The polarized federalism in the health system in Brazil

Abstract

This article proposes a historical-institutionalist perspective to explain the political and institutional dynamics that polarized the intergovernmental relations in health policy in the 1990s. In this perspective, the history of intergovernmental relations is understood as the result of a long chain of decisions made in particular political contexts of the three governments in the 90s: Presidents Fernando Collor, Itamar Franco and Fernando Henrique Cardoso. During this period, the relations between the political actors of the Brazilian Health Unified System (SUS) and the trends of the national federative context established the direction of the decentralization strategy at each stage. The intensity of institutionalization also depended on the choices of these governments regarding the direction of the national policy and the federative coordination structure of the Ministry of Health, the stability and volume of financing, and the institutional capacities of states and municipalities to take on new responsibilities.

Keywords:
decentralization; federalism; health policy; Health Unified System (Brazil); historical institutionalism

Resumo

Este artigo busca explicar a dinâmica político-institucional que produziu um quadro polarizado de relações federativas no SUS nos anos 1990, com base no arcabouço teórico do neoinstitucionalismo histórico. Nessa perspectiva, analisa-se a trajetória das relações intergovernamentais como o resultado de uma cadeia de decisões tomadas em contextos singulares dos governos Collor, Itamar e FHC, onde o jogo entre os atores políticos do SUS e as tendências do contexto federativo nacional estabeleceu a direção da estratégia de descentralização em cada etapa. A intensidade de institucionalização dependeu ainda das escolhas desses governos relativas à direção da política nacional e à estrutura de coordenação federativa do Ministério da Saúde, da estabilidade e do volume do financiamento setorial, e das capacidades institucionais de estados e municípios para assumirem novas responsabilidades.

Palavras-chave:
descentralização; federalismo; política de saúde; Sistema Único de Saúde (Brasil); neoinstitucionalismo histórico

Resumen

Este artículo propone un punto de vista histórico-institucionalista para explicar la polarización de las relaciones federativas en SUS a finales en los años 1990. En este enfoque, la trayectoria de las relaciones intergubernamentales es el resultado de una larga cadena de decisiones tomadas en contextos políticos concretos de los tres gobiernos en los años 1990: Collor, Itamar y FHC. La dinâmica de forças entre los actores políticos del SUS y las tendencias del contexto federativo nacional establecieron la dirección de la estrategia de descentralización en cada etapa. La intensidad de institucionalización resultó también de las elecciones de esos gobiernos relativas a la dirección de la política nacional y a la estructura de coordinación federativa del Ministerio de Salud, de la estabilidad y volumen del financiamiento sectorial, y de las capacidades institucionales de estados y municipios para asumir nuevas responsabilidades.

Palabras clave:
descentralización; federalismo; políticas de salud; Sistema Unificado de Salud (Brasil); neoinstitucionalismo histórico.

1. Introduction

The 1988 Constitution and the Organic Health Laws established a tripartite cooperative federalism, with a regionalized design for the Brazilian Health Unified System (SUS). However, the complex process of decentralization led to different forms of implementation from what was originally proposed.

Throughout the 1990s, the federal government enforced its power by acting with significant regulatory and financial capacity, imposing decontextualized models and patterns of policies and programs, induced through intergovernmental orders and transfers. The process of transferring roles and resources essentially focused on municipalities. The states, played a reduced role in the direct provision of services and presented difficulties in conducting regional planning and coordination. This dynamic introduced a polarization of federative relations in SUS, characterized by direct relations between the federal government and municipalities.

The literature that analyzes the history of SUS’ decentralization, emphasizes a) the political and economic context; b) the correlation of sectoral forces; or c) the configuration of basic operational norms, as the main explanatory factors. When considered separately, these three approaches lead to only partial explanation for polarization (Ouverney, 2014).

This article aims to present an Institutionalist Historical explanation for the polarization of federative relations, using an analytical framework built specifically for this purpose, based on the concepts of critical juncture and path dependence. Such explanation connects the three perspectives of the sectoral literature in a sequential way in the governments of the Brazilian Presidents Fernando Collor de Mello, Itamar Franco and Fernando Henrique Cardoso. This article discusses whether the political actors of SUS and the trends of the national context established the direction of the decentralization strategy at each stage. It also considers the intensity of institutionalization depending on the choices of governments regarding the direction of the national policy and the federative coordination structure of the Ministry of Health, the levels of stability and volume of financing, and the institutional capacities of states and municipalities to take on new responsibilities.

2. Institutional change: critical junctures and path dependence

Critical junctures can be defined as periods of expectation for significant changes in society, in a specific institution or policy, as a consequence of the erosion of current arrangements, which may have become illegitimate or inefficient.

They can be triggered by both external political and economic crises and by expressive changes in the internal correlation of forces. This happens when new actors emerge, or already established actors broaden their political space. Both of them seek to introduce changes to the rules of accessing and exercizing power; to create a new agreement; and to project their legacy (Soifer, 2012; Capoccia and Kelemen, 2007).

At a critical juncture there are more possible directions and the policy can go in several ways at once. This occurs as the choices of the main actors involved acquire greater capacity of changing the direction of the policy (Mahoney, 2001; Mahoney and Thelen, 2010; Capoccia and Kelemen, 2007).

Therefore, the fundamental characteristic of a critical juncture is the emergence of alternative institutional paths. Confirming one direction as the hegemonic path for change will trigger the development and implementation of institutional mechanisms of a new order, leading to a lock in situation for the other paths. Thus, the end of a critical juncture is marked by the stability of the innovations introduced, even though contradictions may appear in new forms (Mahoney and Thelen, 2010).

The production of the legacy continues to unfold through sectoral political disputes marked by the interaction between reactive and self-sustaining political-institutional sequences (Mahoney, 2001; Broschek, 2011; Pierson, 2000b:2006).

Self-sustaining sequences occur when initial steps in one direction induce future movements in the same direction, that is, they are interconnected actions where the probability of a future event grows as the chain of events is followed. This tendency is produced because of increasing returns i.e. as a sequence of events progresses, the costs of exit and strategy change involved increase over time (Mahoney and Thelen, 2010; Pierson, 2000b; Mahoney, 2000).

Reactive sequences are political reactions that produce negative feedback regarding the new institutional standard being established. They can be driven by both the resistance of important actors and by adjustments or contradictions in the ongoing innovations (Mahoney, 2001; Pierson, 2000a; Mahoney, 2000). This movement of self-sustaining and reactive sequences produces a path dependence on the initial events triggered in the critical junctures.

Systems of intergovernmental relations aimed at public policy management, especially in areas such as health, are characterized by expressive levels of resilience. They involve change in constitutional milestones that define the division of attributions between the levels of government, the reallocation of significant amounts of resources, the implementation of administrative, financial and operational structures, the training of specialized technical staff, among others.

Changing these parameters involves qualified voting in the legislative. It also involves complex and difficult negotiations among very significant players on the national scene, such as parties, bureaucracies and trade unions, regional and local governments, suppliers, equipment and services, among others.

