Chapter 6 elucidates why structural reform has remained off the table in São Paulo since Brazil’s transition to democracy. It focuses on three specific events since the 1980s, instances when reform seemed imminent but ultimately fell short. Two instances were selected due to being repeatedly identified in interviews as especially salient cases of egregious police violence that led to calls for reform; the third instance, meanwhile, was an ultimately failed effort by a sitting governor to enact police reform. Chapter 6 presents a comparative sequential analysis of these sets of events to demonstrate how the state’s Military Police exerted pressure to limit policy options and how fragmented preferences and the absence of political competition led political leaders to conclude that structural police reforms would not be electorally advantageous. Considering the cases as sequences of events that fail to bring about comprehensive structural reforms helps to elucidate the police’s remarkable continuity in São Paulo State. This comparative sequential analysis demonstrates how long-term institutional persistence has been driven by the absence of an electoral counterweight to the structural power of the police due to enduring fragmentation of societal preferences and weak political competition in the state.

Chapter 1 develops a theoretical framework for understanding the persistence of police forces as authoritarian enclaves in democracies, distinguishing between authoritarian and democratic coercion. Under authoritarian rule, the primary purpose of coercion is to keep the leader in power, and it is deployed with few constraints. In democracies, where governments must place checks on their own power, coercion ought to be deployed primarily to provide security for the citizenry rather than to serve the interests of the leader, constrained by the rule of law, and subject to external accountability. I disaggregate coercion along three dimensions: the extent to which coercion is governed by law or is applied arbitrarily; the strength of external accountability mechanisms; and whether coercion primarily serves leaders’ interests or to protect citizens. I demonstrate that police reform to promote democratic coercion has been relatively rare in democratic Latin America, even as rising crime and violence made security a salient electoral issue. I assess the problem posed by the endurance of authoritarian coercion for democracies and situate this institutional persistence within the structural power of the police, which leads politicians to engage in accommodation with the police, an exchange relationship that creates entrenched interests that favor institutional persistence.

Incentives within the political system during the 1985–2018 period made exceedingly difficult a shift toward either a more effective developmentalism or a more liberal market economy. Among the political institutions that drove this equilibrium were: 1) an electoral system that fragmented political party representation; 2) coalitional presidentialism, whereby a strong president sought to overcome fragmentation by providing incentives to coalition partners. Many of the 3) president’s tools of coalition formation were derived from the developmental state apparatus, including appointments and fiscally opaque instruments. The fragmentation of party life, alongside the many veto players engendered by coalitional presidential system, enabled 4) the emergence of both pluralist and corporatist forms of interest representation. The complementarities between these four political institutions had concrete effects: a resolute political system, in which change was slow and incremental, marked by long-term reciprocal relations between private firms and public actors, defensive parochialism, weak checks and balances, and weak controls on the developmental state apparatus.

Because of the economic power of incumbent firms and the political power of multiple veto players, changes to the developmental state were usually incremental.This contributed to the protagonism of the civil service as a change agent. Drawing on three case studies, the chapter illustrates how epistemic communities within the bureaucracy guided a variety of innovations across unconnected policy arenas: fiscal, health, and anti-corruption. Although policy innovation by the bureaucracy was incremental, slow, and often restricted to particular “islands of excellence” within the archipelago of state agencies, it was nonetheless essential to the most important accomplishments of the past generation. Civil service incrementalism, however, may have made change away from the overarching systemic equilibrium of the developmental state less likely, by exacerbating the fiscal quandary, sustaining the coalitional presidential system, and suppressing demands for more radical reform.

Chapter 3 probes why São Paulo State’s police have yet to undergo comprehensive structural reform in the decades since Brazil’s democratic transition. Brazil’s transition entailed considerable military reforms in accordance with democratic principles but no comparable effort to reform the country’s police forces. I illustrate how the state’s Military Police leveraged its control over coercion to cultivate structural power vis-à-vis civilian political leaders, through both the selective provision of security in politically beneficial ways and the strategic withdrawal of service. The chapter describes how the Military Police succeeded in constraining policy options available to politicians, and how politicians benefited from accommodating the police. These conditions raised the threshold for police reform in São Paulo over the last three decades, favoring institutional persistence despite the prevalence of widespread extrajudicial killings. I show how the fragmentation of societal preferences over policing and security - rooted in differences in citizens’ experiences with police along lines of race, class, and geography - has yielded little electoral incentive for politicians to enact reform. Instead, a large segment of São Paulo’s citizens demanding weakened legal restrictions on the police’s use of coercion with little external accountability have provided a constituency for politicians favoring authoritarian coercion.

