(Un?)Healthy Politics: The Political Determinants of Subnational Health Systems in Brazil

34
(Un?)Healthy Politics: The Political Determinants of Subnational Health Systems in Brazil Jorge Antonio Alves Department of Political Science Queens College, City University of New York [email protected] Forthcoming in Latin American Politics and Society. **Please do not cite without permission.** Abstract How does political competition shape institutions that govern the expansion of social policy subnationally? Brazilian states have shown a surprising variation in the design of their public health institutions, which regulate the distribution of health resources and citizen access to public health care. While many states have experienced fragmentation, some have remained highly centralized and discretionary, and only a select few have established a coordinated system based on power-sharing and rules-based distribution. Accounts that link public health care expansion to federal government imposition, presence of the public health care movement and leftist parties cannot fully explain this variation. Instead, in the three Brazilian states examined here, the nature of subnational political competition triggered different institution-building strategies. The findings indicate plural political competition yielded incentives for limiting state-level discretion and sharing power with municipal governments, while political concentration reinforced the attraction to centralized and discretionary policy-making. Acknowledgements An earlier version of this article was presented at the Conference on Subnational Research in Comparative Politics hosted by the Center for Latin American and Caribbean Studies at the Watson Institute for International Studies, Brown University, May 9-10, 2013. I appreciate helpful feedback from John Bowman, Tulia Falleti, Julie George, Patrick Heller, Wendy Hunter, Pauline Jones-Luong, Eduardo Moncada, Alfred Montero, Lorena Moscovich, Celina Souza, and three anonymous LAPS reviewers. I am especially grateful to Richard Snyder for ongoing advice and support. Support for field research was provided by the National Science Foundation under Grant No. 0819267 and Brown University.

Transcript of (Un?)Healthy Politics: The Political Determinants of Subnational Health Systems in Brazil

(Un?)Healthy Politics: The Political Determinants of Subnational Health Systems in Brazil

Jorge Antonio Alves Department of Political Science Queens College, City University of New York [email protected]

Forthcoming in Latin American Politics and Society.

**Please do not cite without permission.**

Abstract

How does political competition shape institutions that govern the expansion of social policy

subnationally? Brazilian states have shown a surprising variation in the design of their public health

institutions, which regulate the distribution of health resources and citizen access to public health care.

While many states have experienced fragmentation, some have remained highly centralized and

discretionary, and only a select few have established a coordinated system based on power-sharing and

rules-based distribution. Accounts that link public health care expansion to federal government

imposition, presence of the public health care movement and leftist parties cannot fully explain this

variation. Instead, in the three Brazilian states examined here, the nature of subnational political

competition triggered different institution-building strategies. The findings indicate plural political

competition yielded incentives for limiting state-level discretion and sharing power with municipal

governments, while political concentration reinforced the attraction to centralized and discretionary

policy-making.

Acknowledgements

An earlier version of this article was presented at the Conference on Subnational Research in

Comparative Politics hosted by the Center for Latin American and Caribbean Studies at the Watson

Institute for International Studies, Brown University, May 9-10, 2013. I appreciate helpful feedback from

John Bowman, Tulia Falleti, Julie George, Patrick Heller, Wendy Hunter, Pauline Jones-Luong, Eduardo

Moncada, Alfred Montero, Lorena Moscovich, Celina Souza, and three anonymous LAPS reviewers. I am

especially grateful to Richard Snyder for ongoing advice and support. Support for field research was

provided by the National Science Foundation under Grant No. 0819267 and Brown University.

Alves – (Un?)Healthy Politics

2

Brazil’s redemocratization process ushered important developments for expanded citizenship

such as greater responsibilities for social policies by the state—of specific interest, the determination

that health was a right, leading to the foundation of a universal health system (Sistema Único de Saúde,

henceforth SUS)—along with expanded mandates for the provision of public services by subnational

governments.1 But degrading macro and socio-economic conditions resulting from endemic corruption,

bureaucratic gridlock, and the predatory behavior of resurgent state-level political elites (Weyland 1996;

Abrúcio 1998; Hagopian 1996; Samuels 2003; Souza and Dantas Neto 2006) tempered these

developments. National democracy has since become increasingly consolidated (Power and Zucco

2012), as the federal government has implemented important reforms, including macroeconomic

stabilization and inducing programmatic social policies on subnational governments. This has had

important socio-economic and political impacts, including the increasing, though uneven, level of

political competition in Brazilian states (Souza 1997, 2009; Borges 2007, 2011). In some cases the

political competition has led to leftist party victories in former conservative strongholds (Montero 2012;

Zucco and Samuels 2014). But while existing research explains these important shifts in the nature of

political competition in Brazil, this literature says little about the impact of shifting competition on

subnational institutions and public policies.

This article argues that equitable social service expansion at the subnational level in Brazil is

determined by the nature of state-level political competition. The equitable distribution and expansion

of public health services obtain when state-level governments limit their own discretion over health

policy by building governance institutions that include municipal governments in decisions of policy

design and resource allocation, especially when those institutions govern the distribution of state as well

1 Brazil is a three-tiered federal system with 26 states, a Federal District, and 5,570 municipalities. Each level has

elected executives, legislatures, and a bureaucracy composed by career civil servants and political appointees.

Alves – (Un?)Healthy Politics

3

as federal resources. But actors2 shape health governance not only with an eye on health policy, but as

part of broader strategies to engage shifting electoral scenarios both nationally and subnationally.

National institutional blueprints are not fully implemented in the majority of states, but especially in

those where political competition is concentrated. Equitable governance institutions result where

political competition is plural. The argument is based on evidence from a comparative study of health

care governance using data collected during fieldwork in the states of Bahia, Minas Gerais and São

Paulo.3

These findings supplement and challenge existing studies of the political determinants of public

health and social policies. Past research emphasizes the importance of leftist partisanship (Navarro and

Shi 2001; Cameron and Hershberg 2010; Pribble et al. 2009; Levitsky and Roberts 2011; Pribble 2013),

programmatic networks (e.g. Hassenteufel et al. 2010), institutional factors such as concentration of

executive power (Immergut 1992), or policy legacies (Haggard and Kaufman 2008; Pribble 2013). In

Brazil, social and health service expansion has been linked to either national or local-level factors, such

as the infiltration of the state by programmatic actors of the Movimento Sanitário (Escorel 1998; Paim

2008; Falleti 2010), proliferation of leftist ideology and rule (Pimenta 2007; Sugiyama 2008), and the

spread of participatory governance institutions (Dagnino 2002; Coelho 2006; Avritzer 2009). While

important, these factors do not fully account for how patterns of political competition at the subnational

level shape the governance structures in which local health policies are derived and implemented. A

focus on subnational political competition strengthens our understanding of the conditions under which

2 The main agents in this study are political elites (incumbents or challengers for state and municipal governments) and their immediate political circles, especially the cabinet-level appointments such as public health managers (Secretários Estaduais e Municipais de Saúde) and the main lower-tier appointments that make policy decisions. 3 Data was collected during hundreds of hours of participant observation in 61 public meetings, over 80 semi-structured elite interviews—ranging from elected officials, appointed health managers, career civil-servants and public health specialists—as well as archival research, in two years of field research in three Brazilian states and the national capital between August 2008 and June 2010. Due to space restrictions, I cite the most relevant source, but most claims presented have been corroborated by multiple actors.

Alves – (Un?)Healthy Politics

4

equitable social policy expansion is more likely to take place.

Institutions, Political Competition and Social Policy

The public health policy literature considers a wide variety of determinants of equitable health

policy, ranging from global actors (e.g. Ruger 2004) down to local managerial and technical capacities

(e.g. Arretche and Marques 2002), epidemiological or social-economic determinants of health (e.g. WHO

2008). An important subsection of the public health and comparative literature has focused on the link

between political institutions, especially democracy and expanded health and other social services

(Brown and Hunter 1999; Sen 199; Franco et al. 2004; Ruger 2005; Huber et al. 2008; Haggard and

Kaufman 2008; McGuire 2010). However, the variation in approaches to health policy across both

established and third-wave democracies points to the importance of contextual factors shaping health

policy, ranging from social features such as ethnic composition (Gauri and Lieberman 2006) or the

prevalence of interest groups such as doctors or programmatic networks (Falleti 2010; Hassenteufel et

al. 2010), public support (Schlesinger and Lau 2000), to political and ideological aspects such as Leftist

rule (Navarro and Shi 2001; Cameron and Hershberg 2010; Levitsky and Roberts 2011; Pribble 2013).

