend to be more interested in the people who have diseases than in the diseases themselves, and that is the first requirement of reasonable economy and a humane scale of priorities.
Return to the Market ?
The past ten years have seen a spate of papers urging that the N.H.S. be returned wholly or partly to the operation of the market. Jewkes," Lees, Seale, and the advisory planning panel on health services financing of the British Medical Association have all elaborated on this theme. Their arguments consist in a frontal attack on the policy of removing health care from the market, together with criticism of faults in the service that do not necessarily or even probably depend on that policy at all, but on the failure of Governments to devote a sufficient part of the national product to medical care. These faults include the stagnation in hospital building and senior staffing throughout the 1950s, the low wages throughout the service up to consultant level, over-centralised control, and failure to realise the objective of social and geographical equality in access to the best medical care.
None of these failings is intrinsic to the original principles of the N.H.S.; all have been deplored by its supporters, and with more vigour than by these critics. The critics depend heavily on a climate of television and editorial opinion favouring the view that all but a minority of people are rich enough and willing to pay for all they need in medical care (but not through taxation), and that public services are a historically transient social form, appropriate to indigent populations, to be discarded as soon as may be in favour of distribution of health care as a bought commodity, provided by competing entrepreneurs. They depend also on the almost universal abdication of principled opposition to these views, on the part of its official opponents. The former Secretary of State for Social Services, Mr. Richard Crossman, has agreed that the upper limit of direct taxation has been reached, and that "we should not be afraid to look for alternative sources of revenue less dependent on the Chancellor's whims. . . . I should not rule out obtaining a higher proportion of the cost of the service from the Health Service contribution." This is simply a suggestion that rising health costs should be met by flat-rate contributions unrelated to income – an acceptance of the view that the better-off are taxed to the limit, but also that the poor can afford to pay more in proportion. With such opposition, it is not surprising that more extravagant proposals for substantial payments at the time of illness, for consultations, home visits, and hospital care, are more widely discussed and advocated than ever before.
Seale proposed a dual health service, with a major part of hospital and primary care on a fee paying basis assisted by private insurance, and a minimum basic service excluding the "great deal of medical care which is of only marginal importance so far as the life or death or health of the individual is concerned. Do those who want the Health Service to provide only the best want the frills of medical care to be only the best, or have they so little understanding of the nature of medical care that they are unaware of the existence of the frills?" Frills listed by Seale are: "time, convenience, freedom of choice, and privacy". He says that " it is precisely these facets of medical care-the 'middle class' standards-which become more important to individuals as they become more prosperous ". Do they indeed? Perhaps it is not much that they (and other frills such as courtesy, and willingness to listen and to explain, that may be guaranteed by payment of a fee) become more important as that they become accessible. The possession of a new car is an index of prosperity; the lack of one is not evidence that it is not wanted. Real evidence should be provided that it is possible to separate the components of medical care into frills that have no bearing on life, death, or health, and essentials which do. Life and happiness most certainly can hang on a readiness to listen, to dig beneath the presenting symptom, and to encourage a return when something appears to have been left unsaid. And not only the patient-all patients value these things; to practice without them makes a doctor despise his trade and his patients. Where are the doctors to be found to undertake this veterinary care? It need not be said; those of us who already work in industrial areas are expected to abandon the progress we have made toward universal, truly personal care and return to the bottom half of the traditional double standard.
This is justified in anticipation by Seale:
"some doctors are very much better than others and this will always be so, and the standard of care provided will vary within wide limits . . . the function of the state is, in general, to do those things which the individual cannot do and to assist him to do things better. It is not to do for the individual what he can well do for himself …I should like to see reform of the Health Service in the years ahead which is based on the assumption of individual responsibility for personal health, with the State's function limited to the prevention of real hardship and the encouragement of personal responsibility."
