1In practice, health sector reform encompasses a range of strategies and policies, although a concern to improve efficiency, quality of care and equity are overarching objectives. A review by A. Cassels (1995) succinctly outlines these different strategies and policies: actions to improve the performance of the civil service; decentralisation, including working with the private sector; actions to improve the functioning of national ministries of health; universal delivery of a core set of essential services; broadening health financing options; and adopting sector wide approaches to aid rational planning.
2Once confined to the northern European region and Switzerland, decentralisation has become a central principle of health policy in countries as diverse as Portugal, Spain, the United Kingdom, Poland and France (Saltman and Bankauskaite 2006).
3Decentralisation is a transfer of resources, functions and political, fiscal and administrative authority from the centre to the periphery (Saltman and Bankauskaite 2006). It is a strategy that can be employed for achieving particular goals rather than an end in its own right (Peckman et al. 2005). In the health sector it is usually the administration and management functions that are handed over to locally based offices, with political authority remaining at the centre (deconcentration). Some countries have gone further and created sub-national levels of government that are substantially independent of the national level with respect to a defined set of political functions and legal status (devolution).
4Decentralisation is also promoted as a precursor to greater involvement of the private sector in the health system. The premise is that decentralisation frees managers to manage and will give rise to local bureaucracies that can better adapt to local needs and demands, leading to an improvement in efficiency, quality and utilisation (Kutzin 1995).
5Decentralisation has administrative (including “new public management” and other innovative approaches leading to managerially independent public firms, among others), fiscal and political dimensions (Saltman and Bankauskaite 2006).
6The Portuguese state has the constitutional obligation to ensure access to health care, with effective coverage of the country, and justice in the financial contributions. It has also to ensure the adequacy of entrepeneurial and private provision of health care and regulate the utilisation of health products.
Portugal has a health care service system characterised by three co-existing systems:
- the national health service (the Serviço Nacional de Saúde, SNS), publicly owned, centralised, universal and, at least in theory, providing comprehensive health care. The national health service is funded by direct and indirect taxes;
- special public and private compulsory insurance schemes for certain professions (health subsystems);
- voluntary private health insurance (Bentes et al. 2004).
■ Evolution of the portuguese health care system
8The Portuguese health care system took place over the past 60 years. It is possible to identify five principal periods (1945 – 1968; 1968 – 1974; 1974 – 1990; 1990 – 2001; 2002 – 2005).
Post – war period to late 1960s (1945 – 1968): the state acknowledges its responsibilities regarding health care
9Since the late 1920s, Portugal was under a one-party-system dictatorial regime with restriction of civil rights and censorship of the press.
10The state had responsibilities in the field of public health ( “sanitary police” functions). In curative health it assumed only a suppletive role, although not always giving an adequate response to curative care needs. The responsibility for assistance in illness lay with the families, private institutions or social welfare medical services (Barreto 1996) (union and state-supported social welfare started in Portugal in 1935, slowly increasing the number of benefits included and people covered and culminating with the institution of a state social security system after the 1974 revolution.
11After the war, it was recognised that the sanitary situation of Portugal was very poor and that private initiative was probably not enough to redress it. This led to the establishment or autonomisation of institutes responsible for vertical programmes aimed at improving maternal and child health, tuberculosis control, leprosy control, malaria eradication and treatment of psychiatric diseases (Sampaio 1981). At the same time, important hospital legislation was passed (1946) (regional organisation, type of hospitals, referral systems) supported by a programme of hospital building (many of the hospitals built by the state were handed over to charity institutions) (Lopes 1987). The state saw health care as hospital care and tried to extend it to the whole country, using, when necessary, “mobile health teams” as envisaged by the Law of 1946. Meanwhile, the trade unions organised their social welfare efforts into a federation (Federação das Caixas de Previdência in 1946), equalising benefits and medical services on a national scale (initially covering only the costs of ambulatory curative care, but later, from 1965 onwards, paying part of the hospitalisation costs of its beneficiaries) (Sampaio 1981).
