Roemer 20 Years Later: When a Classical Health-System Typology Meets Market-Oriented Reforms

In 1990, Roemer came up with a very influential health system typology. From his vast study, emerged three types of health care systems: nationalized, mandated and entrepreneurial. Health care systems are not static; slow changes and reforms somewhat alter values and goals on which those systems were initially established. It is fair to say, then, that over the last two decades, health care reformers have adopted a market-oriented governance model that blends new public management (NPM) and managed competition reforms in the provision of health care services to transform supply- and demand-side actors into “responsibilized” customers, payers or providers. These transformations beg the question as to whether we are witnessing a radical redefinition of health care systems through the implementation of market-oriented governance. We propose to add the evolution of market-oriented health reforms in five case studies to Milton Roemer’s typology of health systems. In light of our findings, we will wrap up the analysis with an assessment of the usefulness of Roemer’s classification for social scientists to grasp the evolution of health systems over the past 20 years, and more importantly, to analyze the current state of these health care systems after years of market-oriented reforms.


Introduction
"Improving our health care system only works if everybody does their part."(Office of the Press Secretary 2009) Barack Obama, when pronouncing these words in September 2009, was announcing to the world that the basic tenet of the only truly entrepreneurial health care system -that is, providing health care mainly as a private benefit -was set to be slightly modified if and only if Americans were to make one outstanding effort: viewing health care as a right that the poorest as well as the wealthiest are entitled to.
Interestingly, leaving aside the American context, Obama's assertion could have been heard in a plethora of different health systems as it embodies the core concept underlining reforms in health care systems across advanced industrialized countries since the 1980s: the transfer of responsibility from an overwhelmed state to non-governmental actors in order to reach certain outcomes (quality, efficiency, costcontainment) with respect to health reforms.One only has to think about Quebec's latest budget (released in March 2010), which strives to "responsibilize" health care system users by implementing an annual contribution of $200 for every taxpayer; the rationale is that citizens must do their part in the face of the dire state of Quebec's public finances.
It is fair to say, then, that over the last two decades, health care reformers have adopted a marketoriented governance model that blends new public management (NPM) and managed competition reforms in the provision of health care services to transform supply-and demand-side actors into "responsibilized" customers, payers or providers (Aucoin 1990;Hood 1995;Pollitt 2001).It seems to us that these transformations beg the question as to whether we are witnessing a radical redefinition of health care systems through the implementation of market-oriented governance.
In order to shed light on this issue, we will first provide an overview of the dimensions of market-oriented governance in health care systems that are regarded as common to most health systems: managed competition, regulated markets and internal markets.Second, our task will be to provide a basic case against which the market-oriented health reforms of the 1990-2010 period will be assessed.Thus, we propose to add the evolution of market-oriented health reforms in five case studies to Milton Roemer's typology of health systems.In light of our findings, we will wrap up the analysis with an assessment of the usefulness of Roemer's classification for social scientists to grasp the evolution of health systems over the past 20 years, and more importantly, to analyze the current state of these health care systems after years of market-oriented reforms.

The "incentive and competition" paradigm and market oriented governance
To grasp the salience of market-oriented reforms in health care over the past 20 years, it is important to ground our theoretical arguments i n an evolutionary perspective of health care in the western world.
Cutler outlines three distinct reform eras in the evolution of health care systems over the past century or so: first, what he terms as "universal coverage and equal access" covers the period spanning from the end of the 19 th century to the post-war era, when most countries established universal health care systems (or programs) to reduce inequalities and improve quality of life.Then, as the Trente Glorieuses or "Glorious Thirties" vanished, this made way for an era of cost containment in the 1980s, reforms aimed at implementing controls, rationing and expenditure caps.According to Cutler, this led to overall dissatisfaction towards health systems and generated a great deal of inequalities and systemic inefficiency; hence, the introduction of a third wave of reforms, "incentives and competition," which is the paradigm advanced industrialized countries have dealt with since the early 1990s (Cutler 2002).
The idea of incentive-based reform has best been described by Inglehart who posits that reforms performed under such guidelines are "[…] based on a belief that economic incentives are the principal determinant of how patients, payers, and providers behave when they seek, finance, or render medical care" (Inglehart 1993(Inglehart : 1210)).From this perspective, if the state remains the thrust for reforms, a reorientation of management practices that change the behavior of supply-side actors and demand-side patients/customers ensures effective and enduring transformations.
In fact, Inglehart and Cutler's concepts are both pointing to market-oriented governance, which is interpreted in this paper as the application of different mixes of policy instruments like managed competition or other new public management (NPM) principles, and which leads to the creation of internal or regulated markets as we will demonstrate below.
Managed competition remains a core concept at the heart of health care governance in the era of incentives and competition.The concept was aptly coined by Enthoven who sought to describe the direction that the US health care system was taking in the 1980s.He defined "managed competition" as public management carried out with stringent rules for competition which reward health plans (or public agencies) that "do the best job of improving quality, cutting cost, and satisfying patients" (Enthoven 1993).Its goal is to divide providers into different economic entities competing against each other to attract customers.To this end, the state introduces incentives to various actors of the health care system guided by specific goals defined by governments.Such a governance model is now found in every health system, including American residual public programs that borrow on "managed care" practices, and Canadian and European systems in which managed competition is clearly gaining ground (Albreht 2009).Simonet provides an account of n ew public management (NPM) reforms in several European countries over the past decades and shows that marketoriented governance remains high on the reform list of policy-makers as of the early 1990s for most countries (Freeman & Moran 2000;Simonet 2008;Albreht 2009).