Lees central thesis is that medical care is a commodity that should be bought and sold as any other, and would be optimally distributed in a free market. A free market in houses or shoes does not distribute them optimally, rich people get too much and the poor too little, and the same is true of medical care. He claims that the N.H.S. violates "natural" economic law, and will fail if a free market is not restored, in some degree at least, and that in a free market "we would spend more on medical care than the government does on our behalf ". If the "we" in question is really all of us, no problem exists; we agree to pay higher income tax and/or give up some million-pound bombers or whatever, and have the expanding service we want. But if the "we" merely means "us" as opposed to "them", it means only that the higher social classes will pay more for their own care, but not for the community as a whole. They will then want value for their money, a visible differential between commodity-care and the utility brand; is it really possible, let alone desirable, to run any part of the health service in this way? Raymond Williams put his finger on the real point here:
"we think of our individual patterns of use in the favourable terms of spending and satisfaction, but of our social patterns of use in the unfavourable terms of deprivation and taxation. It seems a fundamental defect of our society that social purposes are largely financed out of individual incomes, by a method of rates and taxes which makes it very easy for us to feel that society is a thing that continually deprives and limits us – without this we could all be profitably spending…. We think of 'my money'.. . in these naive terms, because parts of our very idea of society are withered at root. We can hardly have any conception, in our present system, of the financing of social purposes from the social product…"
Seale thinks the return to the market would help to provide the continuous audit that is certainly necessary to intelligent planning in the health service:
"In a health service provided free of charge efficient management is particularly difficult because neither the purpose nor the product of the organisation can be clearly defined, and because there are few automatic checks to managerial incompetence. . . . In any large organisation management requires quantitative information if it is to be able to analyse a situation, make a decision, and know whether its actions have achieved the desired result. In commerce this quantitative information is supplied primarily in monetary terms. By using the simple, convenient and measurable criterion of profit as both objective and product, management has a yardstick for assessing the quality of the organisation and the effectiveness of its own decisions."
The purposes and desired product of medical care are complex, but Seale has given no evidence to support his opinion that they cannot be clearly measured or defined; numerous measures of mortality, morbidity, and cost and labour effectiveness in terms of them are available and are (insufficiently) used. They can be developed much more easily in a comprehensive service outside the market than in a fragmented one within it. We already know that we can study and measure the working of the National Health Service more cheaply and easily than the diverse and often irrational medical services of areas of the United States of comparable population, though paradoxically there are certain techniques of quality control that are much more necessary in America than they are here. Tissue committees monitor the work of surgeons by identifying excised normal organs, and specialist registration protects the public from spurious claims by medical entrepreneurs. The motivation for fraud has almost disappeared from the N.H.S., and with it the need for certain forms of audit. A market economy in medical care leads to a number of wasteful trends that are acknowledged problems in the United States. Hospital admission rates are inflated to make patients eligible for insurance benefit, and, according to Fry :
"In some areas, particularly the more prosperous, competition for patients exists between local hospitals, since lack of regional planning has led to an excess of hospital facilities in some localities. In such circumstances hospital administrators are encouraged to use public relations officers and other means of self-advertisement…. This competition also leads to certain hospital 'status symbols', where features such as the possession of a computer; the possession of a 'cobalt bomb' unit; the ability to perform open-heart surgery albeit infrequently; and the listing of a neurosurgeon on the staff are all current symbols of status in the eyes of certain groups of the public. Even small hospitals of 150-200 beds may consider such features as necessities."
And though these are the more obvious defects of substituting profit for the normal and direct objectives of medical care, the audit by profit has another and much more serious fault; it concentrates all our attention on tactical efficiency, while ignoring the need for strategic social decisions. A large advertising agency may be highly efficient and profitable, but is this a measure of its socially useful work? It was the operation of the self-regulating market that resulted in a total expenditure on all forms of advertising of £455 million in 1960, compared with about £500 million on the whole of the hospital service in the same year." The wonderfully self-regulating market does sometimes show a smaller intelligence than the most ignorant human voter.
All these trends of argument are gathered together in the report of the B.M.A. advisory planning panel on health services financing, which recommends another dual service, one for quality and the other for minimum necessity. It states its view with a boldness that may account for its rather guarded reception by the General Medical Services Committee of the B.M.A.:
The only sacrifice that would have to be made would be the concept of equality within the National Health Service . . . any claim that the N.H.S. has achieved its aim of providing equality in medical care is an illusion. In fact, absolute equality could never be achieved under any system of medical care, education or other essential service to the community. The motives for suggesting otherwise are political and ignore human factors."