12The year 1958 saw the foundation of the ministry of health and welfare, probably to calm the movement of criticism towards state’s neglect of the country’s sanitary situation, which had been launched by the opposition during the previous presidential election campaign. Between 1956 and 1961, medical doctors in the region of Lisbon, with the support of the Portuguese medical council (Ordem dos Médicos) developed an important document of analysis of the health situation in the country and defended the need for a national health service (Ferreira 1990; Campos 1983).
13Therefore, during this period, state intervention in the health sector grew but not to the level demanded. The colonial war effort (1961 – 1974)
progressively isolated Portugal from a world increasingly sympathetic to the cause of colonial independence and conditioned the resources available for non-military developments. Nevertheless, the war and the need for healthy soldiers might have been an incentive for some of the developments observed in the health sector.
Period from the late 1960s to the April Revolution (1968-1974): primary health care emerges
14In 1968, the political softening of the dictatorial regime under Marcelo Caetano permitted the emergence of change (Rosas 1998). At this time, health services were scattered, dependent on different national structures and ministries and not interconnected.
15The health system was made up of
- public health services (with a “sanitary police” function), – public institutes coordinating vertical programmes, – social welfare medical services (Caixas de Previdência, providing the bulk of ambulatory care),
- private hospitals (not-for-profit charities were responsible for the bulk of hospital care) and private consulting services, – pharmacies and diagnostic laboratories. There was no national health policy and no goals or even studies pertaining to the health situation and needs.
16In 1970 an office for health planning was created within the ministry of health. This structure still exists today, with some changes.
17In 1971, an important law on the reorganisation of the ministry of health services was approved. Health was, for the first time, formally recognised as a right guaranteed by the state. A concern with accessibility was also made explicit. It states that there must be a unitary health policy under the responsibility of the ministry of health and that all the health and welfare activities should be coordinated. Health planning and the choice of priorities were also defended. The law created health centres in a spirit of what was later formalised at Alma Ata (Almaty) (Declaration of Alma-Ata 1978) as “primary health care”. Despite its importance, this law was very difficult to implement because of the absence of political will to do so. It became impossible to connect or integrate the different systems that were dependent on different ministries or private organisations. The practical consequence of this law was that it added public sector-controlled heath centres without changing anything else in the health system.
18These health centres were chronically under-resourced and provided only preventive and promotive maternal and child health care, including vaccinations, as well as legally required medical examinations (e.g. to be accepted as civil servant or to work as a food-handler), but they did not provide curative care, which remained under the responsibility of the social welfare medical ambulatory services.
19This cleavage would only be overcome with the legal integration in the 1980s of the health services into the national health service, under the social security system.
The softening of the harsh political regime lasted only until 1972. The simmering war in the African colonies and a growing movement against a military solution to the conflict were crucial in triggering the Carnation Revolution.
Period from 1974 to 1990: establishment of the national health service
20The revolution in April 1974 marked the birth of the modern democratic system in Portugal. The wars in the Portuguese colonies stopped immediately and the colonies’ attained independence soon after.
21With this decolonisation process, many people living in the colonies were forced to come back to Portugal, including a great number of health professionals. This provided the extra resources needed to exand the coverage of the public sector health system.
22In 1975 a new law created a compulsory period of service in health centres for newly graduated doctors (Serviço Médico à Periferia). This further contributed to improve the care provided by health centres.
23A new constitution in 1976 reaffirmed health as a right of all citizens – “all are entitled to health protection and have the duty to defend it and promote it”. This entitlement is embodied through a comprehensive, universal national health service, free at the point of delivery (later, 1990, becoming “mostly free at the point of delivery”).
24The national health service law was passed in 1979. With it, the ambulatory health care provided by the social welfare system was integrated into national health service health centres. This met with enormous resistance and even today there are historical remnants of this social welfare system in the form of parallel compulsory health insurance schemes for which membership is based on professional or occupational category. These are often refered to as health “subsystems” (still covering today about 25% of the population) (Bentes et al. 2004).
