%20-%20https://sketch.com%20--%3e%3ctitle%3e3ECBBF50-FD08-412A-943D-2CFC5E025A16%3c/title%3e%3cdesc%3eCreated%20with%20sketchtool.%3c/desc%3e%3cg%20id='Stylesheet'%20stroke='none'%20stroke-width='1'%20fill='none'%20fill-rule='evenodd'%3e%3cg%20transform='translate(-926.000000,%20-668.000000)'%20id='BUTTONS'%3e%3cg%20transform='translate(100.000000,%20597.000000)'%3e%3cg%20id='BTN-/-RIGHTS'%20transform='translate(816.000000,%2065.000000)'%3e%3cg%20id='ICON-/-Rights'%20transform='translate(10.000000,%206.000000)'%3e%3cpath%20d='M13.5752841,0%20L1.34943182,0%20C0.59375,0%200,0.586666667%200,1.33333333%20L0,22.6666667%20C0,23.4133333%200.620738636,24%201.34943182,24%20L17.6505682,24%20C18.40625,24%2019,23.3866667%2019,22.6666667%20L19,5.36%20L13.5752841,0%20Z%20M13.5752841,1.89333333%20L17.0838068,5.36%20L13.5752841,5.36%20L13.5752841,1.89333333%20Z%20M1.34943182,22.6666667%20L1.34943182,1.33333333%20L12.1988636,1.33333333%20L12.1988636,5.36%20C12.1988636,6.10666667%2012.7926136,6.69333333%2013.5482955,6.69333333%20L17.6235795,6.69333333%20L17.6505682,22.6666667%20L1.34943182,22.6666667%20Z'%20id='Shape'%20fill='%23595959'%20fill-rule='nonzero'%3e%3c/path%3e%3ccircle%20id='Oval'%20stroke='%23595959'%20stroke-width='1.5'%20fill='%23595959'%20cx='9.5'%20cy='13.5355339'%20r='5.5'%3e%3c/circle%3e%3cpolyline%20id='Path-2'%20stroke='%23FFFFFF'%20stroke-width='1.5'%20transform='translate(9.500000,%2013.035534)%20rotate(45.000000)%20translate(-9.500000,%20-13.035534)%20'%20points='8%2016.0355339%2011%2016.0355339%2011%2010.0355339'%3e%3c/polyline%3e%3c/g%3e%3c/g%3e%3c/g%3e%3c/g%3e%3c/g%3e%3c/svg%3e)
1. Introduction
In a normal competitive insurance market (e.g. car insurance), insurers practise experience rating and risk selection, which means that those at high risk either pay large premiums or are uninsured. In an unregulated health insurance market, annual premiums may range from less than €400 to more than €40,000 per person, which in most societies is unacceptable. In Europe the solution to this problem was to remove choice in both the Beveridge and Bismarck models. For countries with Beveridge systems multiple insurers were replaced by a state monopoly in a single-payer scheme and there was limited choice and no competition among providers. For countries with Bismarck models of multiple insurers there was neither choice nor competition among the different insurers, as rules of eligibility determined, which sickness fund or government scheme would apply to each individual; and although patients had free choice of physician, there was no price competition among providers.
In the 1970s, in the USA, the Federal government sought to develop the idea of Ellwood et al. (Reference Ellwood, Anderson, Billings, Carlson, Hoagberg and McClure1971) for competing integrated care organisations modelled on the Kaiser Permanente Health Maintenance Organisation (HMO) (Starr, Reference Starr1982: 393–398). Enthoven (Reference Enthoven1978) articulated this idea in his Consumer-Choice Health Plan as a way to provide universal coverage by regulated competition in the private sector. In the late 1980s, governments in the UK, Denmark, the Netherlands, Sweden and New Zealand launched proposals for major reforms in health care based on choice and markets. These reflected ideas of New Public Management (Hood, Reference Hood1995) and in particular two different models based on the ideas of Enthoven (Reference Enthoven1978 and Reference Enthoven1985). For the UK National Health Service (NHS), Enthoven suggested creating non-competing HMOs to which people would be assigned based on where they lived, which could thus avoid the problems that arise from experience rating and risk selection (Enthoven, Reference Enthoven1985). This emerged as the original model of an ‘internal market’, from 1991 to 1997, which transformed hierarchies into markets by separating ‘purchasers’ from ‘providers’; and enabled purchasers to contract selectively with competing public and private providers, with the objective of ‘money following the patient’ (Secretaries of State for Health, Wales, Northern Ireland and Scotland, 1989). In England, the idea of provider competition was suspended from 1997 (Secretary of State for Health, 1997) but reintroduced with an emphasis on patient choice from 2002 (Secretary of State for Health, 2002). In the Netherlands, the model of choice and competition among insurers was developed over 20 years (Ministry of Welfare, Health and Cultural Affairs, 1988), which also allows insurers to contract selectively with competing providers.
The reforms in England and the Netherlands created new transactors, responsible for insuring defined populations for their costs of health care, but the markets were designed so that these transactors would not practise experience rating: in England by virtue of these being monopolies, and in the Netherlands by the development of a risk equalisation scheme and mandatory community rating (with everyone paying the same premium for the same insurance product per insurer). In the rest of this paper, we describe these transactors generically as Mutual Healthcare Purchasers (MHPs), which includes the sickness funds and private health insurers in the Netherlands, and ‘purchasers’ and ‘commissioners’ in England.
