Why is Sweden rethinking its NHS style reforms?BMJ 1997; 315 doi: http://dx.doi.org/10.1136/bmj.315.7113.935 (Published 11 October 1997) Cite this as: BMJ 1997;315:935
- Margaret Whitehead (), visiting fellowa,
- Rolf Å Gustafsson, associate professora,
- Finn Diderichsen, professora
- Correspondence to: Professor M Whitehead The Old School Ash Magna Whitchurch Shropshire SY13 4DR
- Accepted 6 March 1997
During the 1990s Sweden has embarked on a series of changes to health policy heavily influenced by the British NHS reforms. These have included the separation of purchasers and providers, an internal market regulated by contracts, competitive tendering, and the encouragement of the private sector. We discuss the origins and main features of the Swedish reforms, examine the subsequent developments which have led to a rethink, and consider what other countries can learn from this experience. To help answer these questions we interviewed senior politicians and policy makers in Stockholm County, as well as analysing official policy documents and carrying out empirical analysis of activity and cost data.
The run-up to the reforms
Why was reform considered necessary? One of the triggers was growing discontent among influential sections of the population, partly rooted in the effects of cost control during the 1980s.1 Through deliberate policies the proportion of gross domestic product spent on health care in Sweden fell from 9.6% in 1983 to 8.8% in 1991—Sweden and Ireland were the only countries in the Organisation for Economic Cooperation and Development to achieve any reduction at all over this period.2 A further decrease to 7.6% in 1993 was largely the result of shifting care of elderly people from counties to municipalities.
The 1990s reforms of the British NHS stimulated similar market-style changes in Sweden
In the largest county, Stockholm, problems started to surface almost immediately, requiring a series of modifications
By 1996 a rethink was evident; as a result, there was a swing towards cooperation and away from competition
Financial factors behind the rethink included a loss of cost control and problems with productivity incentives
Concerns about maintaining equity and trust within the system were also decisive in the rethink
Although this stabilised the healthcare budget, there were negative effects. As relative priority in healthcare was given to elderly and chronically sick people, waiting lists for elective surgery and access to primary care worsened for younger people and some services seemed unresponsive to their demands.
There were also shifts in some attitudes, again mainly in the middle class. Although national surveys still showed broad support for universal programmes, feelings about means tested benefits restricted to a minority and public bureaucracy were increasingly negative.3 Ideas about introducing market forces and competition into public services in general started to gain ground and the British NHS reforms came along as a possible solution to the dilemma. Great attention was therefore paid to the white paper Working for Patients when it was published4 and to British experts.5 6
At the end of the 1980s, the governing Social Democrats saw market reforms as a way of making the system more responsive and efficient and thereby dealing with some of the expressed criticism. The Conservative-Liberal government, which came into power in 1991, embraced these ideas even more warmly.
The main responsibility for funding and organising the Swedish healthcare system lies with the county councils, regional self governing bodies with elected politicians. The 26 counties organise the provision of services as they wish as long as the overall management is judged to be in line with national principles and guidelines.
Many councils introduced at least some aspects of market-style reforms such as purchaser-provider splits and performance related payment systems.7 Only in Stockholm County, however, did the reforms go as far as a managed market system introducing competition between providers, and consequently this model has received considerable attention in the public debate throughout Sweden.7 8 We analyse the experience of Stockholm in the discussion that follows. Table 1 illustrates the main developments.
In 1989 Stockholm County was decentralised into nine semiautonomous district health authorities, each with a local political board. By January 1992, these authorities—still directly managing primary care, geriatrics, and psychiatry—were transformed into purchasers with responsibility for medical care, public health, and health promotion for their population. A new weighted capitation formula was devised to allocate resources from the county council to the districts. The purchasers were required to establish contracts with providers in which the volume and quality of care were to be specified.
The one private and nine public hospitals in the county were assigned as providers. Only research and education resources were guaranteed; the rest of their income had to come from contracts for services to be negotiated with purchasers in competition with other providers.9
Performance related reimbursement
As a financial incentive designed to stimulate productivity, a performance related reimbursement system based on diagnostic related groups was introduced for hospitals. This started in 1992 with five surgical specialties and was extended incrementally to include more.10 At the same time, providers were no longer allowed to obtain services and rent facilities free from other public services but had to buy these from other providers at cost price. This was done to encourage cost consciousness and so that private providers were not at a disadvantage.