An important subset of this broader literature focuses on how institutional features within the

state determine social policies, such as concentration of power in the executive and the proliferation of

veto points (Immergut 1992; Castiglioni 2005; Huber et al. 2008), the legacies of pre-existing policies

(Haggard and Kaufman 2008; Pribble 2013), and participatory governance (Dagnino 2002; Coelho 2006;

Avritzer 2009). Such institutions—formal legal or organizational constructs that constrain actors by

laying out rules or expectations of behavior—make outcomes more or less likely in part by regulating

participation in the policy discussion, what powers participants are granted, and how interests are

funneled into the decision-making process.

I argue that two main institutional features shape subnational health care systems in Brazil. First

is how state-level officials approach health governance institutions within the state, especially the

Alves – (Un?)Healthy Politics

5

Intergovernmental Joint-Management Commissions (Comissões Intergestores Bipartite, CIBs) that are

tasked with coordinating health policy among state and municipal managers (Levcovitz et al 2001; Lima

and Viana 2011). Though these institutions are nationally mandated, the subnational approach applied

here identifies and structures the explanation of local variation in how state-level governments

approach their CIBs or, more broadly, these institutions’ “process of becoming” (Abers and Keck 2013).

While some states included many facets of policy, negotiating and at times even jointly deliberating

policy with municipal governments, others included some but not all aspects of policy, and others

avoided them all but entirely. The face of state systems also depends on the extent of municipal

independence; therefore I also examine the extent to which state-level governments acquiesced to

further decentralization of health resources that granted municipalities control over higher-complexity

services.

Existing literature provides limited leverage in understanding the context in which decisions

shaping such governance structures take place. Due to the existing theoretical link between leftist rule,

programmatic governance and social service expansion (Navarro and Shi 2001; Sugiyama 2008; Pribble

et al. 2009; Cameron and Hershberg 2010; Levitsky and Roberts 2011; Pribble 2013), one leading

hypothesis is that health governance institutions will be strengthened in those states or municipalities

where leftist parties take office. In Brazil, scholars have classified the PT as the relevant case of

programmatic party in the left (Keck 1992; Hunter 2010), and linked PT-led administrations with the

local adoption of participatory governance (Abers 2000; Avritzer 2009; Baiocchi 2005) and of municipal

health reform (Pimenta 2007). Yet, these studies have a harder time explaining why subnational PT

governments do not equally pursue the creation or strengthening of participatory institutions, nor the

way in which they pursue the implementation or expansion of particular policies, such as public health

care, championed by sector-specific interests within the party.

Analysis of Brazilian health politics also highlight the role of the public health movement known

Alves – (Un?)Healthy Politics

6

as the Movimento Sanitário, or Sanitaristas, in driving national-level universal health reform (Weyland

1996; Escorel 1998; Arretche 2004; Paim 2008; Falleti 2010; McGuire 2010), or local-level reforms (e.g.

Gibson 2012). The importance of the Movimento Sanitário in expanding the concept of citizenship

during Brazilian redemocratization and in particular to health reform cannot be overstated.

Nonetheless, accounts that attempt to capture its salience in shaping public health in democratic Brazil

tend to ignore its fractious nature, especially as disagreements broke out during the process of

implementation of the SUS,4 and do not focus on the inherent tensions generated by the often

conflicting interests of state and municipal governments—despite the fact that many managers share

Sanitarista credentials and values. Finally, they cannot properly account for the proliferation of

Sanitaristas serving as career civil servants in even the most dysfunctional state health offices.

To build and extend on these studies I unpack subnational electoral competition to establish the

factors that shape the behavior of a broader set of political and bureaucratic actors. One strand of

scholarship predicts association between increased and/or more plural competition and implementation

of state and policy reforms, such as expanded public goods (Borges 2008; Diaz-Cayeros 2008; Arvate

2013), decentralization (O’Neill 2003), transparency (Berliner & Erlich 2015), or building strengthened

oversight institutions (Grzymala-Busse 2007). Other studies, however, argue that the relationship

between political competition, uncertainty, and reforms might be curvilinear (Sartori 1976; Geddes

1994; Montero 2001; Grzymala-Busse 2007; Alston et al. 2008): when competition is too heated, actors

revert to short-term and centralizing strategies to win political support. Only when political forces—and

therefore the costs of reform—are fairly balanced, are incumbents likely to build capacity in a way that

limits discretion.

I find that the nature of subnational political competition, more so than political party label or

4 See for example the discussion about the best way to decentralize the national social security health network in

MS/SGEP (2006). Arretche (2004) is an exception.

Alves – (Un?)Healthy Politics

7

ideology, determines how state-level governments shape healthcare governance institutions. In

politically concentrated settings, incumbents face few incentives to shed centralized and discretionary

control over policy, since the bulk of the costs would be imposed on the group currently in control. In

such an environment, opposition forces are less able to challenge the status quo because the

incumbents have a strong grasp over the state and continued exclusion from resources makes co-

optation ever more tempting to opposition mayors and legislators. Incumbents therefore seek to build

weak institutions, ensuring continued control over policy and greater discretion in how they allocate

resources. Similarly, the longer the continuity (incumbency) of the executive, the more likely that one

group will penetrate further into the state apparatus. And while this might generate further probability

of corruption and malfeasance, it also creates strong disincentives for policy experimentation or risk-

taking.

Conversely, settings with plural political competition generate incentives for building institutions

that contain executive discretion and share power, even if not directly. This is partly linked to a more

balanced division of political power, which implies that opposition forces are more viable in challenging

and constraining the incumbent’s discretion. Facing a greater number of viable political actors, both in

terms of a wider distribution of political power but also greater turnover at the state executive, provides

incentives for building institutions that distribute resources based on a clearer set of rules. Incumbents

do so to claim credit with voters who live in municipalities controlled by competitors or placate other

relevant political groups, while preventing future exclusion in the case of electoral defeat. A greater

division of political forces not only spreads the costs of reform (by the loss of spoils of discretion) but

turnover—especially repeated turnover—creates incentives for taking risks; risks which very well might

be introducing good-governance policies and trying to rise to the top of state politics by giving up a bit

(discretion) to get a lot (continued control over the state).

Variation in State-Level Health Systems

Alves – (Un?)Healthy Politics

8

Universal health care in Brazil envisioned not only an expansion of service provision, but also an

ontological shift in the system by emphasizing local-level governance, cost-effective primary care, and

civil society participation and oversight (Paim et al. 2011). Fulfilling this expansive mandate continues to

be a major challenge for Brazilian states and municipalities. Expanding care entailed overcoming

perverse institutional legacies, including fragmentation, geographic concentration, and the reliance on

an expensive, overly-curative network of private providers generated under one of the most centralized,

corrupt and unresponsive bureaucracies under the military regime (Weyland 1996; Falleti 2010).

The initial centrality of local governments to the SUS project also created unforeseen problems.

As municipal governments absorbed the provision of basic care services, they committed a growing

proportion of their revenues (16% in 2000-03, and almost 20% during 2004-2010)—vastly exceeding

their legal share (SIOPS 2013).5 Shortages in physical and human capital led to a scramble to build

expensive facilities, further atomizing health care investments in small, inefficient and low-quality

hospitals (La Forgia and Couttolenc 2008) and a bidding war to hire a limited number of existing medical

personnel. Predatory behavior, including free-riding by sending patients to receive care in other

municipalities with no compensation and municipalities accepting compensation and then failing to

provide services (PPI-BA 2009; PPI-SP 2009), became commonplace. Such issues proliferated in the

absence of intermediary organization between federal policy design and municipal service delivery. Seen

by national reformers as part of the problem, state-level governments were in some instances left out

and at times actively cut out of the public health system. Unsure of their mandate and losing access to

health resources, state-level health offices lost purpose and prestige (Government officials, SES-SP 2009;

SES-MG 2010).