The second and third sections of this paper outline the experiences and progress in implementing different models of choice and competition in England and the Netherlands, and discuss the obstacles to implementing these models successfully and the progress that has been made. The fourth section considers what each country might learn from each other, and explores a ‘thought experiment’, of how competition between MHPs might be introduced in England by learning from the Dutch experience. The final section makes concluding observations on the experience of different models of choice in the two countries.
This paper mainly considers technical, economical and political obstacles and preconditions to implementing provider and purchaser competition in England and the Netherlands. Analysis of how institutional context and governance logics shape changes in governance approaches is, for the most part, beyond the scope of this paper. Dixon et al. (Reference Dixon, Robertson and Bal2010) report a comparison of current policies in England and the Netherlands that aim to implement choice for patients over where and when they are seen by specialists in hospital outpatient clinics, which have been supported by electronic referral systems. Or et al. (Reference Or, Cases, Lisac, Vrangbæk, Winblad Spångberg and Bevan2010) contrast current policies to increase patient choice of providers in three countries with the Beveridge model (England, Denmark and Sweden) and to reduce this choice by ‘soft’ gatekeeping in two countries with the Bismarck model (France and Germany).
2. The purchaser/provider split in England: 21 years of fruitless endeavour?
2.1 Two models of an ‘internal market’
In 1989, prior to the introduction of the original model of an ‘internal market’ in England, the NHS was organised into a hierarchical structure with each organisation defined geographically. There were 14 Regional Health Authorities, which were responsible for Hospital and Community Health Services (HCHS) and primary care. Beneath the regions there were nearly 200 District Health Authorities (DHAs), responsible for meeting the needs of their populations and running the HCHS, and 90 Family Health Service Authorities, responsible for paying general practitioners (GPs) as independent contractors (Klein, Reference Klein2007: 66–72 and 98–100). Patients’ access to HCHS was via referral from GPs who acted as gatekeepers. The system was designed to contain costs of HCHS by capping the global budget, and achieve an equitable distribution of resources by funding DHAs using a weighted capitation formula (Bevan and Robinson, Reference Bevan and Robinson2005). This system, however, lacked financial incentives for efficiency by providers (Enthoven, Reference Enthoven1985). In the winter of 1988–1989, a perceived ‘financial crisis’ was produced by a combination of two policies: constraining ‘real’ growth in total spending on HCHS and redistributing resources between DHAs. The government’s eventual policy response to this ‘crisis’ was to introduce an ‘internal market’ (Secretaries of State for Health, Wales, Northern Ireland and Scotland, 1989; Bevan and Robinson, Reference Bevan and Robinson2005).
In the original model of the ‘internal market’ (1991–1997), DHAs became purchasers and hospitals independent ‘NHS trusts’. The separation of purchasing from provision meant that DHAs, as MHPs, would be funded equitably for their populations and be empowered to contract selectively with NHS and private hospitals, which could thus be funded for their performance. But DHAs mainly used ‘block’ contracts with NHS hospitals that did not vary directly with the numbers of patients treated, and hence created a new kind of global, incremental budgeting for hospital services (Day and Klein Reference Day and Klein1991; Glennerster et al., Reference Glennerster, Matsaganis, Owens and Hancock1994; Raftery et al., Reference Raftery, Robinson, Mulligan and Forrest1996; Croxson, Reference Croxson1999), although these did develop later into ‘sophisticated block contracts’ to incorporate ranges of volumes with related payments (Raftery et al., Reference Raftery, Robinson, Mulligan and Forrest1996; Croxson, Reference Croxson1999). This meant that ‘patients followed the money’: that is, patients’ choices were constrained by DHAs’ contracts. The exception to this was from the innovation with most potential where GPs opted for various forms of GP fundholding and were free to choose providers of diagnostic and elective care subject to managing expenditure within cash budgets. GP fundholders were typically small-scale MHPs based on general practices (Glennerster et al., Reference Glennerster, Matsaganis, Owens and Hancock1994; Audit Commission, 1996; Mays and Dixon, Reference Mays and Dixon1996; Mays et al., Reference Mays, Wyke, Malbon and Goodwin2001; Bevan and Robinson, Reference Bevan and Robinson2005).
Since 1991, there has been continuing organisational turmoil and policy turbulence inflicted on the NHS in England (Bevan, Reference Bevan2006; Klein, Reference Klein2007; Audit Commission and the Healthcare Commission, 2008). Following the 1997 election, the new Labour government rejected the idea of provider competition and developed new organisational forms to replace DHAs and GP fundholding (Secretary of State for Health, 1997). There are currently about 150 Primary Care Trusts (PCTs) acting as MHPs, which contract with providers of primary and secondary care (some run community health services). The Labour government announced in 2002 the current model of an ‘internal market’ that emphasises patient choice (Secretary of State for Health, 2002). Practice-Based Commissioning (PBC) is being developed within PCTs, with the objectives of encouraging GPs and other primary care professionals better to manage referrals and commission and redesign services (Audit Commission and the Healthcare Commission, 2008). PBC has echoes of the ideas of GP fundholding except that PBCs have only indicative budgets (unlike GP fundholders’ cash budgets).
The current ‘internal market’ introduced four innovations (Audit Commission and the Healthcare Commission, 2008):
• An emphasis on patient choice for elective care.