Maarsee agrees with this statement, yet emphasizes that Europeans largely oppose any serious violation of risk and income solidarity in funding (Maarsee 2006); this explains the incremental evolution of privatization in Europe over the past two decades.Indeed, Maarsee contends that market competition in health care is thriving in Europe and is expected to grow despite the slow-paced privatization occurring on the continent.Therefore, it seems clear that advanced industrialized countries fully entered the paradigm of incentive and competition despite opposing values and internal social dynamics (Aucoin 1990;Pollitt 2001;Maarsee 2006).
The inevitable question flowing out of such a sweeping public governance ideology is whether there is convergence; in other words, are health care systems poised to resemble one another on the governance dimension across the board?To be fair, those who attempted to study and compare contemporary reforms in ideal types of health care systems are confronted with roughly similar findings.There is a move from distinctive features demarcating each system to mixed-types and a convergence with respect to the use of market-oriented governance (Blank and Burau 2004;Rothgang et al. 2005;Wendt 2009).This theory is appealing yet it assumes away the fact that even though market-oriented governance can be (and has to be) considered a common denominator when comparing health care systems, it can possibly alter those systems in different ways (Hood 1995;Pollitt 2001).Pollitt's categorization of convergence shows the extent to which one has to take into account context and outcomes when studying managed competition in health care through the comparative lens; indeed, following Pollitt's reasoning, market-oriented governance exhibits a certain degree of convergence at the levels of discourses as noted in the introduction, and of practices such as managed competition or performance-related budgeting.However, it is much harder to achieve convergence across the board with respect to results since on the one hand, governments enact market-oriented reforms according to different objectives and, on the other hand, unintended results occur and are influenced in part by national contexts (Pollitt 2001).
With the above caveats in mind, our contention is that market-oriented governance is a concept that is put into practice everywhere, but that it is embedded into at least two types of reform strategies; first, in some systems, waves of market-oriented reforms lead to internal markets whereby agencies are created at the service-provision level and are accountable to the centralized bureaucratic structure.Reforms occur through such managerial principles as decentralization, delegation and deregulation (Aucoin 1990;Rothgang et al. 2005).Managed competition, in such systems, appears when those agencies are given incentives to compete with one another to attract patients/customers and/or to earn extra funding.Second, the other type of reform strategies leads to the creation of semi-regulated markets which provide governments with sufficient powers to introduce incentives for efficiency and quality in a primarily private market of health insurers and providers.These changes are necessarily modifying health care systems' building blocks whose main characteristics can be traced back by investigating basic health system typologies.

Typologies
Roemer's typology is one of the main basic health care system classifications in the field of comparative health systems, which was developed in the early 1990s.The reader should keep in mind, at this point, that Roemer's typology is used in this study because it is representative of the main characteristics of health systems up until the early 1990s, characteristics that may have been modified to a certain extent by marketoriented governance in the era of incentives and competition.
Over the past 20 years, the literature on health systems generated increasingly sophisticated typologies.A prominent approach in comparative health care identifies supranational historical and cultural factors that combine to generate today' social welfare systems.If Wilensky (1975) and Korpi (1980) were both instrumental in the identification of ideal welfare system worlds, it is especially with Esping-Andersen's " Three Worlds of Welfare Capitalism" (1990) that social system typologies mushroomed in the 1990s.Indeed, the exciting research agenda spawned by Esping-Andersen's influential research constitutes the basis on which a major strand of comparative health care literature built up hypotheses (Wendt et al. 2009).Most studies emulating this framework have been oriented towards the state as the main driver of health care reforms (Bambra 2005;Moran 2002;Wendt 2009;Wendt et al. 2009); indeed, these approaches denote the importance of state intervention relative to other actors on one or several dimensions of health care systems.If this highlights the evolving roles of actors, it does not provide sufficient clues as to how generalized policy changes and management practices affect the structure of these systems over time.In addition, the development of new dimensions of health systems is useful to uncover main trends, but tends to downplay the importance of the changes at the level of governance in health systems over the past 20 years.

Back to basics: Roemer's typology
In order to assess the extent to which the era of incentive and competition in health care governance changed health systems, this paper takes us back to a simple yet very influential typology of health systems: Roemer's categories of national health systems.Roemer came up with his typology at the end of the 1980s and as he acknowledges himself, merely takes into account the paradigm of incentives and competition that appeared around the same time (Roemer 1993); as such, this typology represents an opportunity for researchers to work from a blank slate to uncover the effect of market-oriented reforms on post-war health care systems.
Roemer defines health systems as the result of economic developments, history and dominant ideologies in countries where they are created.He distinguishes three ideal types of systems that are of interest for this analysis: welfare-oriented/mandated, entrepreneurial and comprehensive types.Comprehensive systems are financed through progressive taxes.Health care coverage takes the form of a universal health insurance that covers all citizens or residents of a country according to their needs.Because of funding arrangements and goals, these systems tend to be administered centrally by the health ministry or department of the state.Private insurance still exists in these systems, although it is usually supplementary and covers only care that is not provided under nationalized insurance.