Encouraging private providers
Several measures aimed to stimulate competition among providers. In particular, regulations were introduced to make it easier for private providers to compete for public contracts. Firstly, a system of competitive tendering began in 1993 in which all nine authorities were required to go out to tender for any contracts worth at least 20% of their total budget. Secondly, the government introduced what proved to be a highly controversial measure to give patients greater direct access to private providers outside hospitals. From January 1994, registered physicians and physiotherapists could establish a private practice and be paid on a fee for service basis by the county council without negotiating contracts with the purchasers. In effect, purchasers had a passive role as payers for private services and payments were deducted from the health authority's contracted services budget. The government concurrently increased the fees that could be charged by private practitioners.11 This must be interpreted as a publicly subsidised build up of a private market for health care.12
Other relevant policies
Several other policy changes during the 1990s interacted with the internal market reforms. These included the “freedom of choice” policy that the Swedish Federation of County Councils adopted in 1991.13 Stockholm County had already given patients free choice of primary care centres in 1990 before implementing an internal market system. At the same time patients were also given direct access to all acute hospital care in the county without referral from a primary care “gate keeper.” This policy was integrated with the market model during 1992 when it was decided that “the money should follow the patient.” In practice only marginal changes in patient flows occurred,14 but the result was an inbuilt potential conflict between consumer choice and a purchaser's ability to foresee and control flows of patients.
Major community care reforms (the Ädel reforms) in 1992 transferred responsibility of continuing care of elderly people from the county councils to the municipalities in order to relieve the blocking of acute hospital beds by elderly patients who no longer required such intensive care.15 The reforms included a system of fines for community care providers who were not able to accept patients back from acute hospitals.
In 1991 extra funds from national sources were used to shorten waiting lists for elective surgery, and by 1992 a guarantee of a maximum waiting time was introduced by the government. This specified a maximum wait of three months for 12 elective procedures.16 17
Implementation and retreat
At implementation in 1992 productivity improved almost immediately as elective surgery increased and waiting lists fell. However, most purchasers perceived that costs were increasing, particularly as providers overshot contract volumes.18 19 The observed effects were attributed to the strong financial incentives, such as the performance related diagnostic related group system, coupled with the weak position of purchasers.19 By early 1993 purchasers had to introduce penalties for providers who exceeded the contract volume stipulated (purchaser's discount), but plans to extend the diagnostic related group system to more specialties still went ahead.
In 1993 and 1994 the original aim of stimulating activity was replaced by freezing the level of activity and spending less. Quantity related ceilings were therefore put on payment levels and diagnostic related group system prices were lowered. Success in reducing activity and improving productivity was limited,20 but there was an important side effect. Providers were confronted with uncertainty because of fluctuating prices and rules as well as competitive tendering. This fostered mistrust in the provider-purchaser relationship. The county council auditors in their 1995 annual report concluded that the changes “have drastically diminished the trust in the system.” 20
The fear of loss of cost control was heightened in 1994 when the freedom to establish private specialists came into force. There was concern about the way the measure was consolidating the concentration of specialists in more prosperous areas and transferring more public spending from the public to the private sector. In 1995 the Stockholm county auditors concluded that the increased fee levels for these specialists, rather than an appreciable increase in their numbers, had caused the increased costs.11 21
During this period an influential book on markets for health care was published and was widely debated.12 It warned of the risks of privatisation, the potential irreversibility of these changes, and the threat this posed to the fundamental principles of equity on which the Swedish healthcare system was founded.12
By April 1995 a task force of eight leading Stockholm County officials was recommending major changes. They summed up their interpretation of the effects: “We experience a current lack of trust in the Stockholm model …. The trust has probably been further reduced due to the financial deficit of 1994. This unexpectedly large deficit has both external and internal causes. The external factors are related to consequences of the raised level of pay to private specialists, the free right to establish private practice and the family doctor legislation. Among the internal factors we find the problems related to lack of total cost control and deficiencies in administrative systems for monitoring.” 22 The task force recommended abandoning the diagnostic related group reimbursement system in favour of block budgets and moving towards cooperation rather than competition.
