Document(s) 3 of 14

The Social Protection Model in Latin America

As Flora and Alber (1981) have pointed out in their comparative study of the development of welfare states in Europe and America, modern western systems of social protection originated in the profound economic, political, and cultural changes that accompanied the processes of industrialization and urbanization. Alterations in authority patterns, changes in community values, weakening family ties, and the emergence of class identity were some of the features that caused responsibility for the provision of individual and family needs to shift from the private to the public sphere. Thus, the role of the state became more interventionist in order to ensure emerging social rights.

Although the development of social protection has a common origin, the institutional form it takes varies according to prevalent values and previously existing institutions. It is also determined by the dynamics of the power struggle between the important classes and social actors — their goals, strategies, alliances, and ability to organize and create social consensus.

Nevertheless, analysts have tried to extrapolate patterns of social protection from specific cases and come up with three ideal types of social protection systems that group similar styles of governmental intervention and organizational arrangement as an outcome of former political choices. These are: the corporatized system, the nationalized system, and the segmented system (Heidenheimer et al. 1983). These indicate, respectively, an insurance-based arrangement, a state-supported system, and a market-based system combined with public assistance.

Some critics feel that such classifications are purely descriptive and cannot properly be called instruments for comparative analysis. Others point out that it is impossible to fit any real system into an ideal type, as an ideal type indicates a prevailing tendency rather than a complex historical dynamic. Although agreeing with this criticism, we cannot deny the utility of identifying the common features in the social-protection development process as a useful tool in the analysis of the present situation of health care systems in Latin America.

Other authors recognize the importance of identifying common patterns of social policies, but call attention to the enormous variation in the arrangement of any specific public policy such as health care. This arrangement is the result of political choices on many different issues, such as: ownership of the services; styles of contracting out; means of regulating professional activities; policy-making structure; techniques of intervention; distribution of health care services; and guarantee of access (Heindenheimer et al. 1983, pp. 52-89). That being said, such a classification is undeniably useful in our analysis of the present situation of health care systems in Latin America.

Latin America has traditionally been ruled in a oligarchic way, combining patrimonial mechanisms — the private appropriation of public goods by the elite — with the populist exchange of privileges — the co-opting of workers by government — and the social exclusion of certain groups of people, mainly the poorest.

Nevertheless, social demands have been incorporated by governmental authorities since the beginning of the 20th century, as part of the process of urbanization and the launch of the industrialization process. Some of the most powerful, complex, and long-standing social security systems are to be found in the region, especially in those countries considered pioneers: Argentina, Brazil, Chile, and Uruguay (Mesa-Lago 1978). There, social security institutions date back to the turn of the century, and have formed one of the most important channels in the relationship between the populist leader and the urban working class.

The government bargained with each labour faction separately, exchanging benefits for political support and thereby deriving its legitimacy. Thus, the social security system was built up on fragmented grounds, with benefits being expanded to the same entitled workers and coverage being enlarged as part of the political game of pressure and bargain, while its financial basis was entirely rooted in taxes and contributions from salaries.

The assimilation of the social demands of urban workers has resulted in different levels of social protection in each country, determined mainly by the degree of homogeneity and organization within the working class (to compare the development of social protection in pioneer countries, see Fleury 1994).

Thus, while social policies and social security systems in Latin America played an important role in the process of state-craft, they did not succeed in spreading civil culture or in extending social coverage. Although the concept of citizenship was based on an egalitarian notion of rights, the concept of social protection in Latin America rested on social and institutional mechanisms of differentiation. Nevertheless, this political give-and-take constituted the first instance in which the demands of the working class were considered in the political arena and incorporated into the government agenda.

From this process emerged the main actors in the social policy arena: the technical bureaucracy; middle-class professionals; unionized workers; and, last but not least, the traditional populist politicians. Social protection was rooted in a political system wherein the state played a key role in the industrialization process by combining industrial protectionism with a controlled political incorporation of urban workers’ demands. This apparently strong state interventionism was weakened, however, by the compromises that had to be made with the many conflicting and powerful interests that lent the government legitimacy, but also held it back.

The need to cater to vested interests was responsible, on the one hand, for increased state intervention in social and economic life. On the other hand, it prevented the government from imposing progressive taxes on the productive sector, resulting in chronic fiscal deficit. This populist game, performed by state and society and based on mobilization and concessions, was paradoxically responsible for both the power and the weakness of the state.

As the course of industrialization proceeded, the ever-growing complexity of the political structure, dissatisfaction due to unfulfilled expectations, and the exacerbation of contradictions between co-optation and control made it impossible to fulfill emerging political and social demands. As a result, the expansion and maturation of the social protection system in the region manifested itself as a crisis that could not be resolved within the existing political framework.

With the collapse of populism in the 1960s, social protection institutions suffered an intentional demobilization of popular participation followed by several attempts to concentrate political power in the hands of a bureaucratic–authoritarian central executive.