• A new reimbursement system, ‘Payment by Results’ (PbR), for elective hospital care based on standard tariffs by Health-Care Resource Groups, (the English modification of Diagnosis-Related Groups) to ensure that ‘money does follow the patient’ (Bevan and Robinson, Reference Bevan and Robinson2005).
• NHS Trusts can, subject to approval and scrutiny by a new regulator, Monitor, become NHS Foundation Trusts (FTs) and enjoy greater independence and flexibility.
• Independent Sector Treatment Centres (ISTCs), which provide diagnostic and elective services, were created with the objective of developing an independent sector market that delivers value for money.
2.2 Obstacles to an effective ‘internal market’
Experience in the English NHS (and other countries) shows that there are nine principal obstacles to the implementation of an effective ‘internal market’:
1. From hierarchy to market. One objective of the purchaser/provider split was to replace hierarchical with market arrangements between autonomous purchasers and providers. However, as all political accountability is vested in one individual in central government, namely the Secretary of State for Health (Tuohy, Reference Tuohy1999), this makes it difficult for MHPs to make hard choices or for hospitals to develop into self-governing independent bodies (Enthoven, Reference Enthoven2000).
2. Choice of provider. In many areas, there is limited choice of providers of elective care, and such choice is much less relevant for emergency care, disease prevention and chronic diseases (Ham, Reference Ham2008a). The government aimed to increase choice in the current ‘internal market’ through ISTCs (Audit Commission and Healthcare Commission, 2008).
3. Provider exit. A vital way in which normal markets deliver efficiency is by forcing failing providers to exit the market. However, it is typically politically impossible for Ministers to let the market determine closure of a whole hospital; or the risk of this happening by it being destabilised by loss of contracts; and even closure of a hospital department can be politically difficult (Tuohy, Reference Tuohy1999; Enthoven, Reference Enthoven2000; Ham, Reference Ham2007).
4. Provider incentives. In the original ‘internal market’, block contracts meant that there were weak provider incentives. Patient choice and PbR in the current ‘internal market’ generates financial incentives for a hospital to increase activity where its cost is below the tariff and this applies in particular to day surgery (Street and Maynard, Reference Street and Maynard2007).
5. Purchaser incentives. It has been argued that the Achilles’ heel of attempts to introduce selective contracting by MHPs in England is that, as MHPs do not compete, they lack incentives to do so. (Maynard, Reference Maynard1994; Enthoven, Reference Enthoven2000). The introduction of purchaser regulation on provider costs in the original ‘internal market’ (by the ‘Purchaser Efficiency Index’: Bevan and Robinson, Reference Bevan and Robinson2005) and ‘World Class Commissioning’ (Ham, Reference Ham2008b) alongside patient choice in the second may be seen as attempts to remedy the lack of incentives on MHPs.
6. Effective autonomous purchasing. As demand for health care is determined in a process of collegial decision-making by GPs and hospital doctors, this means that it is difficult for MHPs to be effective purchasers (Tuohy, Reference Tuohy1999; Audit Commission and Healthcare Commission, 2008; Ham, Reference Ham2008b), or for patients to drive change by consumer choice (Audit Commission and Healthcare Commission, 2008). Even in the USA, the evidence is that patients do not switch from poor to good hospitals (Marshall et al., Reference Marshall, Shekelle, Leatherman and Brook2000; Fung et al., Reference Fung, Lim, Mattke, Damberg and Shekelle2008). There have been various attempts to involve GPs in (non-competing) MHPs in the NHS (Audit Commission, 1996; Mays et al., Reference Mays, Wyke, Malbon and Goodwin2001; Wyke et al., Reference Wyke, Mays, Bevan, Goodwin, Malbon, McLeod and Street2003; Audit Commission and Healthcare Commission, 2008).
7. Managing MHPs’ financial risk. MHPs have to manage financial risk with each year’s cash allocation being fixed and determined with reference to a weighted capitation formula. In the original model of the ‘internal market’, this could be done by block contracts, but PbR and patient choice now limits MHPs’ scope to manage payments to providers (Bevan and Robinson, Reference Bevan and Robinson2005).
8. Information on prices. In the original ‘internal market’ there was scope for competition on price but no sound basis for comparing services of different hospitals (Enthoven, Reference Enthoven2000). Now, for 80% of hospital services for which there is in principle scope for competition, PbR has removed competition by price but provided a basis for comparing hospitals.
9. Information on quality. In the original ‘internal market’ there were no comparable good data on quality of care (Enthoven, Reference Enthoven2000). There is now a regulator of quality, the Care Quality Commission, which is responsible for publishing an annual Health Check that assesses the performance of MHPs and providers against national standards set by the government (Healthcare Commission, 2008). This information is not, however, designed to inform patient choice in markets, so there is still a lack of detailed timely information for patients as consumers of health care (Audit Commission and Healthcare Commission, 2008).
2.3 Evaluations of choice and provider competition in the English NHS
Brereton and Vasoodaven (Reference Brereton and Vasoodaven2010) have published a review of the literature on the original and current ‘internal markets’ in which they highlight two key evaluations: the systematic review of the original ‘internal market’ by Le Grand et al. (Reference Le Grand, Mays and Mulligan1998) and the current one by the Audit Commission and Healthcare Commission (2008). As they point out, few studies have sought to evaluate the cumulative effects of either market-based reforms, but there is an abundance of research on the effects of individual policies. Their findings on the original ‘internal market’ may be summarised as follows:
• DHAs lacked the ability effectively to purchase and influence providers.
• NHS Trusts increased productivity, reduced costs and post-surgical length of stay.