The structure of Welfare-oriented/mandated systems boils down to the imposition of a mandated insurance to workers.As Table 1 below displays, the insurance covers a part of health care costs while the balance is generally paid through complementary insurance (often called "mutual" in European states) contracted through employment.Governance mechanisms of such systems are dependent on the fact that workers, employers and the state are financing health insurance funds.Lastly, because these funds are historically associated with professions, the management of mandated systems has traditionally been decentralized (Roemer 1993).
The third type of health care system-Entrepreneurial-mostly relies on the prevalence of the private sector to provide health care insurance and delivery, as depicted in the last row of

Massachusetts-USA
main providers of health insurance in the private sector.In such a system, the market is not the sole player; indeed, residual public programs for very poor, at-risk or elderly populations are to be found.As a result, the administration of these health care systems can be somewhat comprehensive; not only do residual public programs have to be publicly administered, but regulations pertaining to the private sector are usually widely present.The United States is the model par excellence of an entrepreneurial system; indeed, even the regulation of the private sector in this country is difficult.
Thus, as pointed out above, Roemer's typology reflects the evolution of health care system from the golden age of the welfare state up until the early 1990s.This typology hardly takes into account important reforms of the post-industrial era, such as market-oriented ones in health systems.Consequently, it is fair to question its actuality and the extent to which market-oriented governance had any effect on Roemer's categories of health systems-namely, comprehensive, welfare-oriented, and entrepreneurial types of systems.
Case-studies: Great-Britain,1 France, the Netherlands, Quebec and Massachusetts In this paper, the evolution of market-oriented reforms is observed through five health systems that are initially classified in each of Roemer's categories as Table 1 shows above.The health-system approach, as opposed to a country-centered one, is justified by the relevance of analyzing the actual locus of reformmaking; for most states, the health-system level is equivalent to national level, but this is not the case everywhere.Too many studies take the United States and Canada as a whole to explain health system reforms , in fact, the federal structure of these two countries engenders as many health systems as there are jurisdictions; consequently, we believe that Canada and the United States should not be solely studied at the federal level since reforms occur within sub-national entities.Therefore, two of our five cases are found in North America: Massachusetts and Quebec.In Europe, we look at the Netherlands, France and England.These cases were selected for at least three reasons: first, these health systems are qualitatively very different from one another, which gives us the opportunity to assess the impact of market-oriented governance in different contexts; second, extensive background research on these cases was previously conducted by one of the authors (Bourque 2007;2008); and thirdly, all of these health systems recently underwent major reforms which magnify the sweeping governance changes under study.
The act of comparing provincial health care systems with national systems in unitary states is questionable.Provinces or states within a federation fall under the federal umbrella of decisions that supersede them, and which have an important impact on their margins for operation.For example, in Canada, the Canada Health Act of 1984 defines the possible options that provincial governments have in matters of health care delivery.In the same sense, the financial transfers of central governments control the provincial governments in their choice of policy instruments.
Nonetheless, the fact remains that the governments of the states or provinces find themselves facing a range of possibilities in regards to the principles that they want to impose within their reforms.Examples include the regressive health tax of $200 per taxpayer, announced in 2010 by the Government of Quebec and phased in since that year, and the Ontario Health Premium, a progressive tax levied on the salaries of taxpayers adopted in 2004 by the Ontario provincial government.In the same sense, the reform of the Massachusetts health care system was planned and implemented under the federal government of George Bush and not Obama.Despite the disadvantages that this entails, we reiterate the relevance and importance of comparing health care systems based on the distribution of powers within federations as in unitary states.
A typical way of assessing welfare-state changes over time is to investigate the evolution of public expenditures in each case.However, market-oriented reforms in health systems are not easy to uncover with a cursory look at our case studies' broad public spending patterns.This can be attributed to the fact that the impact of market-oriented governance goes beyond financial matters and generally modifies the incentives of actors of health care systems; for example, following a market-oriented reform, an actor may spend a similar amount of money directed at the same institution as prior to the reform yet, the decision-making process behind this financial flux can be subjected to different constraints and pressures.
Therefore, as our analytical approach focuses on policy changes, Cutler provides three different types of incentive-based reforms that have to be highlighted for such a study and that are detectable through an analysis of policy reforms: incentive-based reforms at the patient level, or the implementation of cost-sharing schemes; incentive-based reforms at the insurer's level, which involves the introduction of competition between health insurance entities to enhance efficiency and quality of care; and finally, incentive-based reform within provider communities, which includes a " two-step" allocation of public monies from 1) the state to decentralized agencies or fundholders, and 2) from decentralized agencies or fundholders to hospitals, as well as rewards attached to sound financial results at the fundholder/decentralized agency level (Cutler 2002).Thus, this analytical framework reinforces our contention that market-oriented governance should be found in all of our cases and lead to the implementation of internal/regulated markets over the period spanning from 1990 to 2010.We can now turn to our case studies.