At that time, the social security institutions controlled by workers in the pioneer countries were about to go bankrupt. The first reform movement carried out by the military dictatorships was to take those institutions out of the hands of the trade unions and put them into the hands of the bureaucracy. This was successfully carried out in Brazil, but failed in Argentina, where the workers were able to keep control of the obras sociales, the social security institutions associated with unions.

However, the strengthening of bureaucratic structures was accompanied — to a greater or lesser degree, depending on the country — by a general trend toward de-politicizing social issues, within a context of political suppression and the elimination of representative and organizational channels for expressing social demands.

The rupture with the populist style of relationship between state and society did not change the "corporate pact" of power and control over the political actors by the authoritarian governments. However, bureaucracy was substituted for corporate mechanisms of representation and conflict negotiation, which clearly favored the interests of private entrepreneurs. In the context of social security, this meant private health care providers, whose influence on public policy resulted in a state logic that was predominately private.

In any event, each country was left with a stratified health care system whereby workers in the formal market were entitled to social security benefits while the rest of the population received services provided by the ministry of health. This resulted in significant differences in access to and quality of services.

Regardless of differences in social protection policies from country to country, it is possible to note some common characteristics. These include:

• duality — some services are provided by the ministry of social security and others by the ministry of health;

• fragmented institutional arrangements that vary in terms of duties and rights according to the bargaining power of the labour group involved;

• centralization of the social services system, resulting in inefficient service delivery and poor quality services;

• low coverage by a social security system that provides uneven coverage across regions and income classes, and excludes workers in the informal sector; and

• a complex and intricate power structure, dominated by such powerful stakeholders as politicians, bureaucrats, labour leaders, private providers, international suppliers of health-care inputs, and insurance companies.

Segmented Systems are characterized by the coexistence of the following: social security systems covering workers in the formal sector; a health ministry that carries out some public health interventions and provides health care services for the lower and middle classes; and a very large and diverse private sector that covers the richest, the poorest, and, increasingly, the middle class. Mexico was identified as an example of a segmented system.

Integrated Public Systems

are those in which the public system predominates. Financial resources come out of general tax revenues and are allocated to integrated public providers on the basis of installed capacity. Costa Rica was identified as an example of this type of system.

Subcontracting Systems

are those in which the government plays an important role in mobilizing resources while provision of services is largely private, usually on a fee-for-service basis applied with various degrees of sophistication. Brazil, where abundant private resources complement public efforts, is the only country in the region to approximate such a system.

Contract-Intensive Systems

are characterized by the prominence of social security organizations, for which payroll taxes are the major source of financing. These usually involve a formal separation between finance and service provision, and apply a wide range of service-contracting mechanisms. Argentina was identified as an example of this type of system.

This typology emphasizes the relationship between payers and providers, but also takes into consideration the ownership of health services. As a descriptive tool, it does not indicate the reasons why a health system assumes this configuration, nor its tendencies to reform it. All of them are identified as facing problems in improving the quality of services and achieving efficiency. Nevertheless, it would be misleading to conclude that the different configurations of the health system in the region are coming to the same point of adopting a standardized reform model.

After a Decade of Reforms in Latin America

The structural reforms implemented in the late 1970s and throughout the 1980s saw the emergence of a new style of economic development based on the leading role of private investment and exports, the expansion of private consumption, and a limited role for the state. In most countries of the region, adjusting to the external financial crisis led to regressive income distribution and an acute decline in real wages, further increasing inequality and poverty (O’Donnell and Markovitz 1996).

During the 1980s, the "lost decade," the political and social balance achieved during the postwar era of development quickly broke down. Most of the countries emerged from the adjustment period with greater levels of inequality (the exceptions being Colombia, Costa Rica, and Uruguay) and poverty (except for Chile and Uruguay).

Although economic stagnation and inflation were the main factors responsible for poverty and inequality, adjustment measures taken in response to the fiscal crisis also contributed. The new style of development based on increased productivity and competitiveness deepened existing structural problems such as lack of investment capital (national savings), vulnerability to external shocks, and lack of an income distribution policy to cope with poverty and inequality.

As the region recovered, experiencing economic growth without inflation, purchasing power increased in some countries and the income distribution structure was altered across the region, with varying effects. Only the high-income groups benefited in Mexico, while all income groups benefited in Chile, Colombia, and Uruguay. In Venezuela, the incomes of all groups deteriorated, while in Argentina and Brazil the middle classes and the poor suffered the greatest losses (Altimir 1994). Only in the mid-1990s would the Brazilian stabilization policy succeed in reducing inflation and, consequently, poverty.

According to a report by the Inter-American Development Bank (IDB 1997, p. 17), "after falling continually throughout the 1970s, poverty increased dramatically in Latin America during the 1980s. By the end of the decade, the proportion of the population living in moderate poverty had risen to 35%, and the share of the population in extreme poverty had risen to roughly 17%. During the 1990s the poverty rate has declined very slightly, while the number of poor has increased somewhat...."