The literature on institutional change in federative systems reflects this resilience by showing that choices made in the initial moments of the formation of the nation states establish very long path dependence only altered by very expressive critical junctures triggered by events such as wars, deep economic crises and regime changes (Obinger, Castles and Leibfried, 2005; Benz and Brochek, 2013; Broschek, 2011, 2012; Béland and Myles, 2012).

This is the case of the intergovernmental relations that underlie Brazil’s health policy. The pattern of centralized power, financial resources and expertise at the federal level established in the Vargas Era (government of President Getulio Vargas from 1930 to 1945), and deepened with the military governments, was more heavily challenged when confronted with the international economic crisis of the 1970s and with the emergence of a decentralizing project. The project was supported by the articulation of important actors such as the MRS, governors, mayors, state and local health care managers, part of the federal bureaucracy, social movements and unions.

This moment inaugurated a critical juncture that culminated in the 8th National Health Conference in 1986 and the inclusion of SUS in the Constitution in 1988. However, during the decentralization strategies of the Integrated Actions in Health (AIS) and of the Unified and Decentralized Systems of Health in the States (Suds) both in the 1980s, the federal dilemma remained open, and in the early 1990s the direction of decentralization was not set allowing the possibility of different configurations.

3. Research methodological strategy

3.1 Alternative paths for decentralization

The analysis of the political-institutional dynamic that led to the federative polarization of SUS in the 1990s began by identifying the possible paths for the decentralization process.

An alternative would be to structure the health system based on the deconcentration in the same model as adopted by the National Social Security Institute for Medical Assistance (Inamps), built on the closer partnership and cooperation between the regional offices of Inamps and the state and municipal health secretariats, while maintaining the main functions under federal responsibility. This is similar to what has been implemented in AIS. This path was preferred by private providers and the Inamps bureaucracy, which had an interest in securing their acquired positions to the maximum (Cordeiro, 2004).

Another alternative would be the construction of SUS by the state, based on the political, managerial and financial boost offered by the experiences established within the Suds. The state secretariats would be responsible for implementing the entire decentralization process. Those most interested in this model were the state secretaries and part of the health movement (Paim, 1990).

A third possibility would be regionalization in the form embodied in the constitutional text and in the Organic Health Laws. The Union would focus on strategic policies, states in regional planning and coordination, and municipalities would be in charge of programming, managing, financing, and regulating the health care system in their respective territories. This has been the major project of the health movement (Fleury, 2003).

Finally, the fourth option would be to expand the participation of municipalities, maintaining the expressive role of the Union by inducing strategies at the national level, with less participation of the states in the provision of services, reducing their coordination and role of regional planning. This option was not part of the actors’ agenda at that time, but it ended up becoming the route that was implemented, which requires a more comprehensive explanation than those found in literature.

3.2 Model of analysis: connecting the decentralization explanatory factors

The analysis of the political-institutional dynamics of decentralization was based on a sequential model, in which choices made in certain moments influence the future development around specific solutions, blocking the federative sectoral relations around these solutions. During this period, sectoral disputes led to a reformulation of intergovernmental relations, and the initial situation of intense competition gave way to a more stable political-institutional configuration. This process was developed in three moments corresponding to the governments of Presidents Fernando Collor, Itamar Franco and Fernando Henrique Cardoso.

This article advocates that, at each stage, the convergence or divergence in the game between the political actors of SUS and the trends of the national federative context, in political and economic terms, established the guidelines for the sectoral reform process in terms of the decentralization path to be adopted. Thus, the end of a critical juncture would occur when the configuration of the sectoral political game and the trends of the national federative context allow a convergence around a specific path of decentralization. Such convergence should be intense enough to install an office of the Ministry of Health capable of aligning federal policies and organizational structures of federative coordination with this path. In addition, the Ministry of Health has to be able to provide stability and sufficient financial resources for the strategy, as well as expand the institutional capacities of states to absorb new responsibilities. These latter factors give greater intensity to the path of decentralization chosen and are essential to increase the degree of institutionalization of its legacy. This dynamic is represented in figure 1.

The data for this research were collected from 2012 to 2014 through 18 interviews with key actors of the health policy at that time, including Ministries of Health, Secretaries of the Secretariat of Health Care (SAS-MS) and the Executive Secretariat (SE-MS), leaders of National Council of State Secretaries of Health (Conass) and National Council of Municipal Secretaries of Health (Conasems) and leaders from entities promoting health reform, such as the Brazilian Center on Health Studies (Cebes) and the Brazilian Association of Collective Health (Abrasco). In addition, extensive research was conducted using documents from the Brazilian Federal Senate, Chamber of Deputies, Presidency of the Republic, Ministries of Health, Finance and Planning, Fiocruz Foundation and Cebes, as well as an extensive and systematic review of the national literature on political and economic reforms and on the decentralization of the health policy in the 1990s.

Figure 1
Model of analysis of the SUS’ decentralization path in the 1990s

4. Results: from controlled municipalization to federative polarization

4.1 Controlled municipalization (1990-1992): an unsuccessful attempt to strengthen Union and the beginning of federative polarization

4.1.1 Federative context: impasses and absence of a National Project of Decentralization

Since the beginning of the 1980s the game of forces between the Union and subnational entities has favored the latter, due to the simultaneity of the crisis of the national developmentalism of the Vargas Era and the military regime; the redistribution of the fiscal burden to states and municipalities; and to the centrality of the elections in Brazil’s Nova República (period started in 1985, after the military regime). These movements weakened the presidency in the political, economic and financial fields (Abrucio, 1998).

However, the co-operative behavior of the President of the Republic was fundamental to enable decentralization, because of their power of veto over sectoral legislation that regulated the still unapproved 1988 Constitution, and because of the administrative capacity of the federal government, which was an outcome of a legacy of centralization (Arretche, 2012). President Collor never adopted a position favorable to decentralization, using the available instruments to create a negative agenda in the social area and delay the decentralization process.

The state governments, despite having sufficient political strength to avoid refinancing the state public debt defined by the federal government through Law no. 8.388/91, remained with great difficulties to take on other responsibilities in terms of public policies due to their significant fiscal crisis. The 1990 elections also resulted in few governors allied with Collor, even in the more developed states, creating a deadlock between the two levels of government. In this context, the approach to municipalities became a more viable option for the federal government, since municipalities had smaller fiscal problems and were eager to expand their space in policy management, but they wanted more autonomy than Collor’s government was willing to grant. This government relegated decentralization to the background, leaving it confined to a context of federative conflict.