Comprehensive/Nationalized System: Quebec The province of Quebec, which is a sub-national unit in the Canadian federal system, can be classified as a comprehensive/nationalized health system.Health care reforms, in the past 20 years, were mostly conducted under right-to-center governments (Gaumer 2008).Quebec's health system is subjected to the Canada Health Act adopted in 1984.This law stipulates that federal health funding is conditional on whether provinces provide access to free health care irrespective of where citizens live and of their needs.Quebec's nationalized health care system started to be put into question in the mid-1980s with the Rochon Commission (1985).However, it is only in 1997 that the party in power, the Parti Québécois, was able to carry out an ambulatory shift reform as a natural continuation of retrenchment policies enacted in the mid 1990s (Charbonneau 2001;Gaumer 2008).The main objective of such a sweeping reform was to devise a plan to achieve an optimal utilization of health care resources.Evidently, such an endeavor entailed the deinstitutionalization of health care (by closing hospital beds) and put forward modern initiatives such as home care to prevent hospitalization (Charbonneau 2000).
In 2001, Quebec's government, still led by the Parti Québécois, commissioned a report on health care reform, this time presided by Michel Clair.The report suggested that the government should adopt a greater focus on improving frontline health services and on finding ways to increase the number of family doctors in Quebec (Denis 2001).However, the report's recommendations were relegated to the back burner due to an unfavorable political environment at the time.
Quebecers had to wait for a change in government to witness the sharpest turn towards marketoriented governance and new public management (NPM) of their health care system since the ambulatory shift in 1997.Indeed, the newly elected Liberal Party of Quebec, tabled Bill 25 in 2003.This reform, championed by the health minister, Philippe Couillard, was in line with this government's holistic approach to state modernization set out by Premier Charest, whose rationale for such a plan was the need to adapt the Quebec model of governance to such challenges as population aging and globalization.The impact of this strategy was not negligible for Quebec's health care system.First, the government sought to implement a population-based approach to allocating funding in different areas according to local needs.Therefore, one pillar of the Quebec health system, geographical equity in access, was altered without a profound modification of the features pertaining to the nationalized system-that is universality, provision of care irrespective of needs and financing through taxes.Second, institutional transformations did take place in the organization of the health care system, which affected the centralized management principle of the system (Bourque 2007).Importantly, from this point on, the institutions were modified in such a way that the implementation of some elements of an internal market were to be facilitated in the following years.
In Quebec, one of the main elements contained in market-oriented reforms is decentralization.This is best exemplified by the creation of local health agencies known as Agences de santé et de services sociaux (ASSS), whose mission is to reorganize the system and ensure service delivery based on the needs of the local population.As prescribed by new public management (NPM) practices, the Quebec government created new organizations in order to offload responsibility onto lower-level public servants (Noreau 2008).As a result, the centralized management principle proper to comprehensive/nationalized health care systems was altered.These decentralized agencies make the population-based approach work; they provide socio-demographic and economic data as well as service utilization statistics to facilitate the population-based allocation of the health care budget across the province's regions (Noreau 2008).
The second important aspect of market-oriented reform in Quebec is network integration; as a matter of fact, the creation of centres de santé et services sociaux (CSSS), which are merely mergers of many health service centers (CLSC, CHSLD and CH), aimed at creating health care zones and local networks of health and social services.This is along the lines of the population-based approach, since each sector overseen by CSSS contains a smaller population for whom need-based evaluation is performed on a regular basis.Interestingly, each CSSS must conduct a clinical project whose goal is to improve continuity of care for patients (Levine 2005).One could advance that as a result of these clinical projects, CSSS are in a situation of competition with one another; indeed, achieving good results (and outperforming them) as set out by the government has become a norm with which health care managers comply.
From a patient's point of view, the latest budget in Quebec constitutes a major change-and break-from the past since the inception of a comprehensive health system in 1972.In March 2010, Quebec's Finance Minister Raymond Bachand announced the introduction of two types of user fees in the health care system.The first one is a premium or consultation fee that would act as a gatekeeper for hospital visits; it is, however, unclear at this point how that would be implemented.The second user fee is an annual contribution of $200 per adult towards a centralized health fund regardless of individual income.The fund will serve as a tool to enhance competition between hospitals since a supplementary budget will be allocated according to the number of cases treated in health establishments.This is, without a doubt, one step further in the implementation of managed competition since 2003.Thus, the creation of public agencies by the centralized administration and their subsequent increasing responsibilities in the allocation of funding and health services are evidence of the presence-and growth-of an internal market in the province of Quebec.
Before turning to the British case, it is important to mention recent developments with respect to the opening-up of Quebec's health care system to private interests.In 2006, the government passed Bill 33, which significantly altered health care governance in two ways; first, the bill aimed at increasing publicprivate partnerships by creating norms regarding waiting lists.Second, it left the system's door wide open for private insurance in specific health care sectors.2Only time will tell whether future changes in policy will pave the way for further marketization of health care in Quebec.

Great Britain (England)
Our second comprehensive/nationalized case, the English National Health System, underwent major market-oriented reforms as early as in the 1980s under Thatcher's Tories (Rivett 1998;Pollitt 2001).At the end of the 1990s, a second set of important health reforms was launched by Blair's New Labour party and surprisingly reinforced the market-oriented approach adopted by previous governments.In 1984, Thatcher, relying on new public management (NPM) principles, implemented the g eneral m anagement f unction, which aimed at reducing staff and enforcing result-based management in health facilities.Less than a decade later, the National Health Service and Community Act was an early plan for the creation of an internal market, which entailed the shift from public authority acting as health managers to public authority acting as service buyers (otherwise known as GP fundholding) (Rivett 1998;Robinson an d Dixon 1999).Competition and population-based approaches in allocating public monies were therefore supported in Great Britain much earlier than anywhere else in Europe.In order to make this work, 100 regional health authorities were then created and granted sufficient power to oversee these health regions.