Weak recovery in the 1990s was attributed to the insignificant decline in poverty rates, which may also reflect the manner in which the gains from recovery were distributed (IDB 1977, p. 18). Between 1991 and 1995, per-capita income of the top fifth on the income-earning scale grew by 5%, whereas per-capita income of the middle and lower income groups grew by only 3%. The overall conclusion was that the pattern of income distribution has not shifted in this region, which is noted for having the most unfair income distribution in the world.

Concern about the fragility of the region’s economic recovery centered on the fact the that growth that has been restored in some countries is heavily dependent on private foreign investment. Many experts have suggested that privatization of the social security funds would be the most promising way to raise private savings (Williamson 1997, p. 51) and reduce reliance on public investment.

While other international agencies attribute these failing efforts to reduce poverty to the weakness of the recovery process, a report by the Economic Commission for Latin America and the Caribbean (ECLAC 1996, p. 44) takes a more pessimistic view. The report attributes the persistence of the same pattern of income distribution to some characteristics of the job market, such as the persistently high rate of unemployment, the limited bargaining power of many wage earners, and the increased income gap between workers with different skill levels.

Economic recovery is of critical importance to the social sector, because the characteristics of the labour market determine access to social benefits, as well as the possibility of joining a private health insurance plan. The formal job market not only determines salary levels, but also the level of resources collected by social security that are destined for the health sector. Thus, the low income earned by a high proportion of public sector workers — in Bolivia, Honduras, and Venezuela, between 30% and 40% of public sector wage earners were below the poverty line in the mid-1990s (ECLAC 1996) — limits the scope for state reform of social security, and especially affects the social sectors, where the lowest salaries in public administration are to be found.

An analysis of the process of change in income distribution and the debate regarding the role played by public policies is revealing. The Chilean case shows that even under sustained economic growth, inequality remains higher now than during the pre-adjustment period, while the examples of Costa Rica and Uruguay show that inequality can decrease as a result of income distribution mechanisms.

The evidence points to the conclusion that there is no economic theory of wealth redistribution, and that the positive results achieved resulted from a set of political policies and increased public expenditure on social services. In other words, Latin America’s historic experience illustrates the thesis that, while economic growth is a necessary condition for the struggle against poverty and the reduction of inequities, it cannot by itself improve income distribution. The judicious use of social expenditure is a better way of improving income distribution because it is relative independent from both the level and rate of economic expansion.

The last period of crisis and the implementation of structural economic adjustment measures led to reduced employment and wage levels and increased poverty and inequality throughout the region. Nevertheless, it did not reverse the long-range trend toward improved social conditions. Rather, it reduced the rate of social improvement, particularly in the areas of educational coverage, water and sanitation services, child mortality, and life expectancy at birth. Continued improvement in social conditions must be attributed to the social service networks established in several countries during the period of economic expansion and to socio-demographic characteristics associated with the accelerated rate of urbanization.

Fertility rates have declined rapidly in Latin America since the mid-1970s, reducing population growth and eliminating the threat of overpopulation. The challenge is now to take advantage of this opportunity by improving the quality and coverage of social programs, particularly with respect to young people living in poverty.

In addition to reduced fertility rates, the region has experienced significant economic and social changes in the recent decades, mainly, an intense process of urbanization and the aging of the population. Both phenomena have altered the demand for health care, since urban demands are more visible than the rural demands, and the aging of the population has redrawn the epidemiological profile. To the typical diseases of underdevelopment have been added those prevalent in industrialized societies: cardiovascular disease, chronic illness, accidents, and violence.

In Latin America, the health sector is a powerful economic force, employing about 5% of the economically active population (EAP). Expenditure on health represents on average 5.7% of gross national product (GNP) in the countries of the region (CEPAL 1994). However, public expenditure on health care accounts for only 49% of total health expenditure in the region; private expenditure for 50% and foreign development assistance for the remaining 1% (Govindaraj et al. 1995). In terms of public expenditure on health care, the region does not compare favorably with the world average. Public expenditures on health care in the region is US $121 per person per year, or 2.2% of GNP, less than half of the world average of about US $323 per person per year (World Bank 1993).

While the level of public spending on the social sector increased during the early 1990s in most countries in the region (ECLAC 1996, p. 79), the increases varied greatly between countries. In some countries, expenditures now exceed the levels prevailing in the 1980s. This is the case in Argentina, Chile, Colombia, Costa Rica, Mexico, Panama, Paraguay, and Uruguay. In other countries, expenditures show an upward trend, but have not yet regained early 1980s’ levels. Such is the case in Bolivia, Ecuador, and El Salvador. Social expenditures in yet other countries, such as Brazil, Guatemala, Honduras, and Nicaragua, have fluctuated or declined slightly. Moreover, most increased public spending in the region occurred in the areas of education and social security, rather than health.