• GP Fundholding was associated with reductions in patient waiting times, but there was mixed evidence of its impacts on referral rates and prescribing costs, inconclusive evidence of its longer-term impact on costs, and concerns over inequities between patients from fundholding and non-fundholding practices.
Detailed studies of the original ‘internal market’ were handicapped by limitations of data. Data on prices were available for GP fundholders and for hospitals treating cases from DHAs with which they did not have a contract (Extra-Contractual Referrals – ECRs), but not for most of hospital care as this was for contracts with DHAs for which prices were not made public (Propper and Söderlund, Reference Propper and Söderlund1998). There were limited data on quality. Propper et al. (Reference Propper, Burgess and Green2004) sought to investigate the impact of competition on quality by using in-hospital deaths within 30 days of emergency admission with a myocardial infarction for patients (which they found had a small negative effect). There are two problems with the evaluations of hospital efficiency and productivity: these use data on episodes of care, which are vulnerable to inflation, and we lack a counterfactual. Le Grand’s verdict on the limited impacts of the first ‘internal market’, which remains valid in the light of later evidence, was that ‘The incentives were too weak and the constraints were too strong’ (Le Grand, Reference Le Grand1999: 33).
The principal findings of Brereton and Vasoodaven (Reference Brereton and Vasoodaven2010) for the literature on the current ‘internal market’ based on patient choice were that there is little evidence of improvements that can be attributed specifically to market-based reform, which has little support from patients and the public, and NHS staff have been de-motivated by the succession of shifts in policy. They recognise that it may be too soon to judge the impact of these reforms. The evaluation by the Audit Commission and Healthcare Commission (2008) identified four reasons for this: the development of FTs and patient choice were behind schedule, where patients were offered choice, they lacked detailed information (Dixon et al., Reference Dixon, Robertson and Bal2010 highlight problems of implementing the ‘Choose and Book’ system), the scale of ISTCs is limited, and the effectiveness of those working on commissioning in PCTs has been impaired by four reorganisations since 1997. The Audit Commission and Healthcare Commission found little hard evidence of systemic improvements in the development of effective commissioning by PCTs or PBC; from ISTCs or FTs; from PbR; or from the choice policy. They also identified concerns with three of the four innovations: freeing concerns from government control means that there is no system of governance to ensure supply across health economies; ISTCs were offered more costly contracts than NHS providers; and there was scepticism over the choice policy as an instrument to drive up quality (Bevan, Reference Bevan2008).
A separate evaluation by the Audit Commission (2008) of PbR concluded that this had ‘undoubtedly improved the fairness and transparency of the payment system’ (p. 2) but has ‘yet to have a significant impact on activity and efficiency’ (p. 5) (see also Farrar et al., 2007). The review of the literature on PbR by Brereton and Vasoodaven (Reference Brereton and Vasoodaven2010) identified problems of the incentives it created including a conflict with funding of PCTs, incentives for supplier-induced demand, questions over its impact on quality of care, and disincentives to treat severely ill patients.
The principal conclusion of the review by Brereton and Vasoodaven (Reference Brereton and Vasoodaven2010: 10) of the evidence on the impacts of both ‘internal markets’ is that ‘the reforms have not been proven to bring about the beneficial outcomes that classical economic theory predicts of markets’ such as provider responsiveness to patients and purchasers; large-scale reduction in costs; and innovation in service provision (the reasons being the obstacles we identified above). Their conclusion was that the available research on each form of the ‘internal market’ indicated that the NHS has incurred the transaction costs of seeking to introduce competitive markets without experiencing their benefits.
3. Consumer choice of MHP in the Netherlands
3.1 From top–down cost controls to managed competition
Historically, the Dutch health care system has been characterised by much private initiative both in funding and provision of care, but the mix of corporatist and etatist governance has allowed the government to intervene in various ways to control total costs. From the mid-1970s, the government increasingly regulated supply and price. In the 1980s, the open-ended hospital reimbursement system for inpatient care was replaced by a budgeting system, and the government controlled physicians’ fees and then their total revenues. However, these top–down policies were criticised for creating a system that lacked incentives for efficiency and innovation – similar to Enthoven’s criticisms of the NHS (Enthoven, Reference Enthoven1985). Hence there was broad support for incentive-based reforms and a reconsideration of the role of competition in health care. In March 1987, the government-appointed Dekker Committee recommended a pro-competitive health care reform, which after 12 months of intense national debates, the government accepted (Ministry of Welfare, Health and Cultural Affairs, 1988).
In the early 1990s, the Dutch government had effective systems for controlling prices, the number of physicians, hospital capacity and hospital budgets, which produced effective top–down control of total expenditure on health care: as a percentage of gross domestic product (GDP) this barely increased over 20 years: from 7.4% in 1980 to 8.0% in 2000 (OECD, 2008). An unintended consequence of this was unacceptably long waiting times for hospital care, which resulted in successful legal challenges to the European Court of Justice and Dutch courts. (Flood and Haugan, 2010, explore the issue of legal challenges to policies that seek to constrain choice.) Judgements in these courts required sickness funds to pay for patients to be treated more quickly, which led to an explosion in the costs of Dutch health care: expenditure on health care as a percentage of GDP increased from 8.0% in 2000 to 9.5% in 2004 (OECD, 2008). This provided the justification for urgent reforms. After a few years of intense political debate, the Health Insurance Act was enacted on 1 January 2006, which may be seen as the culmination of a series of pro-competitive reforms, gradually implemented from the early 1990s, and the eventual implementation of the principles of the Dekker report. This Act introduced a package of crucial major systemic reforms including choice among competing MHPs, so that now:
• It is mandatory for everyone to buy private health insurance that covers the standard benefits package as described (in terms of functions of care) in the Health Insurance Act.