4.1.2 Game of forces between the actors: strengthening traditional actors vs defending SUS’ constitutional prerogatives

In the health sector, as a consequence of the centralizing preferences of President Collor, there was strengthening of actors opposed to sanitary health, especially the federal bureaucracy and contracted private providers (Silva, 2001). The federal bureaucracy, affected by the recent transference of Inamps from the Ministry of Social Security and Social Assistance (MPAS) structure to the Ministry of Health, kept important power resources, such as contract management with the private network, technical expertise for decentralization and the institutional space of Inamps (Wayland, 1995). The private sector had larger losses with the advance of decentralization in SUS, because it was interested in re-articulating a recentralization strategy to support Inamps as the only channel of contact to implement contracts (Cordeiro, 2004).

This movement contributed to the survival of centralizing forces, especially since the coalition, formed to promote reforms, had been weakened with the departure of its main leaders from important positions during President José Sarney’s government. However, the main entities of the Brazilian Sanitary Reform Movement (MRS) (Cebes and Abrasco in particular) were articulators of the sectoral reform project, still had in their favour the articulations with the state and municipal managers, as well as the legitimacy and sufficient political resources to defend the legacy of the reform, especially the new constitutional framework. This made it possible to denounce the illegal aspect of the regulations issued by the government. This balance of forces created an impasse in the health sector.

The creation of Conasems in 1988 led municipal managers to an important role in the reformist coalition, while the state secretaries of health, due to Sarney government’s restrictive financial policies on SES, had reduced their capacity to carry out the process of decentralization (Arretche, 2000; Silva, 2001; Souza, 2005). This internal change in the reformist coalition has placed municipal managers on the “front” of the decentralization process, but in the face of a situation not conducive to the decentralization project, forcing them to evaluate a temporary adaptive strategy.

4.1.3 Direction of the national policy in health: insulation of the Ministry of Health

In this context, the direction of the Ministry of Health sought to avoid the pressure of reformist actors through the insulation of health policies, draining the political debate with the MRS, including postponing the 9th National Health Conference (Escorel and Block, 2005).

The Alceni Guerra administration at the Ministry of Health mirrored the federal government’s restrictive orientation towards decentralization, since the Minister was part of Inamps staff and was elected a deputy from parties that opposed the creation of SUS (PDS and PFL) (Cordeiro, 2001). Guerra saw decentralization as a slow process and regulated by the rules defined by the Ministry of Health, since, according to President Collor’s government project, states and municipalities were not prepared to autonomously assume the management of their health systems (Guerra, 1990).

On the legislative front, several provisions of the text of the Organic Health Law (Law 8080/90), which could directly or indirectly expand the role of states and municipalities in the implementation of SUS, were vetoed by President Collor. These included: the institution of conferences and health councils, the definition of percentages and automatic and regular flows of transfer of federal resources, a job and salary plan for SUS staff, among others. Some, but not all, of these vetoed points were recovered later in Law 8.142/90. In short, the Collor Government ignored the impasse with the reformist forces and tried to articulate a strategy for recentralization of health policies, in opposition to the entire sectoral process of the 1980s.

An important part of this strategy was keeping Inamps as a central organizational unit in the management of care and financial resources, management of professionals, contracts with providers, etc. The modus operandi of Inamps prevailed in the early years of SUS, especially in the design of the architecture of resource transfers to subnational entities, because it was the agency that held strategic resources accumulated over several years of sectoral activity (Carvalho, 2002).

4.1.4 The decentralization strategy: deconcentration of the model based on Inamps

The result of this orientation was the deconcentration strategy of the model based on Inamps, whose guidelines are contained in the Basic Operational Norms (BON) 01/1991. The model is considered by MRS as a step backwards in relation to Suds, since it considered the states and municipalities as mere service providers, hired by Inamps, without any autonomous management.

BON 91 established a dynamic of direct relationship between the federal government and the municipalities, without considering mediation of the states or the establishment of political and managerial instances that could implement consistent processes of agreement and integration among the actions of the three levels of government.

The guidelines showed the option for the combined use of agreements signed directly between Inamps and each of the states and municipalities. In addition, there was the option to use information systems to monitor and control the production of services provided as well as financial incentives linked to the fulfilment of a set of prerogatives by subnational entities.

The basis of the federal financial transfers model adopted was the direct payment for the production to the service provider, with different criteria of allocation of resources for outpatient coverage, hospital care, investment in the existing network and promotion of municipalization (Lima, 2007).

The new rules were unattractive to the states as their central role was removed and the management and funding structure of the Suds — organized in the State Secretary of Health — was dismantled.

For the municipalities, which were expanding their importance in the federation, the strategy proposed in BON 1991 represented the possibility of obtaining new financial resources from the federal transfers. The norm also represented an opportunity for affirmation — in the health sector — towards the states that were, until then, in a privileged position within Suds.

On the other hand, the proposal of BON 91, because it contained expressive levels of control of the federal government, reduced the incentives for adherence of municipalities that wanted a more expressive insertion in SUS. Likewise, other factors that could increase the chances of success of the strategy, and increase its potential for institutionalization, were not properly taken into account by President Collor’s government.

4.1.5 Financing, structures of federative coordination and institutional capacity of subnational entities: low potential for institutionalization and permanence of critical juncture

In addition to the strategy proposed by BON 91 to be against the general tendencies of Brazilian federalism, the relationship between the Ministry of Health and MRS was conflicting most of the time due to the incompatibility of projects. MRS was able to find spaces to defend SUS, especially the National Health Council and parliamentary groups in the National Congress. Thus, neither the national federative context nor the sectoral game of political forces favored a strategy of deconcentration of the model based on Inamps. The direction taken by President Collor’s government was incompatible either with the national scenario and the scenario of the health sector.

The administration of Minister Alceni Guerra was also characterized as a period of high instability and irregularity in the flow of resources for health, in addition to the reduction of the amounts allocated by the federal government. This dynamic indicated to states and municipalities that they had to take on responsibilities without any confirmation regarding financial transfers from the Union.

Likewise, there was no significant reform of the federal administrative structure aimed at preparing for a “controlled deconcentration” movement. During this period, there was only the maintenance of inamps, the accomplishment of some specific changes, such as the establishment of the National Health Care Secretariat (Snas); the merger of the Foundation of Public Health Services (FSESP), the Superintendency of Public Health Campaigns (Sucam) and the Company of Technology and Information of Social Security (Dataprev) under a new unit, the National Foundation on Health (Funasa); and the disorderly transfer of personnel and health units to states and municipalities (Machado, 2007; Costa, 2011).

Thus, the attempt to implement a “controlled deconcentration” of the health policy during President Collor’s government was unsuccessful, which was evident from the low adherence of municipalities: 321 in 1991, 565 in 1992 and 188 in 1993, a total of 1,074 municipalities at the end of 1993, or only 21.6% of the municipalities in the country.