During the 1980s, reduced public spending was accompanied by an increase in health expenditures on the part of families, enterprises, and nongovernmental organizations (NGOs). The NGO contribution to the region’s health financing was significant, totaling US $6 billion between 1980 and 1989. In poorer countries, this contribution was similar to or even exceeded that of national governments.

As a result of reduced public spending, public health services tended to deteriorate, the technological gap between public and private hospital services widened, and the efficiency and effectiveness of publicly managed and provided health care declined. This was accompanied by a growth in services supplied by the private insurance sector, which increased significantly both in affiliation and number of hospital beds. Thus the pattern of supply and consumption moved toward greater involvement of the private sector in health care delivery.

In addition to changing market dynamics, there also occurred in Latin America a great transformation in the social fabric and institutional political framework. Since the 1970s, the old relationship between state and society that had developed during the early years of industrialization — known as the "corporate pact" — showed itself incapable of encompassing the complex and pluralistic network of political actors that emerged as part of the urbanization and industrialization processes. Grassroots social movements and NGOs flourished in the region, becoming a channel for organizing and representing interests that did not fit into the old political order. Even traditional political actors, such as trade unions and entrepreneurs, developed new forms of organization and representation, clearly demonstrating the insufficiency not only the old structure but also the political party and electoral systems. Moreover, the increase in unemployment and changes in the work process have weakened the power of the trade unions in the region and limited their ability to control or implement social policies.

The transition to democracy in many countries of the region in the late 1980s changed the existing pattern of authority into a scene of conflict, where consensus, as well as the institutional framework of democracy, had yet to be built up. In this new context, social and grassroots demands had to be considered in the political arena, some of them becoming part of the government agenda. At the same time, however, economic adjustment measures imposed tremendous cuts in public spending, emphasizing the contradictions between necessities and constraints.

Deepening inequity, the "urbanization of poverty," and the failure of the state to guarantee even minimal economic and social functions (a stable currency, taxation, education, health, and housing), led to unprecedented urban social breakdown, manifesting itself in violence, criminality, lawlessness, and the proliferation of drug trafficking such as can be read about any day in the regional press.

The emerging individualist, or self-centered, system of values acts as a barrier to keeping or developing solidarity, destroying the political and family ties that traditionally provided the poor with social and economic support. Another aspect of individualism, that of consumers’ free choice, has had the effect of reducing the power of the bureaucracy to determine the nature of public goods to be offered and the manner of their consumption.

Some Issues in Comparative Analysis

Over the last decade, health care systems in several Latin American countries have undergone a number of organizational changes in line with the worldwide trend toward health sector reform. Such changes were perceived as an answer to the financial crisis of the 1980s, an attempt to make health care institutions more effective and as a way of adapting to the more sweeping process of economic restructuring and the rethinking of the role of the state.

The first question one has to ask about these processes of health care reorganization is whether or not they can rightly be considered as sanitary reform (health reform). Of course the answer will depend on what is meant by the latter.

Sanitary reform encompasses changes at both the institutional and the policy levels, transforming both prevailing values and the authority system that is designed to govern individual actions. These transformations are processed in a sustainable way, no matter whether the strategy has been piecemeal or comprehensive. Reforms are supposed to affect crucial features of health care organization, altering access and utilization, as well as institutional arrangements and political and economic power-sharing among the various actors in the field.

The political process of sanitary reform in Italy and Brazil has inspired a definition that emphasizes the political and the institutional changes (Fleury 1989).

Sanitary reform refers to the process of transforming the laws and the institutional arrangements responsible for protecting the health of citizens, and corresponds to an effective displacement of political power toward less-advantaged groups. The concrete expression of this change is reflected in the establishment of a universal right to health care and the creation of a national health system headed by the state.

The same emphasis on the political aspect is evident in another concept of sanitary reform (Berlinguer et al. 1988, p. 131), although it also calls attention to the transformations of the health care model and services.

When we aim to broaden the equality of citizens with respect to health care; to displace the power of the few to the many; to change productive technology and individual behaviour in a healthy way; to recycle the services, the technical interventions, the treatment and prevention of disease; and to drive conscientiously other transformation processes (without which any reform would stay isolated and be defeated), we will clearly face many difficulties and huge resistance. Consensus is the most important weapon to overcome them.

In other words, after the profound changes in the health sector in Europe in the first half of the century, the discussion about sanitary reform acquired a political meaning in terms of:

• rebuilding the power structure by promoting sanitary conscience (health awareness) and organizing a new hegemony;

• altering the legal and institutional health care apparatus, with the state guarantee of universal right and access to a national health system; and

• shifting from a curative to a more preventive orientation, accompanied by more adequate treatment.

The missing point in this case relates to the microeconomic aspects of health care system organization, i.e. the financing and organizing health care units. Recently, this matter has become paramount for reformers, changing the concept of sanitary reform into that of health care reform.