The institutional complementarities of that marked Brazil’s political economy between 1985 and 2018 period were placed under considerable strain by the recession, fiscal crisis, corruption investigations, and political upheaval that culminated in the 2018 election of Jair Bolsonaro. An open question was whether the multiple and overlapping crises of the 2010s would destabilize the institutional equilibrium that had prevailed over the previous generation. To evaluate this question, this chapter summarizes the contributions of this manuscript to our understanding of Brazil’s political economy under democracy. It then describes how past reform efforts failed to move the political economy away from the status quo equilibrium, and the ways in which institutional complementarities shaped the pace and scope of change. It concludes with an evaluation of the “stress test” posed by the Bolsonaro administration, pointing to the importance of changes in the political realm for resolving Brazil’s long-term economic developmental challenge.

Between 1985 and 2018, Brazilian economic well-being stagnated, with lackluster growth and regressive public policies destroying citizens’ life opportunities. There is considerable consensus about the sources of this low-level economic equilibrium, including low savings, low investment, and modest human capital improvements. Despite this consensus, and despite decades of reform, however, the overall institutional equilibrium changed only marginally. Drawing on the study of varieties of capitalism, this chapter describes how institutional complementarities drove actors’ incentives toward a collectively suboptimal equilibrium. Complementarities within and across five domains sustained the equilibrium: 1) the macroeconomy of a middle-income developmental state, 2) the microeconomy of firm organization; 3) the coalitional presidential political system; 4) the weak control mechanisms this political system set in place; and 5) an autonomous bureaucracy that permitted incremental reform but in consequence, may have moderated demands for more dramatic reforms while deepening fiscal constraints and impelling policymakers to preserve the tool kit of the developmental state.

This chapter analyzes Brazil’s 500 largest firms and financial institutions in the mid-2010s to evaluate how well corporate life fit into a Latin American variety of capitalism that Schneider (2013) termed “hierarchical market capitalism.” While Brazil adhered to the general characteristics of the hierarchical market economy (HME), Brazilian firm life differed from other HMEs in the region due to significant state activity, the presence of large but relatively undiversified business groups, and credit and equity markets with a large dose of state participation that enabled firms to behave and organize in ways that differed from their regional peers. Five characteristics of Brazilian firm life stood out: the segmented firm structure; the muscular influence of developmentalist policy tools on firms; the segmentation of labor markets; the segmentation of skills; and the segmentation of social policy provision. This segmentation had a variety of implications for firms’ incentives to participate in politics.

Developmental states must be politically strong to design, implement, and recalibrate developmental strategies. They must have the capacity to provide rents to firms that nudge them up the innovation frontier, as well as to demand reciprocity, or returns on those rents. Achieving these goals requires effective instruments of control. Analyzing four developmental programs undertaken by various governments during the 1985 to 2018 period – the Manaus free trade zone, the automotive regime, the ethanol program and the Greater Brazil Plan – this chapter demonstrates the endemic weakness of controls. The Brazilian developmental state was ineffective at controlling rents in ways that channeled business energies in strategically productive long-term directions. The causes of weak control included political factors associated with the coalitional presidential system and the weakness of checks and balances, bureaucratic factors such as the fragmentation of oversight, economic factors such as incumbent firm influence, and judicial factors such as the toothless policing of illicit links between firms, the developmental state apparatus, and the political realm.

Chapter 5 explores the prolonged institutional decay of the Colombian National Police, particularly during the 1980s and early 1990s, and considers the factors that impeded comprehensive structural reform. Following a period of generalized loss of legitimacy across the Colombian state resulting from widespread political and criminal violence, the early 1990s saw a range of transformative policy reforms – from health and education to fiscal policy – and an institutional overhaul in the form of an ambitious new Constitution. Reform of the National Police, however, did not gain much traction, despite widespread authoritarian coercive structures and practices, weak institutional capacity, and inefficiency in fighting crime. I demonstrate how the National Police leveraged its structural power – flowing from its primary role in fighting the country’s war against drug cartels and its institutional and financial ties to the United States – to constrain the policy agenda and thwart reform attempts, even in a political environment that was highly conducive to a range of institutional and policy reforms. I then show how fragmented societal preferences over policing and security, often divided along class lines, generated conflicting demands and provided little electoral incentive for politicians to push for police reform, reinforcing institutional persistence.