As soon as he became Great Britain's Prime Minister in 1997, Tony Blair went forward with an ambitious reform agenda entitled " The New NHS: Modern and Dependable" (Government of Great Britain 1997).This first reform propelled by Blair supported, among other things, better cooperation within the many components of the NHS, emphasized service quality, and put forward a plan to review and change the organization of the NHS in order to foster better integration of services (Government of Great Britain 1997).These modifications were to be followed by the replacement of GP fundholdings with primary care groups.In 2000, Blair's government undertook another reform, the NHS plan, which was set to facilitate an increased involvement of the private sector in the public NHS.Indeed, hospitals were to be managed by private firms, the stated goal being to reduce the system's costs and improve efficiency.In practice, then, the government used public-private partnership principles by granting private managers exclusive contracts to run health facilities (Oliver 2005).In 2001, more changes took place in the system along the same lines as preceding ones; for example, salaries became tied to performance and on top of that, evaluation of performance and human resource management started to emulate private-market practices.Up to this point, it can be noted that Great-Britain's pattern of health system transformations started with the creation of decentralized agencies followed by measures to stimulate competition and, in the end, the development of a sophisticated internal market.The latter phenomenon is actually even more obvious with the following reform, in 2006.
The 2006 reform was mostly centered on the idea of patient-customer and supported the principle of benchmarking, which stresses that services are necessarily improved if put in competition and systematically compared one another.The government insisted on the fact that the NHS reform aimed at building up a health care system tailored to the needs of the patients (Department of Health, UK 2010).The goal, as in Quebec, was to improve performance and efficiency.In both cases, then, taking patient needs and preferences into account seems to be the path followed by the latest reforms; in fact, since January 2006, Great Britain's health care system pushed the envelope further by allowing customers to select health establishments based on performance indicators released online.This constitutes, undoubtedly, a clear example of the commodification of health care in this country (Department of Health, UK 2010).
The patient-centered approach to health care governance as seen in Great Britain was crafted so that incentives and users/providers' behavior would resemble those of the private sector.Patients now possess information regarding many aspects of health facilities such as parking availability, access to public phones, quality of care, the management firm operating the establishment and a rating of the facility by previous patients.For example, in the south-west region, Derriford's hospital seems to display less than perfect results: if service provision is neighbouring the average score, resources don't seem to be used efficiently.Moreover, one learns that the hospital regularly cancels operations, has long waiting lists and is considered the less convivial establishment in the area (Directgov 2010).The dissemination of such details obviously leads to important consequences for one hospital; however, in the British case, consequences are magnified since public funding of the hospital is tied to performance.Thus, not only are patients encouraged to pick a local hospital based on performance indicators, but budgets are granted as a function of the number of patients treated in each facility.Consequently, funding goes where patients go, or as the government puts it, "best providers attract more patients" and are rewarded with bigger budgets (Directgov 2010).This particular aspect of health system financing clearly represents, as in Quebec, the principle of benchmarking.
The introduction of new public management (NMP) principles is clearly part of the health reforms in Great Britain and Quebec.It is important to mention, however, that the scope of market-oriented governance differs in both cases: while a sophisticated and all-encompassing internal market was created in England, Quebec's turn to market-oriented governance is still in its infancy when compared to the former case.That said, in the name of efficiency, both countries have adopted result-based governance, implemented accountability within and between decentralized public or semi-public agencies, and developed performance evaluations of health care.Thus, it is not surprising to see the latest reforms promoting and supporting best practices in health care delivery as updated information and quality indicators are made public.As a result, evaluation practices are fundamentally transformed as they are not the sole prerogative of the bureaucracy typical of comprehensive systems anymore; indeed, accountability and responsibility are now central values in those health care systems.

Welfare-oriented/Mandated Systems: France
The French health care system is founded upon the characteristics of the welfare-oriented/mandated system.As in most other health care systems, reforms from the "incentives and competition" era started in the 1990s under a recurring agenda of cost-containment.Prior to 1995, the government was in charge of public hospitals, while the important private sector was largely standing on its own; as a matter of fact, one of the reasons explaining the slower pace of health reforms in this country is that the French mandated system, until 1995, was very decentralized so, the central government in Paris did not initially have a strong hold of the different components of the system to initiate structural reforms.
Things began to change in 1995.The Juppé plan centralized (public and private) hospital management and other elements of the system, thus turning away from the core principles of mandated systems (Sandier et al. 2004).As a result, many new policy elements could be implemented: in 1999, 26 hospitalization regional agencies (ARH) were created in order to organize public and private hospitals, and to manage the allocation of public financing according to local needs and negotiation processes (Bellanger and Mossé 2005).Following this, the French government tabled the Social Security Funding Act, whose objective was to establish a cap on health expenditures controlled by the state and to enforce national priorities in terms of health care goals, a true novelty at that time.