Therefore, to the extent that the decentralization proposed in BON 91 (deconcentration inspired in the model of Inamps) did not obtain enough adhesion to become the hegemonic path, the federative critical juncture remained open. Thus, other paths were still possible, but the dynamics of the Collor period altered the relative probabilities between the available paths. The path of giving more responsibilities for the states was the most impaired one in this period, because together with the continuous fiscal crisis of the states, the State Secretaries of Health (SES), which were already affected by President Sarney’s government’s restrictive financial policies, saw the structures of Suds dismantled by the logic of direct relationship between the MS and the municipalities. The same can be said about regionalization, which requires a more expressive role of states in relation to municipalities. A format organized from larger municipalities would also be unlikely, considering the lack of priority given to policies aimed at the organization of metropolitan regions and to specific provisions in legislation towards the health sector. The municipalization was in a better position, but greater autonomy would be necessary for Municipal Secretaries of Health, to set up structures for federative coordination and to increase transfer of resources, as well as to break the resistance of the central actors reinforced by President Collor’s government. In this sense, the advocates of stronger municipalities took advantage of Collor’s conflicts with the state governors to expand their influence in health policy, even if it was with worse conditions than desired, hoping for a more favorable situation in the future.

4.2 “Courage to follow and to reinforce the law” (1993-94): the time for municipalism

4.2.1 The federative scenario: window of opportunity for municipalism

The context expected by the municipalities occurred between President Fernando Collor de Mello’s impeachment at the end of 1992 and the approval of the Social Emergency Fund (FSE) in March 1994, when the Union regained its strength in the federation. During this period, there was a window of opportunity in the national federative context characterized by the presence of a president who had no preferences for centralization, a timid legislative agenda in the federative reforms, the expansion of available revenues to subnational entities (especially municipalities), and low federal regulation in the field of public finances.

President Itamar Franco maintained close ties with the reformist actors, supporting legislation in the social field (in particular, the Organic Law on Social Assistance — Loas), as well as promoting important programs in the areas of eradication of child labor, generic drugs, family health. In his government, the National Council of Food Security (Consea) was created.

President Itamar Franco’s government also interfered little in federal relations. Out of 59 bills dealing with federative interests discussed in the Congress between 1989 and 2006, only 06 were initiative of his government (Arretche, 2012).

Likewise, in 1993, the tendency to increase subnational entities’ revenues from taxes was maintained. Between 1980 and 1993, states and municipalities together increased their participation of available revenues from taxes from 31.9% to 41.2%, municipalities alone increased participation from 8.6% to 15.8%. In addition, the approval of Law 8.727, in November 1993, defined new rules only for debts contracted with federal institutions, not including the debt of states and municipalities with private institutions, securities debt, and operations of revenues anticipation, which interfered little in subnational public finances.

In the perception of the reformist forces, this context was a rare opportunity to promote decentralization.

4.2.2 Game of forces among the actors: the political consolidation of municipal administrations and the reborn of the decentralization ideal of the sanitary reform

President Collor’s impeachment and civil society’s pressure for change weakened the more conservative sectors aligned with the president’s centralizing project, in particular the private sector and the bureaucracy of Inamps. This resulted in a shift in the correlation of forces in the health sector, strengthening the sanitary movement, and state and municipal administrations. A coalition was formed, capable of breaking the insulation promoted by the federal government towards the project of decentralization of health policies. The coalition was driven by the urgency to promote decentralization and avoid future setbacks, due to the international context of financial crisis of the Brazilian State that favored liberal-like reforms (Melo, 2005; Arretche, 2012).

The 9^th National Health Conference allowed opening a political path similar to that of the 8^th conference, where the expansion of the debate on the field of health with civil society actors, popular movements, and state and municipal administrations, triggered a movement of political articulation. This allowed the resumption of decision-making spaces on health policies at the national level, especially under the guidance of the Ministry of Health (Cebes, 1992).

The 9th National Health Conference represented a moment of consolidation of the political role of the municipal secretaries, an event that indicates the position they would occupy in the direction of the new administration of the Ministry of Health and in the definition of the new SUS decentralization strategy (Faleiros et al., 2006; Cebes, 1992; Ribeiro, 1997).

4.2.3 Direction of the national health policy: the time for municipalism

In this window of opportunity, the approach of municipalism was leading the Ministry of Health, using the central power to avoid future movements towards centralization. The new leaders at the Ministry of Health, for the most part, had served in local governments, had worked at the federal administration, and had close relations with actors of the Brazilian Sanitary Reform Movement (MRS). Consequently, the elaboration of BON 93 was characterized by a process of opening the relationship between the Ministry of Health and the main political forces in the sector. The main evidence of this new relationship was the creation of the Special Group for Decentralization (GED), the extinction of Inamps and the implementation of the tripartite inter-managerial commissions (CIT) and the bipartite inter-managerial commissions (Cibs).

The GED acted as a technical and political group for articulating a new proposal of decentralization to replace BON 91, in response to the demands of the 9th National Health Conference and of various reformist actors (Silva, 2001). Conducted by municipalists, the group brought together a broad set of actors who had interests related to the planned changes, showing that the government intended to incorporate the diversity of demands to produce a broad consensus on municipalization (Carvalho, 2002).

During this window of opportunity, the government sent a bill to Congress to extinguish Inamps, which was heavily opposed by parliamentarians allied with the agency’s bureaucracy, the private sector and the national media. The mobilization of the sector, the strong decision of the leadership of the Ministry of Health and the initiatives of disarticulation of the opposite sectors were decisive for the approval of the bill. Among these initiatives, it is important to mention some concessions to sectors opposed to the extinction of the agency, such as the creation of a career in the National Audit System, meeting the demands of the Inamps auditors, and maintaining a fee for service payments to meet the interest of the private providers.

Finally, the implementation of bipartite commissions (CIBs) was an essential condition for the progress of decentralization, since both the programming of actions and health services as well as the decisions related to the resolution of conflicts between the municipalities and of the municipalities with the states would end up being defined in these instances of agreement. In addition, the CIBs would also be in charge of making decentralization adequate for the diversity in regional and local realities.

4.2.4 The decentralization strategy: establishing power sharing and autonomy of local management

The publication of BON 93 completes this aggressive strategy of decentralization, which stated that the entire process of defining operational guidelines for SUS implementation would be conducted through federative negotiations in the inter-managerial commissions (CIT and CIB) (Carvalho, 2001).

The functions to be performed by states and municipalities were also significantly expanded when compared to the BON 1991. The functions depended on the complexity of the local and regional service structure that allowed the state or municipality to qualify for the different types of participation. The types of participation defined for municipalities were the incipient, partial and semi-full, and the degree of autonomy in local management conditioned to the institutional capacities and the size of the service network.

The coordination of the decentralization process was carried out through a variety of instruments of political, financial, legal and managerial nature, expanding the possibilities for federative coordination available to the Union. BON 93 strategy articulated instances of agreement (CIT and CIB), contracts of participation based on managerial conditions, information systems to monitor and control the production of services, reports on management of the responsibilities taken per management condition and financial incentives linked to the fulfillment of a set of specific requirements per each management condition.