This new wave of health sector reform started in a different context, in the middle of the economic crisis of the 1970s. It was prompted by the increasing cost of health care services and the more complex and expensive demands made on the health system by an aging population. In response to national fiscal crises and to the difficulty of financing health care services, health sector reform is underway or under consideration in countries throughout the world.

Although it is difficult to define precisely, Berman (1995, p. 15) acknowledges that the term "health sector reform" has wide appeal. He defines it as "sustained, purposeful change to improve the efficiency, equity, and effectiveness of the health sector." He refers to the health sector as all the policies, programs, institutions, and actors that together make up the organized effort to treat and prevent disease.

In a comparative analysis of seven countries that are members of the Organisation for Economic Co-operation and Development (OECD), Hurst (1991, p. 14) tries to find similarities in terms of some set of objectives as expressed in their health policies. They areas follows:

• individuals should have adequate and equal access to health care;

• payment should be linked to an individual’s ability to pay;

• health expenditure should consume an appropriate fraction of GDP (macroeconomic efficiency);

• the mix of services provided should maximize health outcome and consumer satisfaction for the share of GDP expended (microeconomic efficiency);

• consumers should be free to choose arrangements from both the public sector and the private sector; and

• doctors and other providers should be given the maximum freedom compatible with the attainment of these objectives (appropriate autonomy).

The concept of reform that grew out of the attempt to overcome the crisis in the welfare state focused on the need to increase the efficiency and reduce the cost of health services. It also proposed changes in the role of the state and private providers, as well as a shift in the status of the user. This concept of reform sees the state as having a minimum responsibility while assuring the providers their autonomy and dealing with the users as consumers with freedom of choice. Equity of access and protecting patients against financially debilitating payments for health care are also on the agenda, mostly in the developed countries although not in the United States. The discourse of empowerment that accompanied the first reform movement has been replaced by one that emphasizes rationality and competition.

According to Chernichovsky (1995), the systematic functions involved in the new generation of reforms are the financing of care, the organization and management of the consumption of publicly funded care, and the provision of care. Reforms in line with this paradigm attempt to combine the advantages of universality, efficiency and consumer satisfaction by offering universal access to a basic package of care through decentralized, mainly private, systems of delivery. To achieve this goal, the two functions of financing and provision must be separated because the first is guided by a public principle whereas the second is based on competition.

International agencies and experts from the region took part in a special meeting in 1995, In order to build a regional agenda for health sector reform in Latin America (Health Sector Reform Seminar 1995). Participants in the seminar defined health sector reform as, "a process aimed at introducing substantive changes into the relationships between and roles performed by the different agencies involved in the health sector, with a view to increasing equity of benefits, efficiency in management, and effectiveness in satisfying the health needs of the population."

The objectives of health sector reform for the region were identified as follows:

• to improve the health and living conditions of all the inhabitants of the Americas;

• to become, along with justice, well being, and equity, one of the pillars of development in the region;

• to reduce health status inequalities, improve access to good quality health services, and foster shared responsibility between institutions, individuals, and communities;

• to modernize and decentralize the organization and operation of public institutions providing health services;

• to balance the public and private health sectors in order to achieve a synergy in their efforts; and

• To ensure that sufficient financial resources are available to the sector at a sustainable level to accomplish the aforementioned objectives.

The emphasis is on regional problems and strategies, such as promoting equality of access; improving the quality of health services; modernizing the state and decentralizing the provision of services; and gathering resources to operate and finance the health care system.

Nonetheless, the concept of health sector reform proposed for Latin America neglects some important aspects of the health sector, such as the development of science and technology, human resources training, and the ever-increasing proliferation of curative treatments.

In this research project, we have avoided choosing an existing concept of health sector reform, because we felt that it would lead to an analytical bias in favour of one concept or another. We have also tried to avoid considering health sector reform as a simple rearrangement of organizational variables such as financing, provision, and coverage. Rather, we have chosen to conceptualize health sector reform in terms of the following dichotomies:

• universal versus segmented policies;

• mechanisms of solidarity versus equity between contributions and benefits;

• integrated systems versus fragmented systems; and

• decision-making processes based on participation versus decision-making processes based on competition.

Actual health sector reforms can be expected to fall somewhere along the continuum between the two extremes, permitting many different arrangements. Also, a tendency in one direction or another will be restrained by the existing institutional and political organization, tending to add to the diversity and complexity of the reforms.

We agree with the conclusion of the Health Sector Reform Seminar (1995, p. 11) about the diversity of processes in the region.

The objectives, scope, and substance of reform, as well as the strategies and mechanisms adopted for their implementation, vary markedly. Some reforms are comprehensive, simultaneously encompassing the organization, the financing and the resources of the sector as a whole. Other reforms are intended to implement only partial changes among some of the institutions or roles of the health sector. In some cases, reform involves the promotion of greater private sector participation in health, whereas, in others making decentralized and democratic government more responsible is what has been attempted.