25The social welfare medical services were managed by a central board and by 18 district delegations. These delegations were the basis for the establishment of 18 district-based coordination structures for the newly integrated health centres.
26The inability of the health centres to respond to the growing demand for health care brought about an inappropriate demand for primary care at hospital emergency services. The emphasis shifted then to trying to improve the hospitals’ response to this demand, failing to acknowledge the integrated logic of the different levels of the health system and placing on the health agenda the resilient problem of waiting lists. The ensuing decade saw investments in hospital construction, hospital health information systems, the first reforms on the financing of hospitals and efforts to define the responsibility of hospital doctors (Santos 1998).
This was a period of extraordinary progress in the field of health care, leading to increased access and equity. Towards the end of this period, soaring health budgets placed cost containment at the centre of the health agenda, where it has remained up to this today.
Period from 1990 to 2001: greater attention to system management
27In 1990 a new legal framework for the national health service was approved, in which health was no longer “free of charge” but “tending to be free of charge”, and described as a responsibility shared by the state, individuals and civil organisations. In practice, this led to an increase in the number of contracts between the national health service and the private sector, the introduction of private practice in public hospitals, the private management of public hospitals and fiscal incentives for the use of voluntary private health insurance.
28This law on the fundamental principles of health regionalised the management of the national health service to five regional health authorities, created by another law in 1993. These regional health authorities become accountable for the health status of their populations and responsible for the coordination of health care provision by hospitals and health centres, according to the allocation of financial resources (Bentes et al. 2004). Nevertheless, some aspects of human resources management and hospital financing, facets such as the choice of institutional directors remained centralised or directly dependent on the ministry of health. The 18 district health authorities also continued their uncomfortable coexistence with the newly established regional authorities.
29By 1995, of the four major political parties in Portugal only the Communists continued to defend the idea that the original model (public integrated model) of the national health service should be maintained and not reformed. The other three supported the necessity of evolving towards a contractual model, with greater separation between purchasers and providers of care (Santos 1998).
30This was reflected in the first draft of a national health stategy for public discussion in 1996. In 1998, the national health strategy and goals for the period 1998 – 2002 were approved and published (Ministério da Saúde 1999). It was the first time in Portugal that such a document was produced. In 1997, the regional contracting agencies were created, formalising the purchaser – provider split – one for each of the five regional health administrations. These agencies were expected to further decentralise resource allocation responsibility through the gradual implementation of contracting arrangements with hospitals and health centres (Bentes et al. 2004), but achieved limited success because of the lack of political support. At present their role and future is not clear.
31Some experiments with more flexible and autonomous forms of management were launched in a limited number of public hospitals. At this level, too, efforts were made to create intermediate levels of hospital management, establishing a mid-level management interface between the professionals and the hospital administration.
32In 1999 a law on the organisation of local health systems (equivalent to the World Health Organization’s health districts) was passed, which sought to promote better coordination of all the local structures involved in a broader concept of health care provision (involving different ministries and authorities: health, social security, education, municipalities, non-governmental organisations, etc.), but until now these initiatives had limited implementation. There is today a single “local health unit”, consisting of one hospital and four health centres with a common management board.
This was a period of experimenting with new forms of managing the health system (recognising the complementarity of central, regional, local and institutional levels of management), organising primary health services and hospital care, financing hospitals and health centres and remunerating health professionals. Some of these experiments led to the passing of laws, such as that of 1998 on peformance-related payment of family doctors, first applied to a limited number of physicians, but forming today the basis, after its evaluation, of a new primary care reform in 2006.
Period 2002–2005: reformulation of hospital management
33The year 2002 marked the beginning of a political cycle marked by intense legislative activity, most of which did not produce any significant changes.