5 Legal public health commitments increased in this same period from 7% of municipal revenues in 2000 to 15% after 2004. The extra municipal burden is partly due to a relative decrease in federal funding, but also to state-level governments shirking their required commitment (12% of revenues) by using a series of well-documented accounting tricks, such as including expenditures on corporatist health plans for civil servants or military, with the acquiescence of legislative and judicial branches.

Alves – (Un?)Healthy Politics

9

National public health activists had sought to ensure the continued growth of public health in

Brazil and to coordinate this expansion by developing a governance infrastructure that ingrained

technocratic expertise and grassroots participation. National legislation required that all health policies

and budgets be deliberated and approved in participatory institutions that included public health

professionals and organized civil society (the aforementioned Conselhos de Saúde, or health councils),

and that technical commissions facilitate horizontal and vertical coordination across levels of

government—the Comissões Intergestores, or intergovernmental joint-management commissions

(Levcovitz et al. 2001, Lima and Viana 2011). By governing the flow of policy and implementation

discussions in such a system, SUS designers hoped to ensure a high quality policy discussion within the

health sector, and therefore minimize political discretion. But what SUS designers did not envision was

that state-level governments might reject or co-opt the SUS project by how they chose to implement

these institutions.

I focus on two features of the institutional choices by state-level governments that shaped the

character of health care policy. The first is the level of municipal independence, most readily visible in

the extent of decentralization of health resources. Since the bulk of public health actions is

constitutionally allocated to municipalities, further decentralization is visible by state governments’

willingness to share responsibility and resources over higher-complexity procedures that entail

increased visibility, greater funding, and where the division of responsibilities between state and local

governments are not clearly defined. Municipal governments could expand their control over these

resources by applying for different levels of certification. While the specific extent of responsibilities

increased over time for each type of certification, the highest level of certification gave municipal

governments “Full Management” (Gestão Plena) of the municipal system, including the right to choose

providers and especially access to the full cadre of per-capita funding to finance medium and high-

complexity services transferred by federal Ministry of Health (otherwise controlled by the state health

Alves – (Un?)Healthy Politics

10

office). As Table 1 illustrates, state governments took noticeably different approaches in response to

municipal interest in further decentralization. In Bahia the state health office vocally rejected and

bureaucratically delayed municipal applications for certification. By contrast, state officials in São Paulo

eagerly embraced decentralization, shedding primary care facilities and personnel to municipalities and

encouraging municipalities to apply for higher certification. As a result, São Paulo’s municipalities

became, as a group, the most independent from the state level—independence that would later

increase the operational and political costs of coordinating state and municipal action. State-level

response in Minas Gerais was intermediate; it supported certification for a small number of larger,

higher-capacity municipalities, but the state office’s skepticism to additional decentralization is visible by

the slow expansion of certifications over time.

[Table 1 about here].

The second institutional feature is the extent to which state offices favor rules-based distribution

strategies for capacity and financial resources by empowering their joint management commissions

(CIBs). This requires examining how health institutions actually functioned in local settings. By bringing

together state and municipal health managers and requiring consensus to approve policies before they

could access federal funding, CIBs sought to guarantee participation by municipal governments –

including those in the political opposition. State-level officials did not always approach these institutions

willingly. In Bahia, ruling elites favored having state health officials guard central control over policy by

perpetuating lack of institutional controls; hence state-level leaders stunted and co-opted participatory

institutions. In São Paulo, state and municipal agents developed robust joint management institutions

that were highly capable of equitably negotiating the distribution of federal resources, but that lacked

power over important policy areas and the ability to coordinate state and municipal action. The CIB in

Minas Gerais matched São Paulo’s capacity to allocate federal resources, but also counted on additional

powers because state officials gave municipal managers a greater role in policy design and governance

Alves – (Un?)Healthy Politics

11

of health resource allocation within the state. Using the rules jointly developed within the CIB gave the

state health office in Minas Gerais the power to coordinate municipal action in a way São Paulo’s system

could not achieve.

Table 2 shows a summary of three distinct state-level systems that challenges the vision among

SUS founders and national-level designers of a fully functioning, cooperative governance of subnational

health care policy based on clear rules and power-sharing institutions. Only in Minas Gerais did plural

political competition enable such an outcome, whereas low political competition in Bahia and São Paulo

led governance institutions to be either coopted or realize limited success.

[Table 2 about here.]

Analysis of a few critical cases, especially within one country, is an established strategy to

examine internal variation while accounting for confounding national phenomena. The ability to

generalize hinges on careful case selection. While the cases are not strictly representative of the

population of 27 Brazilian states, they do fall contiguously along the fault-lines of “three Brazils,” a

developed and economically dynamic one to the south (São Paulo), a historically undeveloped and

stagnant north and northeast (Bahia), and an intermediate region that shares some features with each

extreme (Minas Gerais). Despite the socio-economic disadvantage of poorer states like Bahia, it is

important to study these distinct environments in conjunction exactly because national blueprints were

disseminated uniformly with the expectation that these institutions would function in similar fashion.

Yet, as the next section shows, the distinct political environments in these states generated sharply

contrasting institutional outcomes.

The Impact of Political Competition on State-Level Health Systems

Identifying Key Differences in State-Level Politics

Because differences in electoral environments are central to my argument, I first illustrate the

differences in political competition in Bahia, Minas Gerais and São Paulo using a disaggregated set of

Alves – (Un?)Healthy Politics

12

indicators to capture the distinct though inter-related facets of political contestation. First, I measure

political continuity at the state executive by calculating the proportion of gubernatorial elections that

resulted in either the reelection of an incumbent or a victory by a member of the governor’s party, out

of the overall number of elections. I also include the electoral margin of victory in first-round

gubernatorial elections, since high electoral margins have also been associated with lower quality

democracy. Next, I measure the executive’s legislative support. There are three legislative arenas where

the governor can have influence: the state-level Assembly (Assembléia Legislativa), the national

Chamber of Deputies (Câmara dos Deputados) and the national Senate (Senado). I calculate the

governor’s party share in each legislative body, and present them as a simple average. This measure

therefore captures not only the legislative control at the state level, but also incorporates a governor’s

political importance at the national stage.6 Finally, I also measure a governor’s clout over municipal

governments by the share of mayors held by the governor’s political party. Together these measures

provide a comprehensive and nuanced picture of state-level patterns of political competition.7 All

measures are averages for seven electoral cycles for governor and legislators between 1982 and 2006,

as well as four municipal elections between 1996 and 2008, from the Brazilian electoral authority.8 Table

3 provides a panorama of state-level politics in Bahia, Minas Gerais and São Paulo that shows sharp

differences in political environments.9

[Table 3 about here.]

For most of the period since redemocratization, politics in Bahia have been highly concentrated.

6 Some scholars challenge the notion that governors influence national-level legislators (e.g. Figueiredo and Limongi 2000). While conceding that governors are not the only actors shaping legislator behavior, I nevertheless follow Samuels’ (2003) insight that they do have extensive sources of influence over state delegations. 7 Though operationalized differently, these indicators are in line with other studies of subnational political

competition, such as Borges’ (2007) and Giraudy (2010). 8 There is no electronic access to the municipal electoral data prior to 1996. 9 These findings are similar to those in Borges’s (2007, 2010) studies of electoral competition in Brazil’s states, where Minas Gerais scored in the most competitive quartile, São Paulo in the second highest, and Bahia in the lowest.

Alves – (Un?)Healthy Politics

13

Most elections in this period were controlled by the right-wing Partido da Frente Liberal (Party of the

Liberal Front, PFL), led by charismatic state political boss Antonio Carlos Magalhães (nationally known by

his initials, ACM).10 ACM’s political group, known as Carlistas, won five of the seven gubernatorial

elections in Bahia since 1982, generally with sizable margins. The PFL singlehandedly held a majority of

the state’s legislative seats, an uncommon feat in Brazil’s highly fragmented system of proportional

representation. High as they may seem, the average results presented in Table 3 actually understate the

extent of Carlista dominance throughout most of the period. For one, the values are lowered by the lack

of political dominance when the PT’s Jacques Wagner surprisingly won the state gubernatorial election

in 2006. During the Carlista peak (1990-2005), the PFL held on average 57% of Bahia’s national and

state-level legislative seats. Furthermore, the focus on the PFL alone misses a unique feature of Carlista

political control in Bahia. Magalhães used a series of small parties in the state as de facto satellites to

the PFL to accommodate intra-alliance political disputes (Souza 1997; Dantas Neto 2006a). Incorporating

these colonized parties expands Carlista legislative dominance to 63%, with a peak of 73% during the

1998-2001 term. The weight of political dominance is even stronger at the municipal level, as the

broader Carlista alliance controlled an average of 71% of municipalities in the same period.