In 2000, the French government implemented universal i llness coverage (CMU) which aimed at providing social assistance recipients with national public health insurance, thereby replacing the old decentralized administrative structure (formerly held by the d épartements).Here again, the underpinning principles of reform-namely, universal coverage and centralized management-was clearly borrowing on those found in comprehensive/nationalized health care systems.After this period of centralization deemed necessary to move forward with reforms, the government, through the "plan hospital 2007", introduced in 2003 and 2004 competition measures targeting health care facilities.Among other things, the goal of this reform was to finance hospitals upon the number of cases treated rather than through global budgets (Direction de l'information légale et administrative 2010).Patients were also targeted by reforms as the introduction of " gate-keeping" measures in 2004 attests.Two elements in the latter reform are especially important for our analysis: first, patients became monitored in order to control number of visits and second, access to specialists was restricted as general practitioners became the first point of contact for patients wishing to get full reimbursement of care (Departement de l'information légale et administrative 2009).
The sequence of reforms spanning from the mid-1990s to the mid-2000s denotes a pattern worthy of attention: generally, reforms up until 2003 led to a greater hold on the national health care system for the state, followed by an incremental decentralization of provision management and, in parallel, by the creation of market-type incentives to cut expenditures and improve efficiency in 2003-2004; for example, health care professional groups contracting with health insurance companies were then created.Therefore, once Paris came to restructure the system so as to have more control over its main entities, decentralization strikingly similar to our comprehensive/nationalized cases occurred (Sandier et al. 2004).
Finally, in 2009, the government moved forward in market-oriented governance when it revealed its intentions regarding the financing of the newly decentralized health regions: as in Quebec and Great Britain, a population-based approach in funding was crafted as part of an overarching plan to curb public expenditures in the face of a context of global crisis.Thus, the bill on "Hospital reform and patients, health and territories" was tabled to tackle the so-called modernization of hospital management.Transformations include a further move towards a population-based approach and a closer cooperation between the private and public sectors (Direction de l'information légale et administrative 2010).In addition, a simplified and consolidated governance structure is being built-up on the previous trials of the past years; this time, regional health agencies are given more autonomy since they coordinate the bulk of health resources in each territory.This constitutes another example of power devolution to newly-created local managerial levels and hence, a reformulation of governance principles with respect to mandated health care systems.In final analysis, this denotes the will of French governments to implement an internal market in the health care system.

Netherlands
If one assesses the evolution of health care systems through the "incentives and competition" lens, the Netherlands is certainly a case in point.The Dutch system as it stood in 1990 was, as Roemer outlined, part of the welfare-oriented/mandated family.Until 2006, one could depict the contemporary Dutch health insurance system as comprised of three different compartments: national health insurance for exceptional expenses or long-term care (100 percent population covered), sickness funds or private insurance as main coverage for curative care (53 percent of the population covered) and supplementary health insurance on a voluntary basis (3 percent only) (Den Exter 2004).A Dutch citizen was expected, at a very minimum, to get health insurance through sickness funds and assistance from the state when the cost of care far exceeded income.
The contemporary Dutch scheme was developed during the Second World War when the Netherlands was still occupied by Germany.Already then, regional health funds played a central role in health insurance and perpetuated two defining characteristics of Dutch health care: the prevalence of private funding and " the long tradition of non-government provision of care" (Okma 2009).Indeed, this system facilitated the development of a flourishing private insurance market and, until 2005, the Dutch government accepted this situation so long as private insurance enrollees remained below 65 percent of the entire population.
The pre-2006 system was thus prone to concerns regarding inequity in financing and access to health care in the Netherlands since higher-income individuals clearly held a serious advantage relative to poorer individuals in terms of insurance access.With the context of economic crisis of the 1980s in the background, reforms were clearly needed at the time and this culminated in the enactment of the Health Insurance Access Act (1986), which enforced the establishment of affordable minimal baskets of service within privately-run insurance funds (Den Exter 2004).Already in 1986, then, the state was expressing its will to offload the burden of the so-called poor risks onto the private market.Early on, the market was a tool employed by the government to modernize health care.
A cornerstone in this evolution is the release of the Dekker Commission report in 1987.In a nutshell, this report proposed the creation of a " regulated market" within the framework of social health insurance (Okma 2009).The commission, which based its conclusions on Enthoven's model of managed competition, advocated for guaranteed universal access to basic health care services while "[…] it was thought that regulated competition would create incentives for both insurers and providers to improve the efficiency of health care delivery" (Schut and Van de Ven 2005: 65).The report's recommendations were to be partially implemented by subsequent governments; in fact, most students of the Dutch health care system term the 15 years following the release of the Dekker report " the pre-managed competition phase" leading up to the actual turn towards regulated markets, achieved in 2006 (Bartholomée and Maarsee 2007; Enthoven a n d Van de Ven 2007).
Following the release of the Dekker report, the government sought to encourage consumer mobility by terminating the " regional boundaries of sick funds," which meant that all of those health funds were to become available to the population irrespective of location; moreover, this change implemented competition between health funds, which could, from then on, use the whole country as a market in which to attract enrollees (Okma 2009).As in Great Britain and Quebec, to make this early regulated market work, government-sponsored websites were designed to enable patients-consumers to make comparisons Despite all these market-oriented changes, sweeping reforms had yet to be implemented and discontent regarding the health care system stood firm in the population.As the 1990 Simons Plan asserted, the state had to find a way to simplify the structure of the public system with one consolidated mandatory insurance package and " [ …] introduce competition between health insurers" (Den Exter 2004: 78).First, more regulation had to be passed to frame private insurers' range of action.Also, the government had to reinforce measures aiming at encouraging providers' efficiency, such as medical pricing.Next, a reinvigorated system of evaluation not only focused on cost, but also on "outcomes and quality of care" had to be firmly established so as to make the information on health providers even more widely available.Finally, the terms of competition and a governance structure adapted to such a quasi-market had yet to be determined (Schut and Van de Ven 2005).