With BON 93, subnational entities began to carry out activities of planning, programming and management and evaluation of their own networks, as well as the networks contracted. This led to the change in the way the payments by federal transfers were made. Federal transfers had been made directly to service providers, with BON 93 the transfers started to be made from fund to fund, i.e. from the National Health Fund to the State and Municipal Health Funds. This change was designated only for municipalities adopting semi-full management (in terms of type of participation) and only from August 1994 when the mechanism was regulated. BON maintained the format of fractionating the transfers to outpatient care and hospital care, the incentives for state and municipalities to take on responsibilities for the system and the format of fractionating the resources for investments.

4.2.5 Federative coordination structures, institutional capacity of subnational institutions, and financing: high potential of institutionalization and closing of the critical juncture

The direction of the decentralization process defined by the BON 93 strategy reflected the trends of the federative context and met the expectations of the coalition for reform in the health sector. In addition, throughout the Administration of Jamil Haddad in the Ministry of Health (President Itamar Franco’s government), the federal leadership of the ministry was able to make changes in its organizational structure, changes aligned with the decentralization strategy and promoting a transition between the monolithic structure of Inamps and a more diversified framework composed of specialized structures and more integrated with a single leadership body.

Among the changes aligned with the decentralization strategy are the transfer of INAMPS’ responsibilities to the National Secretariat for Health Care (Snas), the choice for the National Health Fund as the main federal unit of transfer of intergovernmental resources, and the creation of the National Audit Service (SNA) and of the Ombudsman’s Office. Although no planned administrative reform took place, these measures were functional and allowed the federal structure to be adapted to promote the most adequate federal coordination for the decentralized SUS.

Unlike BON 91, the strategy designed in BON 93 took advantage of the institutional capacities of states and, especially, of municipalities, granting greater autonomy to the management of resources. States and municipalities had been expanding their institutional capacity to manage health service networks since the early 1980s.

According to the Survey of Medical Assistance (AMS) of IBGE for 1992, out of a total of 49,676 health facilities in the country, 27,092 (54.5%) were public and 22,584 (45.5%) were private. In the public system, 25,750 (51.7%) were units owned by subnational entities, of which 18,662 were municipal (37.6%) and 7,043 owned by states (14.2%). Only 1,387 (2.8%) were federal units.

This pattern of distribution meant both the potential of subnational entities to take on new responsibilities. The different modality of qualification allowed the participation of municipalities with a variety of levels of installed capacity, adapting the strategy to the Brazilian regional asymmetries.

Finally, the only factor that reduced the possibilities for subnational entities to join was the instability of federal funding. In 1993 the Minister of Social Security, Antonio Brito, discontinued the deposit of resources from the contribution on payroll, a source of funding established by article 195 of the Federal Constitution of 1988 for the financing of all social security areas (Brazil, 1988). This attitude triggered the largest crisis in health sector financing in the period of SUS, since contributions from this source accounted for 55% of the Ministry of Health’s budget in 1992 (Carros, Piola and Vianna, 1996; Carvalho, 2002).

Emergency solutions such as a loan from the Workers’ Assistance Fund (FAT), the support from National Treasury Bonds and the increase in resources from traditional sources of the Ministry of Health’s budget, were not able to reduce the insecurity generated by the irregularity of federal transfer flows (Barros, Piola and Vianna, 1996).

Despite the difficulties posed by financial instability, the BON 93 strategy was well received by the municipalities, leading to a high level of participation, when compared to the numbers obtained by BON 91. In 1995, there were already 2,799 municipalities qualified, which corresponded to 56.3% of Brazilian municipalities. The high participation established a sectoral federative lock-in around the municipalization, making it the definitive route for decentralization of SUS and ended the federative critical juncture that began in the late 1970s.

Once a significant number of municipalities took over the management of their health care systems and started to receive resources directly from the national fund to the municipal funds, the political and economic costs of reversing this process were enormous for any political elite that would take over the administration of the Ministry of Health.

From then on, municipalization has been the basis for the implementation of SUS. Versions of municipalization a little more regulated from the federal or state levels of government were still possible, but without significantly interfering in the role of municipalities. The municipal administrations had in their favor the rules of BON 93, the inter-managerial commissions and the control of a wide network of health services, in addition to a significant volume of financial transfers.

The deconcentration of the model based on Inamps would no longer be feasible, and transferring the system strictly to states would be a remote probability. States continued to be indebted, which encouraged fiscal warfare practices, in addition to the fact that State Secretaries of Health were significantly affected by President José Sarney’s restrictive financial policies, and the BON 91, which reduced the state’s role in SUS. The participation of the states in BON 93 was low, with 09 states habilitated until 1995. Considering that expanding the state’s importance would demand a more expressive role in offering more complex services of SUS (which would require substantial financial investment), it was unlikely the states would become the main actor during this period.

4.3 Centripetal and centrifugal trends shape a polarized federalism (1995-2000)

4.3.1 The national federative scenario: all power to the Union

In the second half of the 1990s, there was a centralizing inflection in the federative dynamics in Brazil with movements in terms of legislation, fiscal/financial and of public policy coordination. These movements were characterized by the strengthening of the Presidency in relation to the National Congress; a new national public finances regime; and by the use of federative coordination mechanisms.

Since 1995, the Union has extended legislation that affected the interests of states and municipalities, impacting on various aspects of intergovernmental relations. Such changes in the national legislation resulted in: (1) increasing of non-shareable revenues with states and municipalities (e.g. Social Contribution on Net Income — CSLL, Contribution for the Financing of Social Security — Cofins, Provisional Contribution on Financial Transactions — CPMF); (2) increasing of the retention of resources in the Union (e.g. Emergency Social Fund / Fiscal Stabilization Fund / Untying of Union Resources — DRU); and (3) reducing the autonomy of subnational entities to collect taxes (e.g. Kandir Act, ISS on tolls, fees on public lighting), to manage costs (e.g. Law of Concessions, Guidelines and Framework Law), and to increase expenses (e.g. Law Camata, social security regimes) (Arretche, 2012).

The new public finance regime included: (1) federalization of states’ debts; (2) maintenance of the DRU (formerly called the Social Fund for Emergency and Fiscal Stabilization); (3) increased available revenues for the Union through the expansion of social contributions (Cofins, CSLL, etc.); and (4) the approval of the Fiscal Responsibility Law (Complementary Law 101/00).

Finally, the establishment of a set of federative policy coordination mechanisms has sought to introduce a vertical coordination relationship in the implementation of social programs, such as the Payment Floor for Primary Health Care (PAB) and Fund for Maintenance and Development of Basic Education (Fundef), which conditioned the receipt of resources to meet goals and standards. These changes sought to implement detailed national standards to allow federal government to regulate the role to be played by states and municipalities in the management of public policies (Fleury, 2001). This new federal context generated conflicts with the sectoral process in which municipalization was already institutionalized, producing both pressure to expand federal regulation and a sense of urgency to consolidate municipalization.

4.3.2 Game of forces among the actors: deepening the municipalist legacy versus adjustments and strategic adaptation

In this space of contradiction, the different actors reoriented their political strategies around three positions: (1) to defend and deepen the initial format of municipalization; (2) to direct municipalization to a format more suited to their interest; or (3) to adapt strategically and preserve the areas where they obtain gains.