• Subsidies make health insurance affordable for everyone.
• Individual consumers have an annual choice among MHPs and among different insurance arrangements.
• MHPs can decide to offer various insurance arrangements (such as a reduced annual premium for a voluntary higher deductible) and different packages of entitlements for different premiums (e.g. packages may differ in terms of the type of primary care provider seen, the setting for types of care, whether there is direct access to specialists or whether the GP acts as the gatekeeper, and arrangements for paying providers). Each package of entitlements must cover the standard benefits package as described in the Health Insurance Act. To protect consumers the Health Care Authority monitors whether the contractual conditions of each package of entitlements are in accordance with the Health Insurance Act.
• MHPs can decide to concentrate on specific population groups with entitlement packages tailored to their specific needs (such as diabetics and the elderly) and become skilled purchasers for their enrolled populations.
• MHPs are required to allow open enrolment and apply community rating.
• A risk equalisation fund distributes funds to the MHPs according to the risk profile of their members and thus ensures solidarity and deters ‘cream-skimming’.
• MHPs are allowed to contract selectively or integrate with health care providers.
• There is active and effective regulation to maintain competition.
3.2 Preconditions for effective managed competition in health care
Several complicated preconditions had to be fulfilled in order to reap the benefits of managed competition in health care: that is, combine competition with universal access and create the appropriate incentives for consumers, providers and MHPs. As few were fulfilled in 1987 (Schut and van de Ven, Reference Schut and van de Ven2005), the original timescale for such ‘radical’ reform was unrealistically optimistic. We now outline the preconditions, the extent to which they were fulfilled in the early 1990s, in 2010 and what still needs to be done.
3.2.1 Risk equalisation
If a government tries to tackle the symptom of high premiums for high-risk individuals by premium regulation and open enrolment, this results in ‘cream skimming’: that is, insurers seeking only young, healthy enrolees. This problem can be tackled effectively only by removing the incentives to do this, ideally by risk-adjusted equalisation payments to the MHPs (van de Ven and Schut, Reference van de Ven and Schut2007). The risk equalisation formula does not need to be ‘perfect’, but has to be good enough to deter ‘cream skimming’ by making MHPs’ perceived costs of risk selection (including its loss of good reputation) outweigh its benefits.
Before 1992, the sickness funds were non-competing non-risk-bearing administrative organisations. From 1993, the sickness funds became risk-bearing MHPs and each was free to set its own premium. This revolution had to be implemented gradually for two reasons: each MHP needed to build up financial reserves and the risk equalisation system was initially based on age and gender only, which left scope for ‘cream skimming’. Gradually, together with improvements in the risk equalisation formula (van de Ven et al., Reference van de Ven, van Vliet and Lamers2004; Prinsze and Van Vliet, Reference Prinsze and Van Vliet2007), the government increased MHPs’ financial risk from 3% in 1993 (with the equalisation fund primarily responsible for 97% of all profits or losses) to 75% in 2010. The Dutch government has announced further improvements in risk equalisation that will further reduce the incentives for risk selection and allow for a further increase in MHPs’ financial risk.
3.2.2 Market regulation
Market regulation ought ideally be by four authorities at arm’s length from government, of which only the first was in place in the early 1990s:
• A quality authority to protect consumers against substandard quality of care: the Netherlands Health Care Inspectorate (Inspectie voor de Gezondheidszorg).
• A solvency authority to require MHPs to have sufficient financial reserves to fulfill their financial obligations: from 1992, De Nederlandsche Bank (DNB).
• A competition authority to prevent cartels, anti-competitive mergers and abuse of dominant position (which were problems in the early 1990s, see Schut et al. (1991, Tables 1 and 2): from 1998, the Netherlands Competition Authority (NMa) covers all sectors. Ensuring competition in health care now accounts for about a third of this regulator’s time.Footnote 2
• A consumer protection authority to safeguard consumer interestsFootnote 3: from 2006, the Netherlands Health Care Authority (Nederlandse Zorgautoriteit).
3.2.3 Transparency
Effective competition requires transparency in medical products and insurance. In 1990, hospitals received a global budget and specialists were paid on a fee-for-service basis. Transparency was improved in 2003 through a system of combined payments for both the hospital and the specialist based on a classification system of Diagnosis Treatment Combinations (DTCs), and from 2006, the requirement that each insurer offers insurance arrangements that cover the standard benefits package as described in the Health Insurance Act (but voluntary supplementary insurance is not standardised). In 2011 the number of DTCs will be reduced from about 30,000 to 3000.
3.2.4 Consumer information
Consumers ought to be well informed about their right periodically to choose a health insurance option and their entitlements. There ought also to be relevant, reliable, objective and easily understandable information on the quality and services of providers and MHPs. In 1990, there was hardly any relevant information to enable consumers and MHPs to choose providers, and there was no consumer choice of sickness fund. From 2005, much more information is available on providers and MHPs (e.g. http://www.kiesbeter.nl), but there is still a lack of detailed timely accessible comparative information on quality of different types of care.
