Select Committee on Health Written Evidence

Memorandum submitted by Professor Peter C Smith (CP 36)


  Most health care is directed at individual patients, seeking to improve the duration and quality of life. It is therefore perfectly feasible to charge patients a fee for their use of health care. Indeed it is worth recalling that until recently doctors in all countries relied mainly on patient fees to provide their income, and that this system is still dominant in much of the developing world. It was only in the latter half of the twentieth century that socialised medical care has become widespread in developed countries.

  User charges in health care have two broad roles: to raise finance for the health system, and to send signals to patients who would otherwise face a zero price for access to health care. Developed countries do not currently rely to any great extent on charges as a significant source of finance. However, there has been a persistent concern with the dangers of "moral hazard" in health care.[12] That is, in the absence of direct prices, patients may use health care when it is not warranted. Moreover, given the power of doctors to influence patient behaviour, moral hazard might be exacerbated by "supplier induced demand", particularly in systems where doctors' incomes rely directly on attracting high levels of business.[13]

  This note assesses the current role of user charges in developed health systems, and possible future options for the English National Health Service.


  Figure 1 shows that direct user charges (out of pocket payments) account for between 10% and 20% of health system revenue in high income countries. Most of the "other" private expenditure relates to voluntary private insurance. In particular, in countries such as France and Ireland, patients are in principle liable for quite high user charges. However, many citizens take out voluntary private health insurance to secure protection from out of pocket payments. Note that the OECD has not been able to report these data for the UK since 1997.

Figure 1

Private expenditure as % total health expenditure 2001

  (Source: OECD Health Data 2004).

  The especially low reliance on user charges in the United Kingdom reflects the founding principle of a NHS "free at the point of access". Eversley relates the fraught history of NHS charges, the imposition of which in 1951 led to the resignation of Aneurin Bevan, hastening the demise of the Atlee government[14] An attempt to abandon prescription charges by the Labour government in 1965 was soon reversed in 1967, and further increases were subsequently imposed in 1975 under pressure from the International Monetary Fund. In 2004 prescription charges in England accounted for income of £446 million, with only 8.9% of prescriptions directly attracting the full charge of £6.20.[15] The vast majority of prescriptions are exempt from charges on grounds of age (young and older people), sickness (certain chronic conditions), maternity, or low income. The Welsh and Scottish government intend to abolish prescription charges.


  Western Europe countries have traditionally sought to model their health systems on the principle of "solidarity". This implies universal coverage, and contributions to the financing of health care according to ability to pay, irrespective of age or level of sickness. User charges appear to contradict the principle of solidarity. Yet, as expenditure on health care has grown inexorably, there has been growing interest in imposing some modest charges.[16]

  In general, these new charges do not raise a significant volume of finance for the health system—indeed in some circumstances the sums involved are outweighed by the collection costs. Rather, the main purpose of these experiments is to encourage patients to use the health system to better purpose, by discouraging treatment when benefits are small and creating incentives for efficient use of services when it is justified. Examples of objectives underlying charging schemes include:

    —    moderating the number of physician and hospital visits;[17]

    —    encouraging use of cheaper generic drugs, through the use of "reference prices";[18]

    —    directing patients through gatekeeper physicians;[19]

    —    encouraging the use of less costly or higher quality "preferred" providers; and

    —    encouraging early discharge of patients from hospital.

  Most of these initiatives have been directed at cost containment, and many other experiments in a similar vein could be envisaged, such as charging patients for outpatient visits, but offering a full or partial rebate if the appointment is honoured (in order to discourage "did not attends"). Moreover, user charges could in principle be used to encourage healthier behaviour on the part of patients. For example, one could envisage a scheme of exemption from charges if a patient complies with a course of treatment in its entirety. There follows a brief sketch of a few recent European innovations.