The municipal secretaries turned their efforts to ensure the implementation of the municipal model, focusing on issues such as the consolidation of shared governance (CIT and CIBs), the debureaucratization of national rules for registering in the system, the extension of managerial autonomy and the creation of a stable source of funding (Carvalho, 2002).

The entities of the Health Reform Movement — Cebes and Abrasco — established two articulated lines of political discourse: they defended the progress achieved with the decentralization strategy implemented by the Itamar Franco Government and denounced the negative impacts of the neoliberal reforms initiated by the Cardoso Government. State administration took on a greater role, with the creation of Conass technical chambers, and organizing thematic workshops with responses to pressures for greater regional coordination resulting from municipalization.

The federal bureaucracy expanded its participation in the game of federative relations due to the high weight of federal financial transfers in financing the sector; to diversification and institutional specialization of the structure of the Ministry of Health; and to the complexity of decentralization strategies.

Finally, private providers adopted a strategy of selective adaptation, which consisted in defending the privileged position in the provision of services and fee for service payment, in addition to obtaining constant monetary adjustment in SUS payments. In the contracts with the state and municipal secretariats, private providers focus their offer on more specialized services.

4.3.3 The national policy direction: a crisis and two styles of federative relationships

The direction of the national policy in this period was characterized by two styles of relationships produced by the acute crisis of the health sector financing. In the years 1995 and 1996, the Ministry of Health was led by technicians who had built their careers with closer ties with the MRS, they were leaders in the sector, and their professional history was connected to the constitution of SUS. This profile favored a greater proximity to the discourse and strategies of sectoral decentralization and allowed for a more expressive resistance to the pressures for strengthening the federal capacity to influence the system. Although, during this time, initiatives were also developed in order to give greater systemic rationality to the process of decentralization in progress, strengthening the role of the Ministry of Health.

After Adib Jatene left the Ministry of Health, motivated by a dispute with the Ministry of Finance about the destination of resources collected through a specific tax (CPMF), President Fernando Henrique Cardoso chose a profile with greater political alignment with the presidency and the economic area. In the following period from 1997 to 2000, ministers with little or no involvement in the health movement were in charge of the ministry, and therefore were outsiders in relation to the political dynamics of SUS, supported by the nucleus of the government. During this period, preferences prevailed for a stronger Ministry of Health and greater capacity for regulation of subnational entities. The differences between these two moments can be identified as follows: (1) in the process of finding a stable source of funding, (2) in the strategy of linking fiscal resources of the three spheres for financing SUS, and (3) in the strategy adopted in the process of preparing and revising BON 96.

The differences in political guidelines in these two periods were initially reflected in the debate around the funding of the sector. In the period 1995-1996, Minister Jatene personally led a political campaign with the main actors in the sector and deputies in order to approve a new tax (CPMF) to increase federal resources to the Ministry of Health, with no commitment of revenues from subnational entities. After the approval of the tax, the use of part of the resources from CPMF for other purposes by the Ministry of Finance put Jatene in conflict with the economic team, leading to his resignation. Under José Serra administration, the underfinancing issue was faced in a different way, leading to the approval of Constitutional Amendment No. 29/00, which designated 12% of states’ revenues and 15% of municipalities’ to health. Thus, the greatest burden was placed on subnational entities, allowing the Union only to maintain the volume of resources already invested in 1999, plus annual adjustment for nominal GDP variation.

In the design of the new decentralization strategy, expressed in BON 96, two different forms of political articulation were observed. Until 1996, the process was more open, starting at the CIBs and ending at the National Health Council and at the 10th National Health Conference, involving the participation of different actors. Under the command of Barjas Negri, as Executive Secretary of the Ministry of Health (in 1997), the Ministry delayed the beginning of the qualifications until the first half of 1998, proposing a more closed agenda for re-discussing BON 96.

In order to give a more rationalizing and inductive feature to federal intergovernmental financial transfers, the Ministry of Health created an operational reorganization agenda that included reviewing codes, reclassifying sets of procedures, revising the criteria for the Integrated and Agreed Programming (PPI), as well as creating new systematic of control and evaluation, among others.

These measures generated a growing conflict between the Ministry and Conass and Conasems, which was accentuated by the publication of a set of orders on 18 December 1997 with the new foundations for BON 96, questioned by the different actors who participated in the process (Carvalho, 2002).

4.3.4 The decentralization strategy: reaffirming municipalization and increasing the Union’s regulatory capacity

The arrangement of intergovernmental relations included in the text of Basic Operational Norms 96 was based on a municipalist federalism, but accompanied by a significant increase of the normative and financial regulatory prerogatives of the Ministry of Health. The model of political negotiations between the federal entities in the inter-managerial commissions was maintained and legitimized. In addition, the arrangement sought to outline a clearer distribution of competencies for each entity of the federation, giving greater systemic rationality to the intergovernmental relations in SUS.

The municipalities participation was based on the level of care offered by the two modalities: Full Administration of Primary Care (GPAB) and Full Administration of the Municipal Health System (GPSM). The same was established for the states in the modalities: Advanced Administration of the State Health System and Full Administration of the State Health System. The Integrated and Agreed Programming (PPI) was introduced in order to allow the state administration, together with the municipalities, to integrate and harmonize the annual schedules for examinations, procedures and hospitalization to be shared among the municipalities.

Primary health care was strengthened by the definition of amounts of investment per capita that were adjusted according to the development indicators of each region. Another measure to strengthen primary health care was the adoption of specific programs promoted by the Ministry of Health through fragmented federal financial transfers (fixed and variable Payment Floor for Primary Health Care — PAB) that significantly regulated the way in which the states and municipalities applied the resources.

The Fund for Strategic Actions and Compensation (Faec) was created in 1999 to cover costs of highly complex procedures in patients with interstate referral and those resulting from the implementation of actions considered strategic. Since 2001, payments to service providers covered with Faec resources have been made directly by the Ministry of Health, centralizing the relationship with suppliers in the federal government.

The increase in the importance of the federal government occur also by using a number of prerogatives and devices, in particular: (1) the maintenance of expressive control over the whole process of elaboration, revision and implementation of BON 96; (2) high power over the definition of agendas and over the dynamics of the negotiation at the Tripartite Inter-managerial Commissions (CIT); (3) the power to decide over the registrations; (4) the control on the accomplishment of the requirements by states and municipalities; (5) and the systematic and increasing use of financial incentives linked to specific policies and programs (Family Health Program — PSF, Community Health Agents Program — Pacs, Sexually Transmitted Diseases/Aids Program, etc.) among others.

4.3.5 Federative coordination structures, institutional capacity of subnational institutions and financing: the production of municipalist legacy from federal polarization

The direction of the decentralization strategy had close relations with the contradictions between the federative context and the game of actors of the health sector. On one hand, it reflected the tendency to strengthen the federal government because of the reform agenda promoted by President Cardoso, and on the other hand, because of the institutionalized municipalization in the period 1993-94.