And yet, politics in Bahia was more volatile than such control would predict. Magalhães

experienced two embarrassing political defeats during this period. In 1986 Waldir Pires, from the

centrist Partido do Movimento Democrático Brasileiro (Brazilian Democratic Movement Party, PMDB),

united oppositions and soundly defeated Carlistas, backed by similar shares of legislatures and

municipalities. Benefitting from support from President Luis Inácio ‘Lula’ da Silva, the PT was also able to

edge out the favored Carlista candidate in 2006. The memory and risk of losing elections made ACM

renew efforts in exclusionary and clientelistic practices (Dantas Neto 2006a), with important

10

In 2007, the PFL officially changed its name to Democratas (DEM). Since the study mostly focuses in the period pre-dating the change, I use the old party label.

Alves – (Un?)Healthy Politics

14

implications for institution-building and the relationship with municipal governments in health policy.

Table 3 provides evidence of the competitive nature of politics in Minas Gerais. The greatest

discrepancy between Minas and the other two states is the lack of continuity in the state executive,

where turnover occurred twice as often as continuity, and winners carried relatively thin margins.

Executive control was also weathered by a plural distribution of legislative seats. The party of the

governor of Minas Gerais never controlled more than one of three national senators, and failed to gain

the most seats for the Chamber of Deputies or the State Assembly between 1990 and 2006. Unlike

Bahia, governors in Minas Gerais could not count on long coattails and had to interact with strong

political opponents from distinct groups and with legislative bases spread across the state. The

precarious hold on state-level office generated a high level of political and administrative instability.

Efforts by Minas’ executives to emulate ACM’s strategies failed to co-opt or exclude the other powerful

political groups in the state and lost power in subsequent elections. Observing this trend, Aécio Neves,

heir to a prominent traditional elite in Minas Gerais and up-and-coming leader in the PSDB, gambled on

good governance to change the nature of political competition in his state. To do so, he tasked

technocrats with institution-building while seeking support from opposition groups by sharing power.

Finally, São Paulo displayed a political stalemate. Three of the four measures in Table 3 show a

plural political environment in São Paulo, with power being fairly dispersed in the legislatures and

among municipal executives. The state’s voters consistently elected a plural delegation to the national

congress, including leading figures in both the national government and opposition. There is a

multifaceted balance of municipal forces, with larger cities surrounding the capital supporting leftist

administrations, while important medium-sized wealthy municipalities from centrist and center-right

parties. Electoral margins for gubernatorial elections were also relatively thin. The dissonance is the

continuous hold that one party, the PSDB, has enjoyed over the governorship of the state for most of

the democratic period. As we shall see, this created dual sets of political incentives within the state.

Alves – (Un?)Healthy Politics

15

First, the plural environment yielded some incentives to build rules-based institutions both due to higher

contestation by viable political opposition forces but also as a hedge in case national or local

connections strengthened the opposition. In fact, a temporally disaggregated measure would exhibit a

decreasing hold of the governor’s party on São Paulo’s legislative seats, in part due to the legislative and

local strengthening of the PT following the party’s rise to the presidency. The PSDB’s share of legislative

seats consistently shrunk between 1998 (25%) and 2006 (17%). But second, uninterrupted control over

the most powerful gubernatorial seat in the country also made state-level officials impervious to

opposition and municipal forces, leading to a substantial degree of discretion in formulating some state

policies.

Political competition was thus concentrated in Bahia, plural in Minas Gerais, and configured in a

stalemate resulting from plural competition and executive continuity in São Paulo. How did these

distinct configurations impact the design of public health institutions in each state?

Concentrated Politics and Discretionary Health in Bahia

Concentrated politics in Bahia led to the construction of a state health system that was similarly

concentrated, with discretionary policies derived to serve state-level political incumbents’ policy

preferences and by distributing health capacity and funding as pork and patronage. Concentrated

control over the state-level government allowed Carlista incumbents to distribute public health

resources as one more spoil of political victory in order to control local allies and punish opposition

forces. As a result, Carlista elites resisted the SUS’s expansive agenda and the pluralizing effects of

national democratization for over two decades.

A prominent force in Bahia’s politics dating back to Brazil’s bureaucratic-authoritarian regime,

Magalhães maintained political control through a combination of clientelistic strategies and exploiting

links to the national government and media (Motter 1994). Despite initial concern with the

Alves – (Un?)Healthy Politics

16

modernization of Bahia, he became increasingly focused on establishing tight political control over the

state (Dantas Neto 2006a). As specialists in Bahia’s politics have shown (Souza 1997; Dantas Neto 2006a,

b), ACM’s large electoral margins diminish when one considers the large abstention rates in Bahia’s

elections. These studies show that while Magalhães could control a steady 30% of Bahia’s total

electorate, mostly in the state’s poor interior, electoral defeat was a threat if the opposition could unite

and tap this large share of alienated voters, as with Pires’ election in 1986. Dantas Neto argues that

Magalhães regrouped from electoral defeat in José Sarney’s national cabinet, strengthening his pre-

disposition for personalistic and centralized control when he once again took control of Bahia’s state-

level government in 1990.

The concentrated nature of political competition in Bahia not only shaped ACM’s policy

preferences but also translated into a highly centralized, discretionary and ultimately exclusionary

health care system in Bahia. The state government strongly defended centralized government despite

substantial coaxing by the federal Ministry of Health. One such strategy by the Ministry tied resources to

the development of a series of technical and planning instruments that ensure equitable and consistent

distribution among municipalities. Bahia’s state health office (known by the acronym SESAB) under the

Magalhães government centralized and insulated this process, drafting a state health plan with no input

from municipal managers, purposefully omitting programs that would bring federal funding for

municipalities, delaying an influx of health resources that would in practice make them less dependent

on the state government.11 It was an unofficial but widely known SESAB policy to delay municipal

applications for service accreditation, withholding even the most basic form of certification for 85% of

municipalities until 1997 (Guimarães 2003), while holding more advanced certifications to a minimum.

11 A telling example is that Bahia’s state health plan, does not make use of the term municipalização (decentralization to the municipal level), a ubiquitous term during this period of heavy emphasis on local health governance (see for example the final report of the 1993’s 9

th National Health Conference, titled “Municipalization

Is The Way” (Municipalização É o Caminho)).

Alves – (Un?)Healthy Politics

17

Centralization was also visible in other institutional decisions: As municipalities in Bahia were relatively

weak compared to the financial and technical capacities of the state government, this generally meant

that health services had historically been provided either by publicly-funded non-profits or directly

owned and operated by the SESAB. The health office in Bahia maintained control of these facilities,

distributing the appointment of their managerial posts to political allies who could in turn use them as

local sources of patronage (Civil servant, SESAB, 2009).12 While subsequent governments have since

decentralized more basic units such as local clinics, the state office still controls a disproportionate part

of the facilities in Bahia.

Discretionary control over the health system was reinforced by the concurrent undermining and

co-optation of health support institutions. Bahia was a laggard in national terms in the implementation

of each of the legally-mandated institutions, begrudgingly founding them only when they became

requisites for accessing federal funding. Delay was followed by co-optation, ensuring the rubber-

stamping of top-down decisions (Members of CIB-BA, 2008-2009). The institutions were populated with

allies from within the coalition, mostly managers from small municipalities, at the expense of larger

cities controlled by the opposition.13 Even then, co-optation was paired with an emptying of actual

power. Within the coalition, intergovernmental negotiations consistently took place outside of

participatory institutions and quite often sidestepped the state health office altogether (Government

official, SESAB, 2008). Likewise, the main policy decision of the Carlista period—to outsource the

expansion of hospital capacity to nonprofits—was introduced as a bill to the acquiescent State Assembly

before it was ever considered by the CIB or the state health council, inverting the legal decision-flow.