  Sweden was one of the first of the traditional public sector systems to experiment with quite small user charges across a wide range of health services. Children and young people are generally exempt, and the maximum annual liability for charges has traditionally been set at quite a low level (€90 in 2001). Charging on this modest scale appears to have been generally accepted as reasonable, but it has resulted in reduced utilisation amongst low income patients, and a concern that equity of access may be compromised.[20]

  Some countries in eastern Europe are experiencing especially severe problems with financing health care, and are therefore experimenting with more radical approaches to charging, especially where a tradition of "informal" payments to doctors and other professionals exists.[21] A particularly ambitious scheme of "diagnosis based reimbursement" is being introduced in the Slovak republic. A national tariff for reimbursing providers is set for all interventions, according to diagnosis. Patients will then be reimbursed for a proportion of the costs of treatment, depending on the diagnosis group. The proportion reimbursed depends on the estimated benefits and costs of treatment, and there is full reimbursement for 33% of diagnoses. This scheme is consistent with the prescriptions of the economic theory of "optima" commodity taxation.[22] However, as experience unfolds, it will be important to see whether it is in practice sustainable, whether unintended behavioural responses on the part of doctors or patients emerge, and whether the lack of exemptions leads to especially adverse outcomes for poor and sick people.


  There is in general a shortage of reliable evidence on the impact of user charges on the utilisation of health care and the consequences for the health of patients. The major exception is the celebrated RAND experiment, under which over 2,000 US patients were randomly assigned to one of four charging regimes over an extended period.[23] One group of patients enjoyed complete freedom from charges, while those at the other extreme were charged 95% of fees for virtually all care, up to a maximum annual "catastrophic" liability of about $6,000 at current prices.

  The experiment resulted in consistent reductions in utilisation across all types of health care as the charges became more severe. For example, physician consultations varied from 4.55 per annum amongst those incurring no charges, to 2.73 amongst those in the highest charging scheme, a reduction of 40%. However, with one major exception, evaluation of the experiment did not detect any material variations in health outcome associated with charging. Researchers have therefore concluded that—for most of the population—charges succeeded in encouraging less profligate use of health care without serious health consequences.

  The one important exception was the finding that charging had a seriously adverse effect on those who were both poor and suffering from poor health. The RAND evaluation estimated that for this disadvantaged group there were a wide range of serious consequences, in spite of some cost subsidy for low-income families. For example, when charges were imposed, hypertension was less well controlled in this group, to the extent that the annual likelihood of death rose approximately 10%.[24]

  It has proved much more difficult to evaluate the consequences of user charges where there is no experimental design. However, when statistical analysis has been undertaken in other countries, it appears to corroborate the RAND results. For example, results from Belgium suggest a distinct impact of charges on demand for GP home visits and office visits, except amongst older or disabled patients.[25]


  In the light of the above discussion, the question arises: what is the most appropriate role for user charges in a modern health system? Experience in high income countries suggests a persistent tension between the equity goal of assuring universal access to health care and the efficiency goal of assuring frugal use of health services. In short, unless carefully designed, user charges designed to curb excessive demand amongst the bulk of the population could have ruinous financial or health consequences for a relatively small number of poor people with health problems. It is therefore important to view the design of user charges within the broader objectives and institutions of the health system as a whole.

  With the notable exception of the United States, there is a general consensus that public funding of tightly regulated delivery should lie at the core of the modern health system. However, there is also a growing trend in such systems towards the use of small but symbolically important user charges. Why this should be the case may be a matter for psychologists, sociologists and political scientists to explain, as—from an economic perspective—they appear insufficient to affect demand materially, except amongst the very poor, who are often exempt. Rather, the intention of new charging initiatives seems to be to influence very specific aspects of patient behaviour, and to act as a signal of preferred behaviour. In this respect, in conjunction with a system of carefully crafted exemptions, they may offer an important policy option for influencing demand. Moreover, the may help reassure the taxpayer that patients are being encouraged to use the services they pay for responsibly.