3.2.5 Freedom to contract
Insurers and providers ought to be free to contract selectively and to negotiate the content of contracts (e.g. prices, quality and services) and to integrate vertically. In 1990, sickness funds had to contract with any willing provider (hospitals and physicians), were not allowed to integrate vertically with providers, and prices were fully regulated by government. From 2006, MHPs are allowed to contract selectively or vertically integrate with providers, and prices are negotiable for range of services (e.g. the percentage of the hospital turnover for which the MHPs and hospitals can freely negotiate the DTC-price, increased from 10% in 2005, to over 30% in 2009).
3.2.6 Consumer choice of MHP
Each individual consumer ought periodically (e.g. annually) to have a free choice of MHP without facing high-transaction costs (search costs and filling out forms), and the premium ought to be independent of age, health status or any other risk factor. This requires MHPs to accept each applicant for each insurance arrangement covering the standard benefits of basic health insurance – ‘open enrolment’ – and not to reject high-risk applicants. In 1990, consumers had virtually neither choice of sickness fund nor of insurance product. From 2006, they have choice of both,Footnote 4 and MHPs are required to practice open enrollment.
3.2.7 Financial incentives for efficiency
There ought to be effective financial incentives for MHPs to seek efficiency in managing insurance and in the delivery of care through integration or contracting with competing providers, for providers to be efficient in delivering care, and for consumers to be price-sensitive purchasers of health insurance. In 1990, sickness funds did not bear any financial risk, specialists received a fee for each item of service, hospitals received a global budget, and consumers had no financial incentives for efficiency. In 2010, MHPs are risk bearing and pay a lump sum payment per episode of illness to the hospital for both hospital and physician services for each DTC, which generates incentives for efficiency. The system is designed to create incentives for consumers to be price-sensitive at the margin when purchasing health insurance and a mandatory deductible has been introduced.
3.2.8 Sufficient providers and MHPs
There ought to be no shortage of providers and MHPs, otherwise they will enjoy dominant positions in the market, which limits effective competition. In addition, the market ought to be contestable. This precondition was satisfied in 1990 and continues to be satisfied in 2009 for MHPs, and each Dutch hospital has at least one close substitute (Varkevisser et al., Reference Varkevisser, van der Geest and Schut2010).
3.3 Progress in implementing and assessment of the Dutch model
Although it is much too early for a full evaluation of the Dutch competitive model, some preliminary conclusions can be drawn from the first evaluation of the Health Insurance Act-2006 for the Dutch government. This evaluation was summarised as ‘on balance positive, despite some serious problems’ (van de Ven et al., Reference van de Ven, Schut and Hermans2009).
A key indicator of the success of the Health Insurance Act-2006 is that no political party or interest group has argued for a return to the former system with its distinction between sickness funds and private health insurers. The positive outcomes are: a good system of cross-subsidies (‘solidarity’); a standard benefits package available for everyone, without health-related premium; an annual choice of insurer/contract (major reasons for switching insurer included the size of the premium and the premium rebate in case of group insurance see van de Ven et al., Reference van de Ven, Schut and Hermans2009, Box 5.2); strong price competition among the health insurersFootnote 5; increasing information about price and quality of insurers and providers of care; increasing insurers’ activities in purchasing care (e.g. setting up their own pharmacies, owning primary care centres, experimenting with bonus payments for GPs, and a preferred provider plan with financial incentives to use preferred GPs, an internet pharmacy and 13 hospitals); and that the quality of care is now at the top of the agenda.
A major achievement of the reforms so far is that in choosing an MHP and a specific insurance arrangement each consumer receives the full savings of choosing an efficient insurance package. This provides the MHPs with a strong incentive for efficiency (which is absent when MHPs do not compete). So far, the strongest impact of the reform has been on the price of outpatient prescription drugs. In June 2008 a number of large MHPs each completed a tender among producers of generics, which resulted in price discounts of between 40% and 90%. These successful purchasing activities by MHPs are the more remarkable as, in the last decade, government made many unsuccessful attempts to lower the prices of these drugs.
However, the evaluation also found that not all preconditions are yet fulfilled and identified some major problems.Footnote 6
• Although selective contracting by MHPs is a major objective of the reforms, this has been slow to develop for four reasons. First, the lack of good indicators of quality of care makes it hard for the insurers to justify offering preferred provider plans. Second, the hospital payment system is still largely based on a global budget under government regulation. Third, other insurers also benefit from an insurer’s managed care activities, because providers prefer not to treat patients differently depending on their insurer. Fourth, the insurers have still a restricted financial risk for inpatient expenses.
• Although the Dutch risk equalisation system is one of the most sophisticated in the world, insurers still have substantial incentives for risk selection.
• In practice, the insurers’ ‘legal care duty’ is not sufficiently defined.
• The chronically ill and elderly are hindered from switching insurer because they fear that their new insurer will not accept them for the voluntary supplementary health insurance.
In the last 20 years, the emphasis of the reforms has been primarily on the health insurance market. The Dutch government realises that the reform process is still a work in progress and intends to consistently further implement the managed competition model. Over the coming years, we expect substantial progress to be made by the Dutch government by: further improvement of the risk equalisation system and DTC-classification system; increasing the number of DTCs for which the price is freely negotiable (the percentage of the hospital turnover for which the MHPs and hospitals can freely negotiate the DTC-price, will be increased from 34% in 2010 to 50% in 2011); new legislation to strengthen further the position and legal rights of the consumer in health care; improving the transparency of health insurance products and better information for consumers on the quality of care and providers per insurance product; giving hospitals full financial responsibility for capital investments (buildings and medical equipment), and reducing substantial ex-post cost-based reimbursements to MHPs. There is also scope for further integration between MHPs and the providers of care,Footnote 7 which could result in Enthoven’s model of competing HMOs.