Despite the contradictions, this scenario allowed convergences. The dynamics of the federative context of the period was favorable to greater federal regulation, but it was not an impediment to municipalization, since local power acquired significant appreciation in the discourse of international development agencies in the 1990s. In addition, the decentralization model of BON 96 brought gains or did not imply losses for virtually all the main actors involved in SUS. The reformist actors ensured the deepening of municipalization, the federal bureaucracy expanded its role in defining policy standards and the use of resources transferred by the Union, and the private sector kept the prerogative of “fee for service” payments, acquired in the negotiations regarding the extinction of Inamps.

The other factors relevant to increase chances for the institutionalization of the BON 96 strategy were also taken into account by President Cardoso’s government.

The implementation of BON 96 was accompanied by functional changes in the organizational design of the Ministry of Health in order to adjust its framework to the new role of the federal administration, characterized by greater inductive and regulatory force on intergovernmental relations of SUS.

Under Adib Jatene’s administration, the framework from 1993-1994 was maintained and improved, strengthening the tripartite municipalist federalism. In the period 1997-2000, this structure presented greater changes aiming to provide more capacity of federative coordination to the Union. These changes involved in 1997 the creation of the Secretariat of Policies and Evaluation (SPSA) and of the Secretariat of Special Projects in Health (Brazil, 1997). In 1998, the entire organizational structure of the Ministry of Health was reviewed and the Department of Information Technology of SUS (Datasus) was transferred from the National Health Foundation (Funasa) to the Executive Secretary of the Ministry. In 1999, the Secretariat of Administration of Investments in Health (1999) and the Brazilian Health Regulatory Agency (Anvisa) (Law 9.782, from 26 January 1999) were created. Finally, in 2000, the structure of the Secretariat of Health Care (SAS/MS) was revised (Decree 3.496, 01 June 2000).

Sector financing also showed more regularity during the second half of the decade. In spite of the instability that characterized 1995 and 1996 and the variations in the overall volumes of Ministry of Health expenditures throughout the period in question, the specific spending levels for “medical and health care” in the years 1997 to 1999 were higher than previous years on average, considering the constant values. There was no real growth in per capita spending or appropriate adjustment following the GDP growth. However, the approval of the above mentioned CPMF and the regularity of transfers showed to the states and municipalities that they would be able to afford their financial commitments and hire professionals, service providers, suppliers, etc.

As a result of the alignment of these factors, at the end of 1999, 97% of the municipalities were already following BON 96, 4,854 (90,7%) in Full Administration of Primary Care and 496 (9,3%) in Full Administration of the Health System. This result consolidated the municipalization, with significant federal regulation, such as the predominant pattern of intergovernmental relations resulting from the political-institutional process of SUS decentralization in the 1990s.

5. Conclusion

The model proposed for the analysis of the dynamics that led to the polarization of federative relations in SUS in the 1990s showed that the correlation of forces among actors of the health sector consisted of the main variable in the definition of the direction of change, although the configuration of the federative context external to SUS was capable of promoting lateral changes in sector orientations or creating impasses and blocking the agenda, due to the preferences of the president and the use of the strategic resources of the Union. When there was a convergence between these two variables during President Itamar’s government, decentralization became possible, articulating forces around the municipalization project. Such process put an end to the critical juncture opened by the crisis in Inamps, producing a sectorial lock in.

The direction of the national policy has shown to be the point of convergence of the pressures coming from the federative context and the correlation of forces in the health sector. When there was convergence between the two pressure points, there was greater uniformity in policy decisions, most of them oriented towards municipalization. In cases in which there was no alignment, the results varied. In President Collor’s government, the lack of convergence with the sectoral forces advocating for reforms, blocked the Ministry of Health deconcentration strategy. In President Cardoso´s government, there were conflicts among the leaders and changes in alignment of the Ministry, with greater proximity to the sector’s actors during Jatene’s administration and greater alignment with the president and the economic area in the other administrations.

The relevance of the three other variables used in the model of analysis — the federal coordination structure, the financing and the institutional capacity of the states and municipalities — was strongly associated with the level of institutionalization of the decentralization strategies analyzed, when the direction was already defined and formalized in one of the BON strategies. That is, they were fundamental to increase the degree of participation of subnational entities.

The analysis of the SUS decentralization process in the 1990s, presented in this article, has relevant implications for the very interpretation of the changes driven by health reform in health policy in Brazil, for the historical-institutionalist theoretical explanations of institutional trajectories and for the future evolution of federative sectoral relations.

Differently from analyzes such as those expressed in (Faletti, 2010), which affirm that there have been only incremental changes in the design of the Brazilian health system, this article shows that there were significant changes that transformed a centralized system without popular participation in another decentralized system with significant participation of civil society. The fact that the changes have been done in a piecemail process does not mean that they have been of little relevance. They were conducted respecting the democratic process and articulated with an institutional reform that includes the construction of instances of agreement, reconstruction of management mechanisms, empowerment of new actors, etc.

The historical-institutionalist theory emphasizes how the trajectories of dependency develop and how critical conjunctures emerge. Thus, one of the challenges of this approach is to show how the transition occurs between a critical conjuncture and the formation of the legacy, which results in the installation of a new path of dependence. The analysis conducted in this article brings relevant theoretical contributions in identifying the variables that interfere in this process in the political, institutional, financial and managerial levels. Analyzing the weight of each of them throughout the process of health policy reform in the Brazilian case was possible to identify how different factors contribute to the formation of a new dependency trajectory.

Finally, it is important to emphasize that the changes introduced in this period imply in the redefinition of the role of the three spheres of government that presented expressive sustainability in the following decades. Even in the face of a new context that implied recentralization in the political and fiscal field, the decentralized configuration of the SUS presented significant institutional resilience. Thus, future reform processes intended by the Ministry of Health will only be carried forward if sustained by consistent negotiations with states and municipalities.