34Nevertheless, this period saw the announcement of the construction of ten new hospitals under private/public partnerships (PPP) and a revolutionary transformation of 34 hospitals (out of 97) into 31 public enterprises, taking further than ever before the logic of autonomous hospital management. One of the consequences of these initiatives was the establishment in 2003 of a new independent authority for health regulation (Entidade Reguladora da Saúde) as a new entity (along with the department of health and the general health Iinspectorate) with specific regulatory functions.
35New hospital legislation further strengthened the central power, requiring that all hospital directors be directly nominated by the minister rather than by regional authorities. Equally, the annual plans of actions of the hospitals had to be submitted for ministerial approval, although it was possible for the minister to delegate this to the regional health authority. This greater centralisation countered the trend in recent years towards towards a more participatory management style, announced even in the constitution, which stipulates that “The management of the national health service is decentralised and participatory”.
36The year 2004 saw also the approval of the “National Health Plan 2004 – 2010”, representing a new formal health strategy for the national health system, calling for a more entrepreneurial type health management in health institutions, with a view to greater autonomy.
Since 2005: reformulation of primary care
37Current developments suggest we are entering a sixth period of reform, mostly concerned with the “reformulation of primary health care” (Biscaia 2006). Following on from previous initiatives, the latest reform aims to strengthen the autonomous management of health institutions, reactivate contracting agencies and local health systems, develop a strategy for human resources in the health field, strengthen the intervention of private players in the health system, harness the health budget deficit, give greater relevance to the regional administrations by eliminating the health districts, and increase patient choice.
Overview of the reforms of the Portuguese health care system
38In short, it is possible to identify in Portugal, in the course of the past 60 years, a number of periods of health care reform (Biscaia et al. 2006). The strategies and policies already referred to above show that the reforms’ objectives centred on the improvement of the performance of the civil service; the improvement of the functioning of the central level of the health system; the universal delivery of primary health care services and access to essential hospital services; the broadening of health financing options; cross-sector strategic planning; and decentralisation, including working with the private sector.
39A number of features are common to the different periods of health care reform. There were no abrupt changes. The state slowly assumed responsibilities in providing health care and financing it, eventually becoming the major funder, provider, regulator and purchaser of health care. The reforms were not usually preceded by formal evaluations; they were politically driven and very normative (based on legislative changes). Once legislated, only rarely were the reforms fully implemented. A number of values persisted throughout most of these periods – equity, universal coverage and solidarity in financing. The recently approved strategic plan ( “the National Health Plan 2004 – 2010”) considered as guiding values social justice, universality, equity, respect for human beings, solicitude and solidarity. The principles of sustainability and continuity, as well the as autonomy and humanisation of health care, are also highlighted in this plan.
■ Decentralisation in the reforms of the portuguese health care system
40The period before the 1974 revolution was a period of scattered health services with little central coordination. All district hospitals belonged to charities even if they relied heavily (95%) on state and municipal budgets (Campos 2004) or on welfare payments for services provided; public hospital ownership was limited to central and specialised hospitals, and to dispensaries and sanatoria for social diseases; ambulatory care was provided either by welfare medical services operated by salaried doctors and nurses or private practitioners; pharmacies, diagnostic imaging and laboratory services tended to be privately provided on a conventional fixedfee basis.
The major reforms emerged after the revolution. The initial period, from the mid-1970s to 1990, was a period that saw the development of a very centralised national health service. From the 1990s onwards, decentralisation concerns were reflected in a number of initiatives: the emergence of regional health authorities; internal health markets and contracting agencies; the purchaser – provider split; entrepreneurial management experiences; local services; vertical (local) integration; and regionalised networks of facilities. These decentralisation efforts have ensured that policy, including on fiscal and financial matters, remained centrally controlled, while service delivery was increasingly decentralised, and mostly of two types: devolution and deconcentration. Privatisation was also experimented with but led to recentralisation.
41Devolution, or political decentralisation, has been present in the political design of the country since the 1976 constitution (Campos 2004). The Azores and Madeira islands are constituted as two politically autonomous regions, with full authority for the development of the autonomous health services. The Constitution also anticipated the creation of politically autonomous regions on the continent, where the majority of the population lives, but the initiative was rejected by the Portuguese in a referendum in 1998.