12

This has clear implications for the nature of care locally. In municipalities within the coalition, the mayor or legislator can direct services to their clients. In opposition municipalities, hospital managers can deny or complicate service, practicing de facto opposition within their territory. 13 Records from the municipal health managers’ association (COSEMS-BA) from this period reveal that leadership and seats in the CIB were occupied by managers from at times remote municipalities governed by the governor’s party, while opposition municipalities were virtually absent.

Alves – (Un?)Healthy Politics

18

Unpopular among the public health sector and highly contested in other states’ participatory

institutions, the policy was quickly approved due to state-level influence (Members of CES-BA, 2008-

2010).

The Carlista group also made use of centralized control and co-opted participatory institutions

to punish the few municipal governments openly in the opposition. One such municipality was Vitória da

Conquista, the state’s third largest, and an important first-mover on public health due to historical ties

to the Sanitarista movement. Despite being a net provider of services to neighboring municipalities,

Conquista’s PT government did not receive proper compensation for those patients, since the system of

compensation that governs financial allocations across municipalities (Programação Pactuada e

Integrada, PPI), led by the SESAB, directed funding to other municipalities from within the Carlista

coalition. Likewise, access to federal funding for basic care programs was prevented by red tape and

bureaucratic foot-dragging by the SESAB (SMS-BA 2010). The opposition-controlled capital city of

Salvador was similarly denied accreditations, even in a situation when a SESAB technical memorandum

recommended that the CIB approve at least partial certification (Government official, SESAB, 2009).

When Carlistas won Salvador’s municipal election in 1995, the municipal health office under the PFL

dropped certification applications, focusing on running basic care facilities and acquiesced to the state-

level office running the city’s hospital network (Mayor 2009). Salvador’s isolation made it the last among

state capitals to receive full-management rights (CIT 2008).

The result is that health institutions in Bahia during the ACM years failed to meet the SUS’s

aspirations for equitable expansion of care. The state network continues to be disproportionately

owned and controlled by the state health office. Health expansion was limited and targeted; plans

commissioned to health experts that would be based on rules and health needs were ultimately shelved

(Government official, SES-MG, 2010). To access health funding, municipalities could fall in line, such as

the small town of São Félix that was rewarded with an expensive Neonatal Intensive Care Unit, go at it

Alves – (Un?)Healthy Politics

19

alone and risk limited success, such as Vitória da Conquista, or face abject failure, as in Salvador. All this

was tied to the Carlista response to Bahia’s politically concentrated environment.

Competition Breeding Health Care Coordination in Minas Gerais

In Minas Gerais a competitive and plural political environment produced a rules-based and

highly participatory system to govern public health; a system that limited the discretion of the state-

level government but in return granted it substantial power as coordinator of municipal action. Building

rules-based institutions and sharing policy design with municipal and opposition forces in Minas’ CIB was

the result of a gamble by a political incumbent and his agents in the state health office given the

constant turnover faced by his predecessors and the plurality of viable political forces controlling parts

of the state’s legislative representation and municipal governments.

Despite its recent history of good governance (Montero 2001, Borges 2008), political

competition in Minas Gerais did not always have such beneficial effects. Historical accounts point to the

oligarchic and conservative nature of its politics and the survival of its pragmatic and risk-avoiding

traditional elites (Hagopian 1996; Bates 1997). The electoral free-for-all following redemocratization

initially eroded the quality of public services, including the proliferation of patronage appointments that

threatened fiscal insolvency by the late 1990s. Aécio Neves, a rising political figure in national politics,

thus sought not only to win power in Minas Gerais, but also to keep power by building institutions that

would deliver a modicum of good governance. In doing so Neves would break Minas Gerais’ political

deadlock and emerge as the fulcrum of its new balance.

Neves tasked reformist technocrats to design and implement a “management shock” reform

program, streamline government institutions, invest in human capital, and emphasize governance

through close monitoring of performance indicators. Top positions at the state health office (SES-MG)

were staffed with a tandem of a political insider and a respected public health expert that together

Alves – (Un?)Healthy Politics

20

structured a strategic role for the state office as a manager and coordinator of municipal action. The

state office shed its health care network and supported municipal governments when they applied for

basic certifications and federal funding for basic care. Support for decentralization was limited, as the

SES-MG discouraged all but larger municipal governments from applying for certification of more

complex and expensive services. State-level managers sought to recentralize the strategic management

of the state system by structuring the territorial allocation of resources to maximize regional impact

irrespective of individual municipal interests (Civil Servants, SES-MG, 2010). Creating and strengthening

regional hubs entailed a de facto concentration of investments on fewer providers but also host

municipalities. While it is politically costly for a state incumbent to announce hospital closures, the state

health office revealed this intention by designing a program (Pro-Hosp) that tackled the prevalence of

small inefficient hospitals by targeting a select few (136 out of existing 600) that received strategic

support in order to build regional scale in the hospital sector.

There should have been significant resistance from municipal and private sector providers who

would individually stand to lose from the concentration of resource allocation in such a way. And yet,

programs such as hospital support continue to be very popular among the majority of municipal

managers interviewed and observed for this study. This outcome was the result of the state-level

strategies of creating rules-based allocation criteria for such programs. Rules that would govern the

state’s major health programs were determined in participatory forums such as Minas Gerais’ CIB,

where policy design was shared with municipalities. Municipal representatives from each of the state’s

regions were guaranteed a place at the negotiating table, which ensured that recipients were picked in

territorially equitable fashion and that recipients would have continued incentives to incorporate the

Alves – (Un?)Healthy Politics

21

needs not only of their home municipality, but also of their neighbors.14 The consistent enforcement of

those rules once programs were implemented meant that eligible municipalities, irrespective of political

allegiance, had a fair chance of receiving state support (Opposition SMS-MG, 2009).

A strengthened CIB and the enforcement of deliberated policies yielded re-centralization as the

state health office gained the authority to coordinate municipal action. In addition to including

municipalities in the policy design process, the SES-MG spurred coordination by shouldering the costs of

development and implementation of coordination tools such as the state’s medical referral system

(SUSFácil), generating financial inducements for municipal participation, and by acting as a funder-of-

last-resort by creating a revolving fund (câmara de compensação) that would square away any

discrepancies between predicted and actual patient flows in order to get the system up and running.

Since these incentives were rolled out concurrently with rules-based and deliberated processes and

regulations, municipalities of different political groups nonetheless had incentives to join. Once local

governments experienced the financial and operational relief of cooperative action, the state’s system

became increasingly popular among municipal managers (SMS-MG, 2009b).

As a result of risk-taking in light of high competition and the constraints imposed by a plural

division of political power, Minas Gerais’ state government surrendered complete discretion of state

health policy design in exchange for a de facto degree of centralization. By following allocation rules

devised jointly with both political allies and opponents, the state health office built a coordinated health

system that decreased inefficient investments and increased patient access to existing health capacity.

This system is popular with managers across the political spectrum, despite the fact that the state

government in Minas Gerais is one of only two states that underinvests in health based on minimal

investment requirements. Neves’ government thus harnessed health politics to become a key political

14 For example, one of the eligibility criteria for hospital support was the proportion of patients from other municipalities brought in through the state’s patient referral network (see for example CIB-MG resolution 427/2008).

Alves – (Un?)Healthy Politics

22

broker in an otherwise balanced political scenario.

Stalemate and Fragmentation in São Paulo

In São Paulo, state executive continuity in an otherwise plural environment produced a

fragmented health system. Plural municipal forces pushed for a robust system of participatory

institutions that strongly contest state action and equitably distribute federal funding. However,

continuity at the state-level led to the exclusion of municipal and opposition forces from shaping the

state health office’s (SES-SP) extensive health network, yielding a system that is fragmented both

vertically (in the interface between state and municipalities) but also horizontally, among municipalities.

In this system, municipal governments atomized health investments, duplicated health capacity in

response to local political incentives and continued to prop up low-quality small hospitals. Municipal

systems also do not interface with the state’s hospital network, causing problems for patient access to

São Paulo’s unparalleled health capacity. While higher financial and human resource endowments

certainly play a role in this outcome, it cannot be properly understood without comprehending the

state’s political environment; in particular the state incumbent’s disincentives to risk a powerful position

with reforms that potentially empower municipal opposition.