The panel overlooks the fact that absolute correctness of diagnosis or absolute relief of suffering are also unattainable under any system of medical care; perhaps the only absolute that can be truly attained is the blindness of those who do not wish to see, and the human factor we should cease to ignore is the opposition of every privileged group to the loss of its privilege.
The Inverse Care Law
In areas with most sickness and death, general practitioners have more work, larger lists, less hospital support, and inherit more clinically ineffective traditions of consultation, than in the healthiest areas; and hospital doctors shoulder heavier case-loads with less staff and equipment, more obsolete buildings, and suffer recurrent crises in the availability of beds and replacement staff. These trends can be summed up as the inverse care law: that the availability of good medical care tends to vary inversely with the need of the population served.
If the N.H.S. had continued to adhere to its original principles, with construction of health centres a first priority in industrial areas, all financed from taxation rather than direct flat-rate contribution, free at the time of use, and fully inclusive of all personal health services, including family planning, the operation of the inverse care law would have been modified much more than it has been; but even the service as it is has been effective in redistributing care, considering the powerful social forces operating against this. If our health services had evolved as a free market, or even on a fee-for-item-of-service basis prepaid by private insurance, the law would have operated much more completely than it does; our situation might approximate to that in the United States, with the added disadvantage of smaller national wealth. The force that creates and maintains the inverse care law is the operation of the market, and its cultural and ideological superstructure which has permeated the thought and directed the ambitions of our profession during all of its modern history. The more health services are removed from the force of the market, the more successful we can be in redistributing care away from its "natural" distribution in a market economy; but this will be a redistribution, an intervention to correct a fault natural to our form of society, and therefore incompletely successful and politically unstable, in the absence of more fundamental social change.
I am grateful to Prof. W. R. S. Doll,,F.R.S., for advice and criticism; to Miss M. Hammond, librarian of the Royal College of General Practitioners; and to the clerks of the Glamorgan and Monmouthshire Executive Councils for the National Health Service for data on recruitment of general practitioners in their areas.
 Hart, JT J R Coll gen Practners (in the press)
 Titmuss RM Commitment to Welfare London 1968
 Rein M Am Hosp Ass 1969 43, 43
 Rein M New Society Nov 20 1969 p 807
 James EF Lancet 1961 I 1361
 Collings JS ibid 1950 i555
 Hadfield SJ Br med J 1953 ii 683
 Family Needs and Social Services. Political and Economic Planning London 1961
 Cartwright A Patients and their doctors London 1967
 Cartwright A, Marshall J Med Care 1965, 3,69
 Cartwright A Human Relations and Hospital Care London 1964
 Report of the Royal Commission on Medical Education 1965-8; p139 London 1968
 DHSS Annual report for 1969, London 1970
 General Practice Today , Office of Health Economics, paper no 28, London 1968
 Fry J J.R.Coll.gen Practnrs, 1969, 17,355
 Last J.M. Lancet. 1968,ii,223
 Report of the Royal Commission on Medical Education 1965-8; p139. London 1968
 Balint, M. The Doctor, his Patient and the Illness. London 1964
 Hart J.T. Lancet 1970, ii,223
 Barber G. ibid. 1950, i,781
 Williams, W.O. A Study of General Practitioners Workload in South Wales 1965-6. RCGP reports from general practice no 12 January 1970
 J R coll. Gen. Practnrs, 1970, suppl. 2
 Godber G. ibid. 1970, 20, 313
 Morris J.N. Uses of Epidemiology. London 1967
 Jewkes, J., Jewkes, S. The genesis of the British National Health Service. Oxford 1961
 Lees, D.S. Health through choice. An Economic Study of the British National Health Service. Hobart paper no 14, Institute of Economic Affairs, London 1961
 Seale, J. Br. Med J. 1962, ii, 598
 Report of the Advisory Panel of the BMA on Health Services Financing. BMA, London, 1970
 Crossman RHS Paying for the Social Services. Fabian Society, London, 1969
 Williams R. The Long Revolution. London 1961
 Seale J Lancet 1961, ii,476
 Fry J. Medicine in three Societies. Aylesbury, 1969
 Battisella, RM., Southby, RMF Lancet 1968, i,581
Some of the arguments in this paper were made earlier by Henry Sigerist