One way to avoid the problem of diversity when comparing the reforms is to analyze the purposeful change made by governmental authorities, and expressed in legal instruments. But this may be misleading. Kroneman and Zee (1997) call attention to the troublesome task of investigating health policy reforms at the national level, because of the difficulty of establishing when a certain policy change took effect and of determining the content of the reform. The difficulties are due to regional variations in a decentralized system, as well as to the fact that reforms are a gradual rather than a straightforward process. Moreover, deliberate vagueness in national policies is used by reformers as a means of avoiding confrontation with health care providers and other actors, and of creating broad support for the reform.

Nevertheless, a framework can be constructed that avoids these pitfalls by focusing on the accomplishments and problems associated with various national experiences currently underway. For example, in examining the purpose, content and scope of the sanitary reform processes in Argentina, Brazil, and Mexico, it is possible to discern a common trend toward modifying the relationships between public and private agents and between different levels of government.

In our comparative framework, we decided to concentrate on the changing aspects of the health care system, thereby focusing on the main point of consensus in different concepts and processes of sanitary reform. Both those processes that emphasize the political aspect of reform and those that emphasize its managerial content have in common the fact that all the intended changes should appear in the organizational structure of the health care system.

Although the decision to concentrate the comparative study on the health care system may avoid problems related to the definition of reform, other comparative issues persist because health systems comprise many complex interactions permitting innumerable variations in policy and in delivery and control mechanisms. This complexity leads many authors to doubt the possibility of accomplishing comparative analysis at all.

For Ellencweig (1992, p. 16), "

¼ the only axiomatic statement that can unify all health systems and their subdivisions would be that they differ from each other. Those differences are not due to chance alone, rather they are rooted in the organization of societies." Similarly, Light (1997, p. 109) states that "in some ways, health-care systems are inherently not comparable."

Nevertheless, many studies have attempted to compare national health care systems, mostly using a health service model. The health service model develops criteria to group national health systems on the basis of a cross-sectional assessment of each of their components. (Ellencweig 1992, p. 25). One weakness of this model is its implicit effort to reduce the explanation of a multifaceted health care system to one or two variables. Its other weakness is that it disregards trends over time.

Roemer (1977) is the most outstanding advocate of the health services approach. He distinguishes the evolution of health systems by identifying four major attributes: historical perspective, political process, socioeconomic environment, and cultural characteristics. His comparative study (cited by Light 1997, p. 109) presents a dynamic model of common elements based on five factors: resources, economic support, organization, management, and delivery of services.

Traditional taxonomy of national health systems was based on ownership, including accountability for financing and provision. This classification could accommodate many different variables, from societal values to market or institutional arrangements involved in a national mandatory insurance scheme. More recently, with the increased complexity of health systems, many different relationships have developed between the government, mandatory and voluntary insurance schemes, and private providers. Thus classification of health systems by ownership is no longer considered useful. Instead, more recent typology is based on the functions performed by each component of the system: financing, regulation, organization, and provision (Lodoño and Frenk, 1997).

The health care system is analyzed in terms of its capacity to integrate either populations or institutions. With respect to populations, integration means the extent to which different groups are allowed access to every institution in the health system. Regarding institutions, integration refers to the arrangements for carrying out the functions. Upon this framework, Lodoño and Frenk (1995) created the following typology of health systems models in Latin America:

• the unified public model, featuring vertical integration of institutional functions and horizontal integration of populations;

• the segmented model, featuring vertical integration of institutional functions and segregation of populations;

• the fragmented model, featuring separation of institutional functions and horizontal integration of populations; and

• the atomized private model, featuring separation of institutional functions and segregation of populations.

This new way of classifying the components by function is based on two premises: the necessity of separating the financing and provision functions, and the existence of more than one type of public-private mix. The typology is used as a way to reach a proposal for a new arrangement for health care systems in Latin America called "structured pluralism."

Gonzalez-Block (1997, p. 201) identifies this approach to comparative analysis as the use of an "ideal type" model, abstracted from the institutional arrangements and processes found in existing health care systems in industrialized countries so as to maximize certain desirable attributes. He also points out some problems in the use of an ideal type, for example, the possibility of oversimplifying existing institutions and actors, as well as the normative aspects implicit in it. "Structured pluralism" is put forward as the best arrangement to achieve efficiency, responsiveness, accountability, and solidarity through an appropriate mix of public and private financing, delivery, and regulation. However, this arrangement tends to conceal the political and ideological aspects by treating any possible arrangement of relations and variables as a matter of technical expertise.