We have tried to make an accurate assessment of the trends in total activity and costs compared with the perceived trends by analysing data from Stockholm County Council for 1983-95 (see 2). The evidence suggests that the measurable rise in activity rates began sooner than was commonly perceived and that it was related to other developments such as the maximum waiting time guarantee that were not directly part of the market oriented reforms.16 17 Overall costs did not rise as perceived, but those for specific parts of the service did—for example, in relation to the private specialist development and for some of the secondary care associated with the diagnostic related group system.11 23
What the figure also shows, however, is the dramatic decline in the council's income. It was the gap between expenditure and income which grew at an alarming rate and engendered a widespread feeling of loss of financial control. The decrease in revenue from income tax which began in 1991 but accelerated in 1994-5 was a result of the economic crisis in Sweden and the Ädel reforms. Unemployment in Stockholm rose 10-fold (from 1% to 10%) between 1991 and 1992 and affected council revenue after a lag period. In addition to this the government ordered the councils to freeze regional income tax.
The current rethink
A general hesitation and reconsideration is evident in recent political decisions both at the national and county council levels (1). By 1995 the unemployment situation in Stockholm was so grave that local politicians decided to give all county council employees an employment guarantee. This gave the health sector the onerous task of reducing costs without reducing staff.
Also in 1995, the government (now the Social Democrats) withdrew the free right of establishment for private specialists. No more establishments were allowed, although those already in practice could continue on less favourable conditions.24
In spring 1996 Stockholm County Council took several political decisions aimed at tighter political control rather than market control. For example, in January 1996 it set up a “hospital board” to oversee the provision of services in all county hospitals. This board reports directly to the central political board for health care, which now coordinates all purchasing of hospital acute care over the nine district purchasers.25 Purchasing now also has to take account of the long term structural plan for services in the county which was drawn up by a new council committee.26 These changes represent a considerable blurring of the boundary between purchaser and provider and an emphasis on setting up mechanisms for cooperation and priority setting, moving decisively away from competition.
Policy implications of reforms
Simultaneous policy changes at national and regional level with conflicting objectives caused reform overload and hampered systematic evaluation
Components of the market reforms were ill suited to addressing the underlying problems
Loss of cost control, erosion of trust, and concern about privatisation were major consequences of the reform process
Reforms should not be based solely on technical analysis of efficiency and equity in health care in isolation from wider concerns in society
What can be learnt?
This experience holds valuable lessons not only for Sweden but elsewhere. Firstly, the technical aspects of evaluating the reforms proved far more complex than was at first appreciated, as many policy changes were happening at the same time and were interacting. For example, surgical rates went up and waiting lists went down around the time the reforms began in Stockholm—but closer inspection shows that these trends started before the reforms and occurred in other county councils, though they were stronger in Stockholm.27 It is therefore not clear how much of these changes can be attributed to specific details of the reform package. The experience also emphasises the breadth of the impact. As well as medical and economic effects, questions need to be asked about the effect of reforms on the ethos of the health services—on the development of trust and motivation, for example.28 29
Secondly, some of the components of the reform package were ill suited to addressing the underlying problems. In Stockholm features such as fee for service payments in secondary care and capitation in primary care seemed the wrong incentives when cost control and structural changes in supply became political priorities in the 1990s. These incentives were designed to solve “yesterday's” problems of decreasing productivity and access. But when the reforms were implemented the underlying problems—decreasing tax revenues and rising unemployment in society—were completely different and the reform solutions were counterproductive.
Thirdly, the experience illustrates the need for balance between short term and long term objectives. Some short term political objectives of the reforms were to achieve gains in productivity and efficiency so as to reduce specific areas of discontent expressed by influential sections of the population. But a fundamental long term objective of the healthcare system is to preserve equity and trust, to ensure access for people in greatest need, not just those with the most influence. To do this methods are needed to control costs and set priorities, and these must be both efficient and equitable to prevent strong pressure mounting for privatisation. This is the old “trade off” between efficiency and equity, and the latest reform experiment failed to resolve it.
Fourthly, the experience shows that reform should not be based solely on a technical analysis of efficiency and equity in health care in isolation from the wider concerns in society. For instance, questions have to be asked about the long term role of the healthcare system in sustaining the welfare state, politically and economically. In Stockholm, for example, the county council hesitated to make unemployment worse by making health workers redundant to cut costs. Longstanding equal opportunity objectives in Swedish society might also be threatened by cutting the workforce, as a strong health and social care sector releases women from informal care at home and allows them to participate in the workforce. In these circumstances the healthcare system can be seen as part of the welfare state which has a role to play in national strategies such as maintaining high levels of employment.
Like Sweden, other European countries such as the Netherlands are now signalling a move away from competitive strategies, underlining the importance of taking stock of what has been gained and what has been lost by market style reforms in health care.
We thank the people we interviewed for the insights they gave into policy making and for providing access to data.
Conflict of interest: None.