In 2002, the government revealed its intentions pertaining to health care reform with the release of the plan.A Question of Demand, which constitutes the primary base on which the 2006 health insurance reform was crafted.This report bridges all of the efforts undertaken by several Dutch governments to gear the system towards a true regulated market.Among other things, the report stressed that the 2006 reform was to effectively transform the state into an umpire for a health care market for which tight regulation of private health insurance companies was to be in place in order to prevent risk selection and risk inequality, and to generate incentives for providers to efficiently operate (Ministerie can Volksgezondheid Welzijn en Sport, NL 2002;Custers and Arah 2007).
The 2006 Health Insurance Act thus dramatically transformed the post-war health care system in the Netherlands despite the fact that the system remains true to its welfare-oriented/mandated origins.Dutch citizens are now obliged to enroll in a private plan whose basic package is regulated by the state and is open to all residents without risk selection.In addition, each insurance company offers supplemental services that are mostly unregulated and more costly for users.Competition is brought forward through differences in "premiums, service, and the quality of care offered by contracted providers" (Enthoven a n d Van de Ven 2007: 29).A centralized risk-equalizing fund to which all citizens contribute was created to fund those private insurance plans that draw higher risks.The resulting budget takes risk factors like age, gender and health status into account-all that provided by insurers.Hospital management must focus on performance and efficiency; the state funds hospitals on a prospective basis and health insurers are allowed to freely negotiate contracts with hospitals as well.As a result, the health care delivery sector adopted a fiercely competitive and rational model of management (Enthoven a n d Van de Ven 2007; Bartholomée an d Maarsee 2007;Vaillancourt Roseneau a n d Lako 2008;Okma 2009).The state also ensures that all citizens are effective customers in this regulated market through a program of income-based subsidies.
Finally, two things seem to be especially salient in the Dutch case: first, since the idea of "incentives and competition" seemed to be already widely discussed in the 1980s, policy options remained circumscribed to the development of a variant of regulated market in the following years (Vaillancourt Roseneau and Lako 2008;Okma 2009).Second, as opposed to France, the Netherlands did not have to completely modify the power structure to implement reforms; the approach mostly remained centered on the reorganization of suppliers and providers throughout as changes in health insurance companies' roles in the system attest to with the 2006 reform.It seems clear to us that while France emulated approaches to managed competition taken by our comprehensive-system cases, the Netherlands opted for a different route.
Entrepreneurial System: Massachusetts American states are dominated by private insurance and private provision of health care services; however, they have a residual public health care sector for which they hold some levers to extend the scope of public coverage as they see fit.Medicaid, a program to which states and the federal government contribute financially and that primarily aims at providing worst-off citizens with financial assistance in health care, can be expanded and modified,3 or left at a minimal scope.Massachusetts is one state that undertook a turn towards mandated health insurance in 2006 by reallocating Medicaid funds and enforcing an individual mandate.A less visible but nonetheless important aspect of this reform is the fact that Massachusetts, in 2006, started regulating suppliers and providers of health services similarly to systems where public health insurance is more dominant.
In 2006, the Massachusetts government passed the Act Providing Access to Affordable, Quality, Accountable Health Care.The goal of this reform was two-fold: first, to give birth to a more comprehensive public system and second, to ensure better resource-management (efficiency) in this public sector.The 2006 reform 1) implemented a mandated system of health insurance and 2) created a semi-regulated market for those private insurance plans that jumped the bandwagon of reform (Muiser 2007). 4  First, the " c onnector" is a body that was created to certify insurance products based on quality and value indicators in order to direct publicly-subsidized customers towards the best plans and to provide insurers with more incentives to deliver better, more affordable services.The certification process induces competition since, with the reform, the state subsidizes individuals on an income basis so that they enroll in "connector"-certified private insurance companies (Gruber 2006;Raymond 2007).From a purely economical point of view, it is potentially advantageous for some insurance companies to be "connector certified" since citizens are obliged to enroll in an insurance plan even though they have no access to health insurance in the private sector; and at the same time, the state ensures greater coverage and strongly supports the best/ most efficient insurance plans through its use of market principles.Competition between insurers is also stimulated through consumer-oriented products, such as a state-sponsored website that provide patients with cost and quality information since 2008.
Interestingly, the 2006 bill enforces a host of policy measures whose goals are to address overutilization of health resources in the public sector.For example, the funding of Medicaid facilities ties rate increases to specific performance goals related to quality indicators.Also, Massachusetts adopted a standard measurement of annual health care spending for the state, the "Massachusetts Global Health Cost Indicator," as a means to encourage holistic efforts to attain lower expenditures in the health care system (Raymond 2007).Indeed, the 2006 reform gave rise to a policy agenda focused on cost containment.For example, the state created a Special Commission on the Health Care Payment System which, in July 2009, recommended a shift from fee-for-service to " a system where providers work together to share the responsibility for the patient's care" (Halslmaler a n d Owcharenko 2006;Bebinger 2009).This will translate into the creation of accountable care o rganizations, which would receive global payment for services provided to patients, thereby rendering providers accountable for meeting cost and quality targets.