In addition to the health services approach, Ellencweig (1992, p. 22) identified two other kinds of taxonomy used in the comparative analysis of health systems: the health policy approach and the epidemiological model. The politico-economic approach is a policy-oriented model that has been adjusted to distinguish between national health systems based on the assumption that they really do manifest clear-cut ideologies (Ellencweig 1992, p. 29). As distinct from the static health services approach, the politico-economic approach encompasses historical, political and economic processes, although it fails to provide concrete tools for comparing the elements of health care systems.

The epidemiological approach applied to comparative health system analysis works with multiple dimensions, such as the social (covering political, socio-cultural, economic, and demographic elements), the institutional, and the individual. This multilevel analysis assigns causal effects to a variety of factors and is used to study the interrelationships between several effects and their causes (Ellencweig 1992, p. 35).

Despite the importance of multilevel analysis and the historical perspective, it would be difficult to establish direct causal links between different political and economic orientations and the institutional arrangements of health care systems. The same holds true for linkages between the institutional and the individual levels.

To cope with the difficulties involved in implementing a cross-national multidimensional comparative analysis, Ellencweig (1992, p. 36) proposed the construction of a modular approach.

The modular model is made up of a number of separate modules which can be grouped together as needed.... Comprehensive cross?comparisons of systems might be desirable but are often beyond reach. The study of longitudinal trends is similarly limited. However, we can still make several comparisons that can be valuable on a narrower perspective, even when only a few modules are available for comparison. The big advantage of the modular approach is that it can be used for solving problems of various sizes. From the many modules making up the system, appropriate modules may be chosen for each problem to be solved.

The weakness of this approach is due to the inherent difficulty of linking the results found in several modules. Even so, the approach seems useful in comparative analysis, in that it permits different levels of comparison, from the suggestion of analogical association to the establishment of causal links based on evidence.

Methodological Options

For the reasons described earlier, we have adopted a modular approach to study the ongoing health care reform processes in Argentina, Brazil, and Mexico in a comparative manner.

Chapters 2, 4, and 6 of Section II consist of a politico-economic analysis of the macroeconomic and social transformations in the region over the last 20 years and describe the way that these may have affected the institutional arrangement of the health care system.

The theoretical model underlying the design of the study assumes that each society, at different times in its development, assigns priority to certain organizational forms and regulatory mechanisms of the health care system, associating these instruments with values such as equity, freedom of choice, and efficiency.

The intention is not to establish causal links between macro socioeconomic process and health care system reform. The presumption is rather that health care systems are driven by values, institutional dynamics, political choices, technical interventions, popular demand, and pressure from actors with different degrees of control over important resources in the health care arena.

The aim of Chapters 3, 5, and 7 is to analyze the political, institutional, and health environment into which health care reforms were introduced in each of the three countries, using a politico-economic approach. To maintain a historical perspective, we reconstructed the background influences and primary factors that could have affected the launch and political meaning of health care reform processes.

Chapters 3, 5, and 7 also describe the legislative and administrative tools used to effect health care reform, as well as the strategy adopted to implement it. The implementation process itself is the result of powerful dynamics among different actors in the health care arena, representing several vested interests. Usually, the result is a complex process of give and take, sometimes yielding results that differ profoundly from what was originally intended.

The evaluation of the overall process is intended to point out the stresses and contradictions, as well as the differences between rhetoric and reality, in each country. The discussion of dissimilar realities is limited to a careful description of the reform processes founded on a common analytical framework, although the method could allow the identification of similarities and differences in terms of general trends. To overcome the criticism related to the use of a highly conceptual model and the lack of tools for comparing elements making up the health system, we decided to complement these studies with a second comparative module.

Chapters 3, 5, and 7 take a health care systems approach and apply it in each country. To avoid the tendency to adopt a simplistic model based on a one-dimensional typology, we decided to build a comparative framework of health systems, encompassing the main dimensions that are likely to be affected by existing reform processes. In each dimension, we grouped important variables, aiming to characterize the methods of financing, regulation, organization, and provision.

The model adopted in this component of the study includes the reconstruction of such interactions in a historical perspective, focusing on changes in the following:

• the regulatory capacity of the main actors within the health sector;

• prevalent organizational forms characterizing services with regard to ownership, professional profile, and resource allocation;

• sources of financing, risk-sharing options, financing methods, and monetary flows; and

• conditions of provision of services to different population groups, and the results in terms of equity, accessibility, and efficiency.

The static characteristic of the health system approach could be minimized by illustrating the health care system at different points in its development along the reform process. Using the same framework (financing, regulation, organization, and provision), we describe the health care system arrangements in the 1970s and the 1990s, to achieve a comparative picture of each system’s organizational and institutional transformation.

To understand the dynamics of the reform process through the relationships and interactions between the main actors, we adopted the model proposed for OECD study of comparative sanitary reforms (Hurst 1991). It identifies five principal actors in health care systems:

• consumers-patients;

• first-level providers;

• second-level providers;

• insurers or third-parties; and

• government as regulator of the system.