42Nevertheless, recent laws contemplate the reinforcement of the functions of local authorities with regard to health care (management of health care centres and the commissioning, construction and maintenance of health facilities), but so far these have not been implemented.
43The establishment of the national health service resulted in a highly centralised administration and management supported by 18 district health authorities inherited from the previous social security system. The 1990 reform proclaimed a regionalised structure for the Portuguese national health service. Detailed legislation three years later created five regional health authorities with regional management teams appointed by the minister of health. These regions reflected the effort to deconcentrate administrative and managerial decisions from the central to the regional level and to recentralise management responsabilities from district to regional authorities. This has been a long process, in which, for example, responsibility for public hospitals and health centres – boards of management, budgets, staff appointments, investment decisions, technical orientations,
44annual plans and organisational rules – has been passed back and forth between the central administration and the regional health authorities. These regions have remained as a very controversial form of deconcentration: incomplete, with limited powers overlapping those of the central and district authorities.
45The organisation of the ministry of health remains highly centralised. It has more than 20 central-level organisational structures reporting directly to the minister or vice ministers.
46Parallel to the process of regional decentralisation, recent governments have focused their attention on health facilities autonomy. These have included the development of innovative tools for running health facilities, mostly hospitals, and the transformation of “mid and lower level public sector administrators into active managers who run their units on a more entrepreneurial basis” (Saltman and Bankauskaite 2006), a development in relation to their past inability to act beyond what was normatively or administratively required (Conceição et al. 2000; Gonçalves et al. 2000).
Hospitals have remained “publicly owned, publicly capitalised and publicly accountable” (Saltman and Bankauskaite 2006), but institutional managers have gained greater control over important operating levers, including the hiring of personnel and greater scope for the purchase of goods and services.
47As seen before, Portugal has a long tradition of private ownership, private funding (mostly from out-of-pocket contributions by the service users) and private delivery of health care services (Campos et al. 1987). A modern form of privatisation emerged in late 1990s, for three sets of reasons:
- ideological issues, in relation to the liberal trends of Thatcherism and Reaganism.
- aspects related to the growing perception of the state’s shortcomings as provider of health care (facilities larger than needed, low staff productivity, weak budgeting, slow technological progress, uneven quality, poor accessibility, fraud and corruption, and huge, recurrent deficits) (Campos 2004).
- and finally, the recognition that the frame of reference of the national health system goes beyond the boundaries of the national health service. These factors led to experiments with the contracting out, on an almost trial basis, of one large public hospital to private management during 1995; the conversion of 34 public hospitals into 31 publicly owned private firms incorporating nearly 50% of the country’s hospital beds during 2002; public – private partnerships for the design, building, financing and operation of facilities. Further, in order to shorten waiting lists for elective medical procedures, the establishment of special programmes enabling the outsourcing of elective surgery to the private sector (Atun et al. 2006). These processes are contributing to a “melting of public – private boundaries” (Saltman and Bankauskaite 2006) in the Portuguese health system.
48Attempts at increasing patients’ freedom of choice can be seen as the most radical form of privatisation of the demand function in health care. Gatekeeping by family physicians and referral networks covering different levels of hospital care, integrating local, regional and national resources organised in a logical hierarchy and differentiated on a technical level, limit consumer choice and thus represent a reversal of privatisation at the individual level (Vrangbæk 2006). This contradiction has still not been addressed in the Portuguese health care sector, but the privatisation of health care provison and funding is seen as means of increasing choice.
49Another example of privatisation of the financing functions is the growing reliance on patients’ contributions and voluntary health insurance (Vrangbæk 2006).