  However, beyond the largely symbolic nature of these recent developments, I believe that in the more medium term the accelerating pace of technological innovation and the inexorable rise in patient demands may require a more fundamental rethink of the role of charges. At present, European countries are (just about) able to ensure that most mainstream interventions are included in their statutory package, allowing policy makers to claim that coverage is comprehensive. However, there is growing evidence that such a policy may become financially unsustainable, and that policy may have to resort to increased use of explicit rationing of health care.[26]

  If it does, the central policy problem is to decide which health care technologies should be subsidised from public funds. User charges policy then flows naturally from the choice of the subsidised treatments. Once the "public" package of care is chosen, patients would still be free to purchase the remaining unsubsidised interventions at market prices, or to purchase complementary private insurance to cover such interventions. This is the essence of the Slovak experiment. I have shown elsewhere that, from an economic perspective, the choice of interventions in the public package should be guided solely by the expected health benefits they bestow in relation to costs.[27] Equity concerns should in my view be tackled not by the health care system, but by the tax system used to finance the public package. However, if political considerations demand that the package should be skewed in favour of diseases of the poor, then this does not affect the general principle of explicit definition of the package.

  The scope of the statutory package will be determined by the public's willingness to pay the necessary taxes—in particular, the willingness of the healthy and the rich to subsidise the sick and the poor.[28] It is therefore essential that the package is of high quality, so that richer people do not choose to use private care in preference to publicly subsidised care. If quality is poor, widespread resistance to paying the taxes required to finance the public package may arise, making the public system unsustainable.

  In England, the National Institute for Health and Clinical Excellence (NICE) is charged with evaluating new technologies, and issuing associated clinical guidelines. Therefore, although a daunting technical undertaking, NICE could in principle be given the expanded remit of recommending the entire scope of the publicly subsidised package. Charges (partial or total) would then be paid by patients on interventions that fell outside the chosen package. Indeed one could envisage that—if a technology fails its cost-effectiveness criterion—NICE could nevertheless determine the (lower) price at which the intervention or drug that could be included in the public package. The patient would then be asked to fund the difference between the NICE price and the provider's price.

  Whether charges are symbolic or substantive, the issue of exemptions has proved a problematic issue for policy makers. For example, successive UK governments have introduced exemptions for prescription charges on the grounds of age (young and old), health needs (an apparently arbitrary selection of conditions) and income, resulting in a very low proportion of patients being liable for charges. Clearly exemptions can often be arbitrary and pervert the intended economic signals. Yet equally, the evidence from RAND and other experiments is that at least some disadvantaged patients will suffer catastrophic financial or health effects if some system of abatement of charges is not put in place.

  The solution to exemptions adopted in many countries has been to set a maximum liability for health care in any one year, perhaps as a proportion of total income. The intention is to ensure citizens experience some of the incentive effects of user charges, but are protected from ruinous health care expenditure.

  In summary, therefore, I believe that the publicly funded health system of the future should look something like the following:

    —    an explicit set of interventions is subsidised by public funds (the "health basket"), the choice of which is guided by the criterion of cost-effectiveness;

    —    the size of the health basket is determined by the willingness of the population to pay the necessary taxes;

    —    there should be no compromise on the quality of publicly funded health care, the intention being that all citizens should use the public sector for interventions within the health basket;

    —    charges (partial or total) are paid by patients on interventions that are not deemed cost-effective;

    —    those able and willing may purchase voluntary (complementary) insurance to protect against such "economic" charges;

    —    there may in addition be small "symbolic" charges even on fully subsidised interventions, as signals of preferred behaviour—these cannot be insured in the private market, otherwise the incentive effect is lost; and

    —    there may need to be a carefully crafted system of exemptions from symbolic charges to protect very poor or very sick citizens, perhaps in the form of a maximum percentage of annual income; however exemptions will not apply to interventions that lie outside the statutory package.

  This system may at first glance appear unattractive compared to the stated principle of a comprehensive NHS, free at the point of access. Yet many commentators feel that it will be infeasible to adhere to that principle indefinitely, as the scope of health care increases inexorably and the limits to popular willingness to pay the necessary taxes are reached. If this is the case, the proposals set out here offer policy makers a framework for making the hard choices that follow in a systematic and fair fashion.