Referências

• ABRUCIO, Fernando. Os barões da federação: os governadores e a redemocratização brasileira São Paulo: Hucitec, 1998
• ARRETCHE, Marta. Democracia, federalismo e centralização no Brasil Rio de Janeiro: Editora FGV; Editora Fiocruz, 2012.
• ARRETCHE, Marta. Estado federativo e políticas sociais: determinantes da descentralização. São Paulo: Fapesp, 2000.
• BARROS, Maria E.; PIOLA, Sergio F.; VIANNA, Solon M. Política de saúde no Brasil: Diagnóstico e Perspectivas. Texto para Discussão (Ipea), n. 401, 1996.
• BÉLAND, Daniel; MYLES, John. Varieties of federalism, institutional legacies, and social policy: Comparing old-age and unemployment insurance reform in Canada. International Journal of Social Welfare, 21 (sup01), p. 75-87, 2012.
• BENZ, Arthur; BROCHEK, Jorg. Federal dynamics: introduction. In: BENZ, Arthur; BROCHEK, Jorg. Federal dynamics: continuity, change and the varieties of federalism. Oxford: Oxford University Press, 2013. p. 1-23.
• BRASIL. Constituição da República Federativa do Brasil Brasília: Congresso Nacional, 1988.
• BRASIL. Presidência da República. Decreto 2.284, de 24 de julho de 1997 1997. Brasília: Presidência da República.
• BROSCHEK, Jorg. Historical institutionalism and the varieties of federalism in Germany and Canada. The Journal of Federalism, v. 42, n. 4, p. 662-687, 2011.
• BROSCHEK, Jorg. Historical institutionalism and comparative federalism. World Political Science Review, v. 8, n. 1, p. 101-128, 2012.
• CAPOCCIA, Giovanni; KELEMEN, Daniel. The study of crititcal juncture: theory, narrative, and counterfactuals in historical institutionalism. World Politics, v. 59, n. 3, p. 341-369, 2007.
• CARVALHO, Gilson. A inconstitucional administração pós-constitucional do SUS através de normas operacionais. Ciência & Saúde Coletiva, v. 6, n. 2, p. 435-444, 2001.
• CARVALHO, G. O financiamento público federal do Sistema Único de Saúde 1988-2001 Tese (doutorado) - Universidade de São Paulo, São Paulo, 2002.
• CEBES. IXConferência Nacional de Saúde. Saúde em Debate, n. 36, p. 4-5, out. 1992.
• CORDEIRO, Hesio. Descentralização, universalidade e equidade. Ciência & Saúde Coletiva, v. 6, n. 2, p. 319-328, 2001.
• CORDEIRO, Hesio. O Instituto de Medicina Social e a luta pela reforma sanitária: contribuição à história do SUS. Physis - Revista Saúde Coletiva, v. 14, n. 2, p. 343-362, 2004.
• ESCOREL, Sara; BLOCK, Renata. As conferências nacionais de saúde na construção do SUS. In: LIMA, Nisia T. et al. (Org.). Saúde e democracia: história e perspectivas do SUS. Rio de Janeiro: Editora Fiocruz, 2005. p. 83-119.
• FALEIROS, Vicente et al. A construção do SUS: histórias da reforma sanitária e do processo participativo. Brasília: Ministério da Saúde, 2006.
• FALETTI, Tulia. Infiltrando o Estado: a evolução da reforma da saúde no Brasil, 1964-1988. Revista Estudos de Sociologia, v. 15, n. 19, p. 345-368, 2010.
• FLEURY, Sonia. Reforma del Estado. Rev. Adm. Pública, v. 35, n. 5, p. 7-48, 2001.
• FLEURY, Sonia. ¿Universal, dual o plural? Modelos y dilemas de atención de la salud en América Latina: Chile, Brasil y Colombia. In: MOLINA, Carlos; ARCO, Jose (Org.). Servicios de salud en America Latina y Asia Washington: Banco Interamericano de Desarrollo, 2003. p. 3-39.
• GUERRA, Alceni. Entrevista exclusiva com o ministro da Saúde. Revista Saúde em Debate, n. 29, p. 5-11, jun. 1990.
• HALL, Peter; TAYLOR, Rosemary. As três versões do neoinstitucionalismo. Lua Nova, n. 58, p. 193-223, 2003.
• IMMERGUT, Ellen. The theoretical core of the new institutionalism. Politics & Society, v 26, n. 1, p. 5-34, 1998.
• LIMA, Luciana. D. Federalismo, relações fiscais e financiamento do Sistema Único de Saúde: a distribuição de receitas vinculadas à saúde nos orçamentos municipais e estaduais. Rio de Janeiro: Museu da República, 2007.
• MACHADO, Cristiane. Direito universal, política nacional: o papel do Ministério da Saúde na política de saúde brasileira de 1990 a 2002. Rio de Janeiro: Editora Museu da República, 2007.
• MAHONEY, James. Path Dependence in HISTORICAL SOCIOLOGY. Theory and Society, v. 29, n. 4, p. 507-548, 2000.
• MAHONEY, James. Path-dependent explanations of regime change: Central America in comparative perspective. Studies in Comparative International Development, v. 36, n. 1, p. 111-141, 2001.
• MAHONEY, James; THELEN, Kethleen. A theory of gradual institutional change. In: MAHONEY, James; THELEN, Kethleen (Org.). Explaining institutional change: ambiguity, agency, and power. Nova York: Cambridge University Press, 2010. p. 1-37.
• MELO, Marcos A. O sucesso inesperado das reformas de segunda geração: federalismo, reformas constitucionais e política social. Dados - Revista de Ciências Sociais, v. 48, n. 4, p. 845-889, 2005.
• OBINGER, Herbet; CASTLES, Francis; LEIBFRIED, Stephan. Introduction: federalism and the welfare state. In: OBINGER, Herbet; CASTLES, Francis; LEIBFRIED, Stephan (Org.). Federalism and the welfare state: new world and European experience. Cambridge: Cambridge University Press, 2005. p. 1-49.
• OUVERNEY, Assis M. Federalismo e descentralização do SUS: a formação de um regime polarizado de relações intergovernamentais na década de 1990. Tese (doutorado) - Escola Brasileira de Administração Pública e Empresas, Fundação Getulio Vargas, Rio de Janeiro, 2015.
• PAGE, Scott. Path dependence. Quarterly Journal of Political Science, v. 1, n. 1, p. 87-115, 2006.
• PAIM, Jairnilson. A democratização da saúde e o Suds: o caso da Bahia. Saúde em Debate, v. 27, p. 5-11, 1990.
• PIERSON, Paul. Increasing returns, path dependence, and the study of politics. American Political Science Review, v. 94, n. 2, p. 251-267, 2000b.
• PIERSON, Paul. Not just what, but when: timing and sequence in political processes. Studies in American Political Development, v. 14, n. 1, p. 72-92, 2000a.
• RIBEIRO, José M. Conselhos de saúde, comissões intergestores e grupos de interesses no Sistema Único de Saúde (SUS). Cadernos de Saúde Pública, v. 13, n. 1, p. 81-92, 1997.
• SILVA, Silvio F. Municipalização da saúde e poder local: sujeitos, atores e políticas. São Paulo: Hucitec , 2001.
• SOIFER, Hillel D. The causal logic of critical junctures. Comparative Political Studies, v. 45, n. 12, p. 1572-1597, 2012.
• SOUZA, Celina. Federalismo, desenho constitucional e instituições federativas no Brasil pós-1988. Revista de Sociologia e Política, n. 24 p. 105-121, jun. 2005.
• WAYLAND, Kurt. Social movements and the State: the politics of health reform in Brazil. World Development, v. 23, n. 10, p. 1699-1712, 1995.

Publication Dates

• Publication in this collection
  Nov-Dec 2017

History

• Received
  20 Apr 2016
• Accepted
  10 Aug 2017