The coexistence of prominent opposition forces with a steady stronghold on the state

government by the PSDB has peculiar impacts on health policies in São Paulo. Prominent Sanitaristas

permeate the state’s health sector, and their participation in health care institution-building is

disproportionate even to the left’s increasing political representation.15 As leaders of the national

Sanitarista movement, São Paulo’s local health managers mobilized early to demand decentralization

15 While the PT controlled an average of 9% of São Paulo’s municipal governments in the state of São Paulo in the period 2003-2008, representatives from PT-led municipalities accounted for an average of 39% of leadership seats in the municipal health managers’ association (COSEMS-SP), including its presidency. Despite heavy Sanitarista presence in both other states, the PT block of municipalities did not achieve such levels of representation. (Author’s calculation based on COSEMS-BA, MG and SP records and electoral data from TSE).

Alves – (Un?)Healthy Politics

23

from the state government (Pimenta 2007), and found willing partners in the Sanitaristas within the SES-

SP. The state office yielded an extensive network of primary care facilities to municipal governments,

along with fully-funded personnel to staff them. State officials also encouraged municipalities to apply

for certification to take over responsibilities beyond primary care, allowing virtually all interested actors

to receive them (Civil servants, SES-SP, 2009). A quarter of all municipalities applied and received

certification in the first year of eligibility alone, meaning that almost half of the state’s citizens lived in

municipalities who had full responsibility over health care service provision.

Despite the agreement on decentralization, the SES-SP retained significant control over the

extensive and high-capacity public hospital network. Owned and operated by the state-level

government since the later 19th century (Hochman 1998; Mendes and Oliveira 2009), this system

demands large financial and managerial commitments from the SES-SP, adding impetus for state health

officials to shed what it considered to be its non-essential network of basic health services. This focus

generates a series of conflicts between state and municipal managers. First is a lack of institutional

attention to local problems—abandoning municipalities with unequal capacities to solve collective

action problems. The SES-SP has kept control over its network by making strategic decisions about its

expansion and operation outside the CIB. Exclusion from the management of such an important

component of the state’s health system frustrates local managers, who commonly voiced their

displeasure in meetings of the municipal health managers association (COSEMS-SP), especially those

from PT-led municipalities (SMS-SP 2009). Finally, despite having the resources to finance most of its

own actions, São Paulo still looks to the Ministry of Health for funding whenever possible. This produces

some conflict of interest in the distribution of new federal resources in joint institutions because the

state office functions as both a mediator between municipal governments and a competitor seeking

Alves – (Un?)Healthy Politics

24

resources.16

As a result of these processes, São Paulo’s robust health governance institutions nevertheless

exhibit a confluence of rules-based distribution with a consistently exercised right to self-determination.

The equitable distribution of federal resources is extensively deliberated. Yet both municipal and state-

level health officials are willing and able to hold onto entrenched positions to respect their freedom of

action in the running of their existing systems. Municipal managers equally defend the freedom to

expand local capacity as they see fit, while opposition managers further toughen the deal for the state

to pay the administrative and financial costs of coordination. Comfortable in its political position and

with a large network of health services to manage, the state office has no incentive to engage the

coordination battle. A telling example is that state and municipal officials failed to agree on a formula

for the system that governs financial allocations between municipalities (PPI) until 2007, years after

poorer and less developed states had theirs in operation (Civil servant, SMS-SP 2009; CGR-SP 2009).

In São Paulo, an unlikely stalemate of political features—plural competition in all arenas but the

state executive—has led to a fragmented state system, highly decentralized municipal health

governance but with centralized control of the state’s hospital network. While incremental development

occurs in São Paulo, the polarization of politics between its two most powerful political groups—the PT

and the PSDB—has made actors on both sides more intransigent in their intergovernmental relations.

Lessons from Comparing Health Institutions in Bahia, Minas Gerais and São Paulo

Comparing how the distinct political environments in these cases impacted institutional

outcomes reveals the importance of the governing strategies pursued by political elites to satisfy

political aspirations. Concentrated political control provided both the impetus and the permissive

16 The inherent conflict of interest when discussing distribution of federal funding (especially for hospital services, where the SES-SP is a major player) was brought up regularly by municipal managers in CIB-SP meetings, even during otherwise amicable deliberation.

Alves – (Un?)Healthy Politics

25

environment for a strategy of continued centralization and discretion in Bahia. Plural competition in

Minas Gerais led to repeated political failures until a suitor for state power attempted an inclusive

strategy that shared some power in policy design and catered to opposition forces by limiting state-level

discretion in the allocation of investments and resources. Finally, it was plural competition in São Paulo

that gave municipal and opposition actors enough clout to extract decentralization from the state office

and to ensure an equitable, rules-based division of federal funding. Yet, the PSDB’s entrenched control

over the state-level government gave it control over the state’s public hospital network, generating

incentives to guard centralized discretion, which in turn led to institutional and capacity fragmentation.

The application of these strategies had practical implications for the type of care offered in each

state as well as to broader political contestation. The political allocation of resources in Bahia limited

expansion in overall patient access while existing service capacity remained disproportionately in the

hands of the state government. This centralized discretion forced municipalities to fall in line while the

Carlista elite preserved the boundaries of state political power until losing their links to the federal

government. Health capacity in the capital city of Salvador, home to a fifth of the state’s residents,

remains stunted while the only public hospitals are run by the state government. Meanwhile,

coordinated care in Minas has boosted existing capacity and increased access to pooled resources

regardless of municipal size, location or political association. As a result, the health policies in Minas

Gerais are popular among elites as well as the population, part of the state government’s broader good

governance brand. This helped ensure a stronger political position that led to Neves’ emphatic

reelection (77% of votes), subsequently strengthening his technocratic and political managers to stay

the course while cementing his role as central power broker within a plurally divided political landscape.

Finally, the duality of political pluralization and executive continuity in São Paulo continued to yield a

political stalemate between state and municipal forces, leading to no compromise in the design of

coordination tools that would allow municipalities to pool resources and more credibly rely on each

Alves – (Un?)Healthy Politics

26

other’s capacity, as well as a disconnect between extensive primary care municipal networks and the

high-capacity secondary state network. The strengthening of the PT’s opposition in São Paulo, which is

historically overrepresented in public health circles, has further politicized the relationship between

state and municipal managers, increasing the stalemate and perpetuating fragmentation.

These findings suggest the importance of the form of political competition for the governing and

institution-building strategies pursued by subnational elites. They supplement the study of changing

national and regional patterns of political competition in Brazil by linking them to important and visible

differences in state-level governance structures in public health, which in turn affect the quality of care

provided to local citizens. These findings also have broader implications for the study of social policy in

increasingly competitive but still consolidating democracies. First, despite top-down pressures by an

overwhelmingly powerful national government and the empowering of municipal governments, state-

level governments still exercised wide latitude in how they implemented and expanded social policies.

Second, while institutions matter for the distribution and coordination of resources, they are designed

as strategic responses to local political environments, more so than national pressure, the presence of

local health movement actors (which was ubiquitous), or partisanship. Third, where institutions are

rules-based and equitable, they are built not only on technical expertise, but political viability. Finally,

despite increased political competition, subnational elites continue devising strategies to achieve and

maintain political power, potentially resisting and reversing these trends.

These lessons are tempered by the study’s limitations. While the comparative method allows for

closer examination of the processes and mechanisms within the cases, it cannot fully eliminate potential

alternative explanations or categorizations. Further study needs to test the hypotheses generated here

in the remaining Brazilian states. If the findings are robust, greater efforts need to be made to link them

with a broader comparative literature on health politics, but also to engage the broader conversation of

social policy and institutional reform in federal systems. Furthermore, the suggested connection

Alves – (Un?)Healthy Politics

27

between equitable governance structures and equitable health policies needs to be further explored,

both theoretically and empirically. Nevertheless, while the finding that subnational political competition

shapes governance institutions might appear focused, understanding the ways in which it does so locally

is of great significance for health care but also more broadly.