■ Discussion and conclusions
50It is important to recognise that the search for decentralised models for the Portuguese health care system has resulted from a number of pressures: the expectations generated by the constitution of 1976; Portugal’s entry into the European Community in 1986 and the expectation that European funds, including funds earmarked for health development, should be used in projects implemented on a regional basis; international influences associated with new public management models in the early 1990s and with Thatcherism and Reaganism, also in the 1990s; and the appearance, after 15 years of a centralised national health service, and despite its great success, of the significant managerial problems listed above (Campos 2004).
51These pressures for decentralisation were countered by a number of barriers to effective decentralisation. Decentralisation is not a favoured concept in a country that has existed since the 12th century and whose current borders were defined the 13th century. Decentralisation derives from the command and control model of public administration established in Portugal. Central government units of command and control (e.g. finance ministry units), are apprehensive of the possibility of runaway expenditure and the risk of unaccountability on the part of decentralised services (Campos 2004; Amaral 2005).
52The fears or reasons that prevented decentralisation from becoming a tool of responsiveness and accountability are a little-studied subject in Portugal. There is a need for empirical analyses that can highlight those reasons and also the expectations associated with decentralisation.
53As Portugal explores different forms of decentralisation, the regulatory role of the state will need development and strengthening. The recent creation (2003) of a regulatory entity has not yet led to many practical developments because it is an addition to a complex and uncoordinated network of other agencies with overlapping responsibilities in the regulation of different parts of the system.
54There has also been a lack of clarity regarding the “acceptable decentralised model”. The various players and pressure groups in the health sphere are centrally organised: unions, professional councils and associations, the pharmaceutical industry, private health insurance, the civil servants’ and health-sector business subsystems prefer to deal with a single central authority rather than several decentralised ones. There is a lack of clarity regarding the objectives of decentralisation – to many it is an end in itself, a constitutional requirement, politically driven rather than motivated by health or management objectives. Lastly, there is little evidence to date of the effects of decentralisation.
55As a result, all the experimentation with decentralisation has resulted in a situation where:
56– Decentralised units appear as add-ons rather than as substitutes to the existing central units of the Portuguese health care system. These central units essentially remain unreformed or even strengthened.
57– New models (e.g. the institutional autonomy model) of decentralisation overlap with incompletely implemented models (e.g. the regional health authorities model).
58– Despite all the efforts, there is no evidence of shifts in power to the decentralised units.
59– Privatisation is leading to recentralisation via increased state control.
60– There is considerable apprehension of the consequences of decentralisation at the central level, at the regional and health facility levels and among the various players in the field. Particularly, in a country where taxes are centralised, decentralisation of expenditure has led to significant resistance to these initiatives.
– There has been a great concern about “where” decentralisation should be orientated (on the geographical and institutional levels), but considerably less about “how”, “why” and “for whom” (professionals and patients).
In Portugal, the search for a decentralised model has involved an ongoing re-balancing of national and sub-national decision-making roles. It is not expected that the country will see a major decentralisation drive. The decentralisation process will continue, however. There is no other option for a centralised system like ours. But it will be a minimalist approach, slow and incremental. This will oblige us to clarify, step by step, the type of model we are looking for: a national health service that is a central service locally managed (the regional health authority model); or a national health service that is a local service operating within central guidelines (the institutional autonomy model) (Butler 1992). 
Associação para o Desenvolvimento e Cooperação Garcia de Orta” (Lisbon) and Health Systems Unit, Institute of Hygiene and Tropical Medicine, Universidade Nova of Lisbon (Portugal).
Part of this work was done in the context of a project, “The practice of health care reforms: lessons for the future”, INCO DC programme of DGXII, European Union, contract number ERB-IC18-CT98-0346.
The authors are grateful to Diane Lequet-Slama and Sylvie Cohu for their invitation to present this paper at the International Conference on the Reforms in Social Protection Systems in Continental and Southern Europe on 19 – 20 December 2005, in Paris. We are also grateful to a number of players in the health system who granted us interviews, in order to discuss their understandings and expectations of the processes of decentralisation in Portugal, namely Amílcar Carvalho, Francisco George, Pedro Nunes, Vitor Ramos, Cipriano Justo and Constantino Sakellarides.