  Adoption of such a system may also help convince the public that taxes are being spent wisely, in line with both efficiency and fairness criteria, so maximising the chances of creating a high quality statutory package of broad scope. It will take political courage to implement such explicit rationing, but the alternative may be steadily to reduce the scope and quality of the NHS by stealth, and reduce the widespread support for tax funding of the NHS, an outcome that cannot be to the general public good.

Professor Peter C Smith

January 2006

12   Zweifel, P and Manning, W (2000), "Moral hazard and consumer incentives in health care", in J P Newhouse and A J Culyer (ed), Handbook of health economics, Amsterdam: Elsevier. Back

13   McGuire, T (2000), "Physician agency", in J P Newhouse and A J Culyer (ed), Handbook of health economics, Amsterdam: Elsevier. Back

14   Eversley, J (2001), "The History of NHS Charges", Contemporary British History, 15(2), 53-75. Back

15   House of Commons (2005), Hansard 27 January 2005 : Column 561W, London: The Stationery Office. Back

16   Robinson, R (2002), "User charges for health care", in E Mossialos, A Dixon, J Figueras and J Kutzin (ed), Funding health care: options for Europe, Buckingham: Open University Press. Back

17   Gericke, C, Wismar, M and Busse, R (2003), Cost-sharing in the German health care system. Discussion Paper, Berlin: Department of Health Care Management, Technische Universita­t Berlin. Grabka, M, Schreyögg, J and Busse, R (2005), Verhaltensa­nderung durch Einfu­hrung der Praxisgebu­hr und Ursachenforschung-eine empirische Analyse. DIW Discussion Paper 506, Berlin: Deutsches Institut fu­r Wirtschaftsforschung. Back

18   Kanavos, P and Reinhardt, U (2003), "Reference Pricing For Drugs: Is It Compatible With US Health Care?" Health Affairs, 22(3), 16-30. Back

19   Bellanger, M and Mossé, P (2005), "The search for the Holy Grail: combining decentralised planning and contracting mechanisms in the French health care system", Health Economics, forthcoming. Back

20   Andersen, R, Smedby, B and Vågeröc, D (2001), "Cost containment, solidarity and cautious experimentation: Swedish dilemmas", Social Science and Medicine, 52(8), 1195-1204. Back

21   Lewis, M (2002), "Informal health payments in central and eastern Europe and the former Soviet Union: issues, trends and policy implications", in E Mossialos, A Dixon, J Figueras and J Kutzin (ed), Funding health care: options for Europe, Buckingham: Open University Press. Back

22   Smith, P (2005), "User charges and priority setting in health care: balancing equity and efficiency", Journal of Health Economics, 24, 1018-1029. Back

23   Newhouse, J (1993), Free for all? Lessons from the RAND health insurance experiment, Cambridge MA: Harvard University Press. Back

24   Newhouse, J P (2004), "Consumer-Directed Health Plans And The RAND Health Insurance Experiment", Health Affairs, 23(6), 107-113. Back

25   Van De Voorde, C, Van Doorslaer, E and Schokkaert, E (2001), "Effects of cost sharing on physician utilization under favourable conditions for supplier-induced demand", Health Economics, 10(5), 457-471. Back

26   Coulter, A and Ham, C (2000), The Global Challenge of Health Care Rationing, Maidenhead: Open University Press. Back

27   Smith, P (2005), The statutory health care package under private health insurance. Paper presented to World Bank conference on Voluntary Health Insurance in Developing Countries, Wharton Business School, University of Pennsylvania Back

28   De Graeve, D and Van Ourti, T (2003), "The distributional impact of health financing in Europe: a review", The World Economy, 26(10), 1459-1479. Back

previous page contents next page

House of Commons home page Parliament home page House of Lords home page search page enquiries index

© Parliamentary copyright 2006
Prepared 18 July 2006