References

Abers, Rebecca Neaera. 2000. Inventing Local Democracy: Grassroots Politics in Brazil. Boulder, CO: Lynne Rienner.

Abers, Rebecca Neaera, and Margaret E. Keck. 2013. Practical Authority: Agency and Institutional Change in Brazilian Water Politics. New York: Oxford.

Abrúcio, Fernando. 1998. Os Barões da Federação: O Poder dos Governadores no Brasil Pós-Autoritário. São Paulo: Hucitec.

Alston, Lee, Marcus Melo, Bernardo Mueller, and Carlos Pereira. 2008. “The Predatory or Virtuous Choices Governors Make: Political Institutions and Economic Performance.” Available at: http://www.odi.org/events/docs/2071.pdf

Arretche, Marta. 2004."Toward a Unified and More Equitable Health System: Health Reform in Brazil." In Robert R. Kaufman, and Joan M. Nelson, Eds. Crucial Needs, Weak Incentives: Social Sector Reform, Democratization, and Globalization in Latin America. Baltimore, MD: Johns Hopkins.

Arretche, Marta, and Eduardo Marques. 2002. “Municipalização da Saúde no Brasil: Diferenças Regionais, Poder de Voto e Estratégias de Governo.” Ciência & Saúde Coletiva, 7(3):455-479.

Arvate, Paulo. 2013. “Electoral Competition and Local Government Responsiveness in Brazil.” World Development, 43(March): 67-83.

Avritzer, Leonardo. 2009. Participatory Institutions in Democratic Brazil. Baltimore: Johns Hopkins.

Baiocchi, Gianpaolo. 2005. Militants and Citizens: The Politics of Participatory Democracy in Porto Alegre. Stanford, CA: Stanford.

Bates, Robert. 1997. Open-Economy Politics: The Political Economy of the World Coffee Trade. Princeton: Princeton.

Berliner, Daniel and Aaron Erlich. 2015. “Competing for Transparency: Political Competition and Institutional Reform in Mexican States.” American Political Science Review, 109(1): 110-128.

Borges, André. 2007. “Rethinking State Politics: The Withering of State Dominant Machines in Brazil.” Brazilian Political Science Review, 1(2): 108-136.

_______. 2008. “State Government, Political Competition and Education Reform: Comparative Lessons from Brazil.” Bulletin of Latin American Research, 27(2): 235-254.

_______. 2011. “The Political Consequences of Center-Led Redistribution in Brazilian Federalism: The Fall of Subnational Party Machines.” Latin American Research Review, 46(3): 21-45.

Brown, David S., and Wendy Hunter. 1999. "Democracy and Social Spending in Latin America, 1980-92." American Political Science Review, 93(4): 779-790.

Alves – (Un?)Healthy Politics

28

Cameron, Maxwell A., and Eric Hershberg. 2010. Latin America's Left Turns: Politics, Policies, and Trajectories of Change. Boulder, CO: Lynne Rienner.

Castiglioni, Rossana. 2005. The Politics of Social Policy Change in Chile and Uruguay: Retrenchment versus Maintenance, 1973-1998. New York: Routledge.

CIT (Câmara Intergestores Tripartite). 2008. “Municípios Habilitados em Gestão Plena do Sistema Municipal de Saúde - Maio de 2006.”

Coelho, Vera Schattan. 2006. "Democratization of Brazilian Health Councils: The Paradox of Bringing the Other Side into the Tent." International Journal of Urban and Regional Research 30 (3):656-671.

Dagnino, Evelina. 2002. Sociedade Civil e Espaços Públicos no Brasil . São Paulo: Paz e Terra/ UNICAMP.

Dantas Neto, Paulo Fábio. 2006a. “Surf nas Ondas do Tempo: do Carlismo Histórico ao Carlismo Pós-Carlista”. Caderno CRH, 39: 213-235.

_______. 2006b. “O Carlismo Para Além de ACM: Estratégias Adaptativas de Uma Elite Política Estadual”. in: Celina Souza. and Paulo F. Dantas Neto (Eds.): Governo, Políticas Públicas e Elites Políticas nos Estados Brasileiros. Rio de Janeiro: Revan.

Diaz-Cayeros, Alberto. 2008. “Electoral Risk and Redistributive Politics in Mexico and the United States.” Studies in International Comparative Development, 43(2): 129-150.

Escorel, Sarah. 1998. Reviravolta na Saúde: Origem e Articulação do Movimento Sanitário. Rio de Janeiro: Fiocruz.

Falleti, Tulia G. 2010. “Infiltrating the State: The Evolution of Health Care Reforms in Brazil, 1964-1988.” In James Mahoney and Kathleen Thelen, Eds., Explaining Institutional Change: Ambiguity, Agency, and Power. New York: Cambridge.

Figueiredo, Argelina and Fernando Limongi. 2000. "Presidential Power, Legislative Organization, and Party Behavior in Brazil." Comparative Politics 32(2): 151-170.

Franco, Álvaro, Carlos Álvarez-Dardet, and Maria Teresa Ruiz. 2004. “Effect of Democracy on Health: Ecological Study.” BMJ, 329: 1421-1423.

Geddes, Barbara. 1994. Politician's Dilemma: Building State Capacity in Latin America. Berkeley: University of California.

Gauri, Varun, and Evan S. Lieberman. 2006. "Boundary institutions and HIV/AIDS policy in Brazil and South Africa." Studies in Comparative International Development, 41(3): 47-73.

Gibson, Christopher L. 2012. Civilizing the State: Civil Society and the Politics of Primary Public Health Care Provision in Urban Brazil. Ph.D. Dissertation.

Giraudy, Agustina. 2010. “The Politics of Subnational Undemocratic Regime Reproduction in Argentina and Mexico.” Journal of Politics in Latin America, 2(1): 53-84.

Grzymala-Busse, Anna. 2007. Rebuilding Leviathan: Party Competition and State Exploitation in Post-Communist Democracies. New York: Cambridge.

Guimarães, Maria do Carmo. 2003. “Processo Decisório e Conflitos de Interesse na Implementação da Descentralização da Saúde: Um Estudo das Instâncias Colegiadas na Bahia.” Caderno CRH, 39: 105-132.

Haggard, Stephan, and Robert R. Kaufman. 2008. Development, Democracy, and Welfare States: Latin America, East Asia, and Eastern Europe. Princeton: Princeton.

Hagopian, Frances. 1996. Traditional Politics and Regime Change in Brazil. New York: Cambridge.

Hall, Peter A., and Rosemary CR Taylor. 1996. "Political Science and the Three New

Alves – (Un?)Healthy Politics

29

Institutionalisms*." Political Studies, 44(5): 936-957.

Hassenteufel, Patrick, Marc Smyrl, William Genieys, and Francisco Javier Moreno-Fuentes. 2010. "Programmatic Actors and the Transformation of European Health Care States." Journal of Health Politics, Policy and Law, 35(4): 517-538.

Hochman, Gilberto. 1998. A Era do Saneamento: as Bases da Política de Saúde Pública no Brasil. São Paulo: Hucitec/ANPOCS.

Huber, Evelyne, Thomas Mustillo, and John D. Stephens. 2008. "Politics and Social Spending in Latin America." Journal of Politics, 70(2): 420-436.

Hunter, Wendy. 2010. The Transformation of the Workers’ Party in Brazil, 1989-2009. New York: Cambridge.

Immergut, Ellen M. 1992. Health Politics: Interests and Institutions in Western Europe. New York: Cambridge.

Keck, Margaret. 1992. The Workers’ Party and Democratization in Brazil. New Haven: Yale.

La Forgia, Gerard M., and Bernard F. Couttolenc. 2008. Hospital Performance in Brazil: The Search for Excellence. Washington, DC: World Bank.

Levcovitz, Eduardo, Luciana D. de Lima, and Cristiani V. Machado. 2001. "Política de Saúde nos Anos 90: Relações Intergovernamentais e o Papel das Normas Operacionais Básicas." Ciência e Saúde Coletiva, 6(2): 269-291.

Levitsky, Steven, and Kenneth M. Roberts. 2011.The Resurgence of the Latin American Left. Baltimore: Johns Hopkins.