4. Learning from each other?
So what can each country learn from each other? The story so far seems to be that the Netherlands has succeeded in the difficult implementation of a system of MHP competition, but that quite dramatic changes in the structure of MHPs and the degree of competition among them have, so far, had little effect on hospitals. This is largely because there is still substantial government regulation with respect to prices and budgets, and although GPs act as gatekeepers to specialist care, patients have free choice of physician (either GP or specialist). Hence the real challenge for MHPs in the Netherlands is to be able to channel their patients to preferred GPs and (via them, or via advice, or via financial incentives or via the insurance conditions) to preferred hospitals. MHPs, however, fear that if they were to do so now, they would be suspected of choosing hospitals on grounds of low cost and inferior quality, and, as a consequence, would lose enrollees to MHPs that did not restrict patient choice. Hence, although the Dutch system is now improving the information on hospital costs, this could only be used in selective contracting if MHPs can also use and hence produce good evidence that their selective contracting is guided by quality. Unfortunately, such detailed information on quality of care is not yet available.
The English NHS has made little progress in developing a system of hospital competition despite this having been the principal objective of two different reforms. There seem to be three principal reasons why hospital competition has failed to develop in England: problems of provider exit; collegial decision making by GPs and specialists; and a lack of detailed timely comparable information on quality of care in hospitals. We emphasised above that there are numerous reasons why provider competition has so far largely failed to develop in England, and so introducing competition among MHPs is unlikely to have much impact without overcoming the other obstacles. Furthermore, we now know that competition among MHPs in the Netherlands has not yet resulted in them being able to exploit the potential of selective contracting for hospitals. Schlesinger (Reference Schlesinger2010) highlights the complexities posed by policies seeking to introduce choice into systems of health care and Bevan et al. (Reference Bevan, Helderman and Wilsford2010) distinguish between choice being an end in itself (the evaluative reason) and as a means to enhancing the responsiveness of governments and markets to the needs and demands of their citizens and consumers (the effectiveness reason). We consider here the idea of introducing choice of MHP into the English NHS as a ‘thought experiment’ with three different sets of changes. The first is for the evaluative reason of simply allowing the principle of choice of MHP. The other two are for the effectiveness reason, introducing constrained MHP competition (so that this still looks very much like the English NHS now) and full MHP competition with the wide range of choices of the Dutch system. We consider here the extra requirements to implement these different models additional to those required for effective provider competition.
Simply allowing choice of MHP would require a different method of risk rating for populations: from using data on small areas (currently used in the formula funding of PCTs) to data on individuals. The Department of Health has commissioned the Nuffield Trust to develop such a method to provide guidance in the funding of PBC (Bevan, Reference Bevan2009).
Constrained MHP competition would allow MHPs to offer the same package of health care (regulated by NICE), free at the point of delivery (except for standard national charges). MHPs would define their catchment areas (in terms of clusters of postcodes) where they can guarantee a duty of care to the population of that area: access to both a local GP and the same standard complete package of community and hospital care. MHPs would be able to contract selectively with GPs and providers of community and hospital services, which would restrict patient choice of provider (these contracts could be based on nationally or locally negotiated rates of pay). One natural way of enabling there to be consumer choice of MHP would be to allow each PCT to offer to act as a MHP for its neighbouring populations. Enthoven (Reference Enthoven1985) suggested this as a way of introducing competition among DHAs. We agree with him, however, that offering such opportunities to bureaucratic organisations is unlikely to result in much change. Real competition would be more likely to emerge from two different innovations: allowing entrepreneurial GPs and private insurers to become MHPs. This would require a radical shift in the understanding of the nature of choice in England from choice of provider, GP or hospital, to choice of MHP. For England, this model suggests different modes of regulation in cities, where there is scope for developing effective competitive markets, and rural areas where there are local monopolies. In cities, regulators need to fight to develop and maintain competition; and then use results from that as benchmarks in regulating areas where there are natural monopolies. Regulators would need to protect consumers from the adverse effects of competition in cities and its absence in rural areas. If competition between MHPs were to develop, this would require the same gradual approach as in the Netherlands in phasing in the risk exposure of the MHPs and regulation of the solvency of MHPs.
Seeking to implement the Dutch model in England would mean that MHPs, in contracting selectively with GPs and providers of community and hospital services, are free to negotiate all relevant aspects of contracts including GPs and hospitals (e.g. PbR would be used to define products, but not to set fixed prices) with conditions on quality and consumer responsiveness. This would also open up opportunities for MHPs to offer different types of standardised insurance packages, but each MHP would be required to offer the same basic benefits package of standardised functions of care. These would entail radical changes to the way NHS providers are paid and would confront patients for the first time with having to make choices over what they pay for insurance and the type of coverage that this brings.
Of these three models of choice and competition among MHPs, the first, that simply allows choice of MHP, seems like a natural development of the NHS in England; the second, of constrained MHP competition, entails more radical changes, would be more difficult to implement, and be an essential step before implementing the model of full MHP competition as in the Netherlands. But it may be only through that type of full competition would improvement follow that would justify the transaction and political costs of these changes.