All in all, the 2006 reform of the state of Massachusetts resulted in the creation of a semi-regulated market in a public sector that is still at the margins of the mainly private one.On the demand side, information is disseminated regarding quality and effectiveness of providers.On the supply side, the government slightly modified the role of a key actor in the American health care system-insurance companies-by means of market incentives to improve performance with the reward of attracting more customers.On top of that, new public management (NPM) measures are being implemented to control the state's health care expenditures.Clearly then, elements of a regulated market are noticeable in Massachusetts whose post-2006 health care system displays striking similarities with the Netherlands, even though the two systems are initially classified by Roemer in different categories.

Conclusion
As shown in our case studies, even if reform processes unfolded differently, all five health care systems have been modified to a certain extent with the implementation of market-oriented principles; in fact, we showed how managed competition was especially central to the development of internal and regulated markets everywhere, albeit under very different reform goals and scopes.These changes imply that social scientists have to take into account market-oriented governance in further comparative health system research.
Our main conclusions start with the fact that internal markets, as indicated in F igure 1 below, are becoming the backbone of health care systems where governments provide comprehensive health care services.Great Britain and Quebec implemented market-oriented measures in the 1990s and 2000s which led to the creation of decentralized agencies; the latter were then responsible to oversee health regions as well as were made accountable to health ministries and increasingly, to patients.
When faced with the need for reforms, 1990s governments of welfare-oriented health systems had a host of policy options to choose from due to the decentralized nature of their systems based on health funds.The French government centralized governance to better decentralize health management following "comprehensive-system" type market-oriented reforms.In the Netherlands, early efforts were geared towards the establishment of a regulated market in the 1980s and successive governments sought to implement missing elements required to change insurers', care providers' and patients-consumers' incentives to make the system more efficient and still accessible to all.
Finally, our American-entrepreneurial case is truly fascinating because it shows how marketoriented governance and managed competition can be used to implement regulation in a very small public sector and predominantly private market.As a matter of fact, an individual mandate supported by market-oriented measures to incentivize insurers, providers and enrollees to embrace the reform (and to do so efficiently!) was the outcome of these changes.In light of our findings, Roemer's typology is useful for identifying original health system characteristics from which reforms stemmed in the era of incentives and competition.For example, certain characteristics made health systems resistant or more conducive to the development of regulated and internal markets in health care systems.On the one hand, a higher degree of centralization inherent to comprehensive systems clearly hampered the creation of internal markets in this category; on the other hand, less centralization in other health systems led to different reform processes and different types of health care markets.However, it remains that Roemer's typology does not correspond to current health systems at least on the governance dimension: for example, the Netherlands and Massachusetts evolved towards regulated markets and individual mandates, whereas France seemed to get closer to comprehensive health care systems with the implementation of an internal market.
Of course further research should include more health systems to confirm our hypothesis, but it seems clear to us that structural changes brought about by market-oriented governance in the 1990-2010 period significantly altered health care systems and their pre-"incentive and competition" characteristics.Furthermore, there is ample evidence that other health systems are undergoing equally important changes due to market-oriented governance.Germany, a welfare-oriented system, is one important example.Since 2007, Berlin is in the process of reorganizing the decentralized sickness funds' overall funding structure; indeed, on the one hand, the federal government wants to determine health insurance contribution rates as opposed to leaving this task to the 250 decentralized sickness funds.On the other hand, the implementation of market-oriented reforms is set to encourage competition between sickness funds and increase patients' choice.Interestingly, in American states, things are moving on the health reform front beyond the borders of Massachusetts.For instance, Vermont's legislature passed a law making universal health care a right for all citizens in 2010 (Rudiger 2010).It remains to be seen whether such a health system, which is set to be ready for implementation by 2012, can be designed upon nationalized systems' principles or on market-oriented principles, the latter obviously being favoured by private health insurers.
In o u r final analysis, there is one important element to retain from this paper for further research.As Roemer himself asserted, health care governance underwent profound modifications in the 1990s and has to be incorporated in any comparative work in health system analysis.Moreover, if market-oriented governance is so pervasive in health care systems of advanced-industrialized societies and does modify actors' behaviour, incentives and expectations as evidence seems to demonstrate, then future typologies shouldn't refrain from increasing the theoretical focus on the governance dimension of health systems.After all, this is where the future of health care system is crafted and implemented.
Melanie Bourque PhD. is a professor at l'Université du Québec en Outaouais at the social work department.Her research interests are the evolution of social protection in the developed world.More specifically, she is analyzing health care system reforms, family policies and work and family conflict.She has publications in these areas of study.
Jean-Simon Farrah is an M.A student in political science at McGill University.He has worked at the IRIS lab for two years on health systems as a research assistant.This research has been done at the International Research Infrastructure on Social Inequalities in Health (IRIS) and made possible through the financial contribution of the CIHR.

Figure 1 :
Figure 1: Outcomes of market-oriented reforms for Roemer's categories

workers, state) Public Decentralized France Netherlands Entrepreneurial Private for population, public for limited targeted population Private, out of pocket, collective insurance
Table 1 below.Private insurance can be collective or individual, although it is clear that employment groups tend to be the