Lima, Luciana Dias de, and Ana L. d'Ávila Viana. 2011."Descentralização, Regionalização e Instâncias Intergovermentais no Sistem Único De Saúde." In Ana L. d'Ávila Viana, Ed, Regionalização e Relações Federativas na Política de Saúde no Brasil. Rio de Janeiro: Contra Capa.

Mahoney, James, and Kathleen Thelen, Eds. 2010. Explaining Institutional Change: Ambiguity, Agency, and Power. New York: Cambridge.

McGuire, James W. 2010. Wealth, Health, and Democracy in East Asia and Latin America. New York: Cambridge.

Mendes, José D. and Vanessa E. de Oliveira. 2009. Saúde Pública Paulista: 60 Anos da Secretaria de Estado da Saúde. São Paulo: SES-SP.

Montero, Alfred P. 2001. Shifting States in Global Markets. Subnational Industrial Policy in Contemporary Brazil and Spain. University Park, PA: Pennsylvania.

_______. 2012. “A Reversal of Political Fortune: The Transitional Dynamics of Conservative Rule in the Brazilian Northeast.” Latin American Politics and Society, 54(1): 1-36.

Motter, Paulino. 1994. “O Uso das Concessões das Emissoras de Rádio e Televisão no Governo Sarney.” Comunicação and Política, 1(1): 89-115.

MS/SGEP (Ministério da Saúde/Secretaria de Gestão Estratégica e Participativa). 2006. A Construção do SUS: Histórias da Reforma Sanitária e do processo Participativo. Brasília: Ministério da Saúde.

Navarro, Vicente, and Leiyu Shi. 2001. “The Political Context of Social Inequalities and Health.” Social Science and Medicine, 52(3): 481-491.

O'Neill, Kathleen. 2003. "Decentralization as an Electoral Strategy." Comparative Political Studies, 36(9): 1068-1091.

Paim, Jairnilson. 2008. Reforma Sanitária Brasileira: Contribuição para a Compreensão e Crítica.

Alves – (Un?)Healthy Politics

30

Salvador, Brazil: EDUFBA.

Paim, Jairnilson, Claudia Travassos, Celia Almeida, Ligia Bahia and James Macinko. 2011. “The Brazilian Health System: History, Advances and Challenges.” Lancet, 377(9779): 1778-1797.

Pimenta, Aparecida L. 2007. A História do CONASEMS: da Fase Heróica da Década de 80 ao Desejo de Mudança do Modelo de Atenção e Gestão dos Anos 2000. Brasília: CONASEMS.

Power, Timothy J., and Cesar Zucco. 2012. “Elite Preferences in a Consolidating Democracy: The Brazilian Legislative Surveys, 1990-2009.” Latin American Politics and Society, 54(4): 1-27.

Pribble, Jennifer. 2013. Welfare and Party Politics in Latin America. New York: Cambridge.

Pribble, Jennifer, Evelyne Huber, and John D. Stephens. 2009. “Politics, Policies and Poverty in Latin America.” Comparative Politics, 41(4): 387-407.

Ruger, Jennifer P. 2004. "Health and social justice." The Lancet, 364(9439): 1075-1080.

Samuels, David. 2003. Ambition, Federalism, and Legislative Politics in Brazil. Cambridge: Cambridge.

Sartori, Giovanni. 1976. Parties and Party Systems: A Framework for Analysis. Cambridge, UK: Cambridge.

Schlesinger, Mark, and Richard R. Lau. 2000. The meaning and measure of policy metaphors. American Political Science Review, 94(03): 611-626.

Sen, Amartya. 1999. Development as Freedom. New York: Anchor Books.

Sistema de Informações Sobre Orçamentos Públicos Em Saúde (SIOPS). 2013. Available at: http://siops.datasus.gov.br/evolpercEC29.php.

Souza, Celina. 1997. Constitutional Engineering in Brazil: the Politics of Federalism and Decentralization. New York: St. Martin's.

Souza, Celina, and Paulo Fábio Dantas Neto, Eds. 2006. Governo, Políticas Públicas e Elites Políticas nos Estados Brasileiros. Rio de Janeiro: Revan.

Sugiyama, Natasha B. 2008. “Theories of Diffusion: Social Sector Reform in Brazil.” Comparative Political Studies, 41(2): 193-216.

Weyland, Kurt. 1996. Democracy Without Equity: Failures of Reform in Brazil. Pittsburgh, PA: University of Pittsburgh.

World Health Organization (WHO). 2008. Closing the Gap in a Generation: Health Equity Through Action on the Social Determinants of Health. Geneva: World Health Organization.

Zucco, Cesar, and David Samuels. 2014. “Crafting Mass Partisanship at the Grass Roots, from the Top Down.” British Journal of Political Science, available on CJO2014. doi:10.1017/S0007123413000549.

Author Interviews and Observations

CGR-SP. 2009. Observations. Campinas region, May 14 and October 14.

Civil servant at SESAB. 2009. Salvador, December 16.

Civil servants (two) at SES-MG. 2010. Belo Horizonte, February 23 and 25.

Civil servants (one career, one appointed) at SES-SP. 2009. São Paulo, 2009.

Civil servant at one SMS-SP. 2009. Amparo, May 18.

Mayor of Salvador, BA (former). 2009. Salvador, December 9.

Alves – (Un?)Healthy Politics

31

Members of CES-BA. 2008-2010. Salvador, multiple interviews.

Members of CIB-BA (current and former). 2008-2009. Salvador, multiple interviews.

PPI-BA planning meetings. 2009. Observation. Salvador, December 3-11.

PPI-SP planning meetings. 2009. Observation. Campinas, March 14.

SMS-BA (Municipal Health Managers in Bahia) (two). 2010. Salvador, March 18 and 19.

SMS-MG (opposition municipality). 2009. Belo Horizonte, November 18.

SMS-MG (two). 2009b. Belo Horizonte, October 23.

SMS-SP. 2009. Amparo, May 6.

State government official (former) at SESAB. 2008. Salvador, December 18.

State government official (former) at SES-SP. 2009. São Paulo, May 20 and June 9.

State government official at SES-MG. 2010. Belo Horizonte, February 25.

Alves – (Un?)Healthy Politics

32

Table 1

Decentralization: Share of Municipalities with “Full Management” Certification

1998 2002 2006

Municipalities Population Municipalities Population Municipalities Population

Bahia 1 3 4 14 8 43 Minas Gerais 5 21 6 25 7 47 São Paulo 24 43 25 44 26 77

Brazil 6 25 9 34 12 54

Source: Câmara Intergestores Tripartite (2009) and IBGE Censo 2000.

Alves – (Un?)Healthy Politics

33

Table 2

Political Competition and Health Systems in Bahia, Minas Gerais and São Paulo

Bahia Minas Gerais São Paulo

State Politics Concentrated Competitive and Plural

Stalemate (Competitive + Executive Continuity)

Health System Centralized and discretionary

Coordinated (Centralized and rules-based)

Fragmented (Decentralized and partly rules-based)

Implication for care

Service allocation is determined by central policy preferences and distribution of pork

Service is allocated based on a coordinated plan developed by state and local governments.

Dual and disconnected allocation of services by state and local governments.

Alves – (Un?)Healthy Politics

34

Table 3

Political concentration in Bahia, Minas Gerais and São Paulo since re-democratization

Bahia Minas Gerais São Paulo

Governor: continuity rate17 57% 29% 86% Governor: margin of victory 29% 14% 10% Share of legislature18 51% 28% 31% Share of mayors19 28% 22% 31%

Source: Tribunal Superior Eleitoral (1994-2006) and Jairo Nicolau (1982-1990).

Note: Indicators calculated for 7 gubernatorial and legislative elections 1982-2006, except share of mayors, which is calculated based on 4 elections for which the data is available (1996-2008). Source: Tribunal Superior Eleitoral do Brasil (1994-2006) and Jairo Nicolau (1982-1990). 17

Following Borges (2007), the continuity rate is the number of gubernatorial elections in which the governor is either reelected or elects a candidate from their party divided by the total number of elections. 18 This is an un-weighted average of the governor’s party share of the state’s senate and chamber of deputy delegations, as well as the state assembly. 19

Mayoral elections are not concurrent with gubernatorial and federal legislature elections. Results are for first election after governor is elected.