5. Conclusion
This paper has reviewed different models of introducing choice into health care based on ‘internal markets’ of provider choice in England and of choice among competing MHPs in the Netherlands, which in principle, are able to contract selectively. The implementation of these different models has, however, so far failed to generate a competitive market for hospitals in each country, for different reasons, but, in the Netherlands, there is now managed competition among MHPs. Our analyses of these changes suggest that those introducing competition into health care must give people confidence that competition will not make access more difficult as a consequence of hospital closures, or reduce costs by driving down the quality of care. The first seems to have undermined provider competition in England, and the second has meant that MHPs in the Netherlands are wary of contracting selectively or otherwise channelling their patients to preferred providers. Although there is a technical solution to the second problem, by generating good comparative information on quality of care, the first looks to be more intractable in England.
In the English system of majoritarian government (in contrast with coalition government, subsidiarity and corporatism in the Netherlands) it is easy for there to be a blitzkrieg in the formulation of radical policies (Wilsford, Reference Wilsford1994), but, Tuohy (Reference Tuohy1999) contrasted this with slow progress in implementation citing the observation by Shock (Reference Shock1994) that, ‘A blitzkrieg can certainly achieve conquest but it cannot ensure effective occupation’. In the blitzkrieg of policy formulation of a model of an ‘internal market’ there is a vision of its effective operation in England: bureaucrats contract to shape supply to meet needs of populations, GPs and patients choose providers on grounds of quality; competition means that the good providers thrive and grow, and the poor suffer and exit the market. However, the experience suggests that Ministers are disappointed with ‘internal markets’ as implemented, as they turn out to be unresponsive and bureaucratic; and also with alternatives, because they can never promise the potential benefits of the vision of an effective dynamic market. This may explain successive Ministers’ enthusiasms for ‘redisorganisations’ and fundamental system reform in England every three years or so (Light, Reference Light2008). The continuing problems after 21 years of various attempts at the model of purchaser/provider split raises the question as to whether that model is fundamentally flawed. The model has been abandoned in three countries in which it has been tried: Scotland (Woods, Reference Woods2001), Wales (Welsh Assembly Government, 2008) and New Zealand (Ashton et al., Reference Ashton, Mays and Devlin2005). There are variations in its application within Sweden counties (Harrison and Calltorp, Reference Harrison and Calltorp2000) and Italian Regions (France, personal communication; see also Carbone et al., Reference Carbone, Lecci and Liotta2008).
The Dutch health care system is now in transition from the old well-established system of government supply side regulation of price and capacity, towards managed competition (van de Ven and Schut, Reference van de Ven and Schut2008, Reference van de Ven and Schut2009). The revolutionary proposals of the Dekker-plan are elegant in principle, but the process of implementation has proved to be difficult and lengthy. Nevertheless, the Dutch experience looks much more controlled than the succession of radical shifts in policies in England. The nature of Dutch policy formulation meant that it has taken two decades for successive Dutch governments to work out how to implement the necessary changes, but this also means that implementation has been a process of evolutionary change. A clear conceptual model has, over time, won support by all involved in health care: the government, the private insurers, the hospitals, clinicians, the population and patients. This model is being implemented in a gradual process of learning by doing. The idea of insurer competition, as implemented in the Netherlands, is being developed in Germany and Switzerland (van de Ven et al., Reference van de Ven, Beck, Van de Voorde, Wasem and Zmora2007). However, after 20 years the Dutch reform is still work in progress, because the proof of the pudding is in effective provider competition, which is still in its infancy. As long as there is no sound evidence that competing MHPs effectively control costs, the Dutch government may not give up its ‘classic’ tools for cost containment, even if they interfere with the preconditions of effective competition.
Tuohy (Reference Tuohy1999) identified Ministerial accountability in England as a fundamental axiom of the logic of the NHS and a serious obstacle to creating an effective ‘internal market’. Unlike in the Netherlands, the government in England remains responsible for the many activities of insuring, providing and regulating health care. This results in bureaucratic conflicts between multiple goals, which can only be resolved by Ministers, who are rowing in a 24/7 media storm where the NHS is a reliable and constant source of bad news. In contrast, the Dutch system of subsidiarity means that the multiplicity of activities involved in insuring, providing and regulating health care are now dispersed between various different agencies each with a clear goal. The Dutch Minister of Health is better placed to reply to questions in parliament on problems in health care that responsibility for its governance lies outside the Ministry, and can in future, aspire to focusing on steering this complex system to achieve its public goals of affordable, accessible and good quality care.
We have identified a number of obstacles to the development of an effective ‘internal market’ in England. We have also considered three kinds of developments: allowing choice of MHPs, introducing constrained MHP competition with no other radical changes to the NHS; and implementing the fully competitive Dutch model. The lesson from the Netherlands for countries interested in doing so is that this requires a number of technically and politically complicated preconditions to be fulfilled for this policy to be effective. These preconditions include a good risk equalisation system, an effective competition policy, an adequate system of product classification and medical pricing and transparent consumer information on the health insurance products and on the quality of health care providers. We end by asking two questions. First, does implementation of the Dutch model require governance that is a mix of being corporatist and etatist and based on the principle of subsidiarity? Second, in the early 1990s, if the different models of competition were seen to be in a race to successful implementation, then the English ‘internal market’ model would have been the hare, and the fully competitive Dutch model, the tortoise. Therefore, will the verdict of history be that, as in the fable, the tortoise won?