One can analyze the transformations that have occurred in recent years in the health sector through the changes in the relations between the main actors, as described in the schema shown in Figure 1, which is well known in the literature on health care reform.

Figure 1. Internal relations in health care systems.

Using the interactive approach, we will analyze the links between: financiers and providers of health care; the population and financing entities; and patients and providers. Finally, we will refer to the instances and regulating methods operating in the systems analyzed.

In a comparative analysis of several countries implemented by the Organisation for Economic Co-operation and Development, seven basic arrangements were identified (OECD 1992, p. 19-27), according to the flow of patients, the flow of money, and the relationships of authority:

• the voluntary, out-of-pocket model where there is no insurance and health care is based on direct, out-of-pocket, fee-for-service transactions between consumers and primary- or secondary-level providers;

• the voluntary reimbursement model based on the full or partial reimbursement of patients for medical care bills within a private voluntary insurance scheme;

• the public reimbursement model based on the same reimbursement principles as the above, but within a public compulsory insurance scheme;

• the voluntary contract model whereby independent providers supply services to the insured under private contractual agreements, mostly free of charge;

• the public contract model based on a compulsory public insurance where independent providers supply services to the insured under capitation or fee-for-service payments;

• the public integrated model based on a public compulsory insurance as above, but delivered by non-independent providers, usually paid through salaries and prospective budgets; and

• the voluntary integrated model based on private voluntary insurance, delivered by salaried providers with prospective budgets.

This characterization of models of interaction not only allows us to compare the basic design of various systems, but also the dynamics of their transformation with regard to some of the most important dimensions of health policy. It can be expected that whenever the health care system is segmented rather than integrated, as in Latin America, parallel policies, institutions, and systems for different population groups will result.

The possibilities for comparison between the three countries are restricted at the outset by the unique configuration of the health care system in each country. Nevertheless, similar trends may be evident even if they occur in different organizational settings. Our hypothesis is that the reshaping of health care systems over the last 20 years has broadening freedom of choice between providers, rendered coordination more difficult, and aggravated existing inequities with respect to the accessibility, utilization, and quality of services in the health care system.

Another important criticism of the use of the health care system approach is that its national focus limits its ability to discern variations imposed by ongoing tendencies toward decentralization and privatization.

The third module is a quantitative study of the utilization of health care systems in three cities: Rosario, Argentina; Rio de Janeiro, Brazil; and Mexico City, Mexico (Chapter 8). Questions may be raised regarding the ability to draw conclusions about the national health system based on a study of one city. Aside from practical considerations, our justification for adopting this approach is that the selected cities exhibit the main features of the national health care systems of which they are a part.

Service use, measured at the individual level, was the key outcome. The quantitative approach used in this module allowed us to look for causal links between the configuration of the health care system and the utilization of services. However, being cross-sectional in nature, such an approach does not lend itself to attributing causal relationships between variations in service use and reform measures. For this study, we used a theoretical model akin to Andersen and Newman’s behavioral model on health care utilization (Andersen and Newman 1973) and applied it to service use specific to four tracer conditions. A survey was designed to reveal the causal relations between three sets of components of the Andersen and Newman model:

• The predisposing component comprises those factors that existed before the onset of illness: a set of demographic, socio-cultural, cognitive, and attitudinal variables that are postulated as personal factors that identify the potential risk group but cannot be modified through health policies.

• The enabling component consists of the organizational and financial factors that affect the individual’s ability to access a given service.

• The need component expresses the individual’s perceived need for care.

• The predicted relations between these components are as follows:

• Predisposing variables may influence health services utilization through their effect on the enabling component and on the need variable, as well as directly.

• Enabling variables influence utilization directly and also through the need variable.

• The need variable influences utilization directly.

The hypothesis for our quantitative study is that the health care system will be more or less equitable depending on the weight of the enabling component in the utilization of health care services by patients. The higher the weight of the enabling component, the less the utilization of services is based on perceived need or predisposing factors, and hence the more inequitable the system.

The tracer conditions were selected taking into account the epidemiological characteristics of the population under study; the biological characteristics of the tracer condition, determining the form and frequency of utilization of health care resources; the type of attention required; and the discretionary or non-discretionary use of services. The selected tracers were diarrhea, hypertension, pregnancy, and delivery.

The survey used a two-stage, self-weighted, and probabilistic household sample to identify all individuals having at least one tracer condition. The task of designing the research and making the sample comparable came with recurrent methodological challenges. The challenges arose in each stage of the inquiry, including the definition of the variables, the sample design, and the collection, synthesis, and analysis of the data.

The decision to work with a modular design was based on our concern to consider health care reform in as deep and extensive a way possible. Although we are aware of the difficulties in combining quantitative and qualitative approaches, we prefer to take a broad rather than a narrow perspective on reform. The complexity of the object should be no excuse for a partial approach to the reality.

Copyright 2000 © Held by the Authors

info@idrc.ca

Document(s) 3 of 14