Media Leaks and Corruption in Brazil

Analyzing the political consequences of the most extensive corruption

investigation in recent Latin American history, Operação Lava-Jato, Media
Leaks and Corruption in Brazil answers two central questions about the
contradictory effects news media has on political systems. First, how can
political actors in a seemingly well-functioning democracy quickly override
checks and balances, and replace a head of state with a corrupt vice-
president? Second, how can very active news media, while ostensibly
performing the role of the watchdog, still fail to deliver media
accountability to the public?
Combining a quantitative view of the media sphere with case studies of
the leaks, legal actions, and alliances forming and breaking in the Brazilian
Congress, Mads Bjelke Damgaard demonstrates that the media’s attention
to leaks and investigations of corruption paved the way for Dilma
Rousseff’s impeachment. By timing the disclosure of information in
scandals, actors with inside information were able to drive the media
agenda and let some scandals escape from the limelight. The book delivers
an in-depth study of how scandals become political weapons in a time of
media personalities and post-politics.
This book will interest scholars of Latin American Studies and Brazil,
and the broader fields of media studies, democracy studies, and journalism
studies.

Mads Bjelke Damgaard is PhD fellow in the Department of English,

Germanic and Romance Studies at the University of Copenhagen,
Denmark. He has previously worked at the University of Southern Denmark
as well as the Department of Cross-cultural and Regional Studies, Science
of Religions Section at the University of Copenhagen. When living in Rio
de Janeiro and Brasília, he was visiting PhD fellow with the Universidade
de Brasília and has collaborated with the media research group LEMEP
(Laboratório de Estudos de Mídia e Esfera Pública) of the state university
UERJ in Rio. He has previously published on corruption and Brazil in the
King’s College-based journal Brasiliana and the journal Ephemera –
Theory & Politics in Society.
Routledge Studies in Latin American Politics

19 Challenging the U.S.-Led War on Drugs

Argentina in Comparative Perspective
Sebastián Antonino Cutrona

20 Manipulating Courts in New Democracies

Forcing Judges off the Bench in Argentina
Andrea Castagnola

21 Crime, Violence and Security in the Caribbean

M. Raymond Izarali

22 Young People and Everyday Peace

Exclusion, Insecurity, and Peacebuilding in Colombia
Helen Berents

23 Indigenous Peoples and the Geographies of Power

Mezcala’s Narratives of Neoliberal Governance
Inés Durán Matute

24 Government and Governance of Security

The Politics of Organised Crime in Chile
Carlos Solar

25 Media Leaks and Corruption in Brazil

The Infostorm of Impeachment and the Lava-Jato Scandal
Mads Bjelke Damgaard

26 Public Debt and the Common Good

Philosophical and Institutional Implications of Fiscal Imbalance
 James Odom

www.routledge.com/Routledge-Studies-in-Latin-American-Politics/book-
series/RSLAP
Media Leaks and Corruption in Brazil
The Infostorm of Impeachment and the Lava-Jato
Scandal

Mads Bjelke Damgaard

First published 2019
by Routledge
711 Third Avenue, New York, NY 10017

and by Routledge
2 Park Square, Milton Park, Abingdon, Oxon, OX14 4RN

Routledge is an imprint of the Taylor & Francis Group, an informa business

© 2019 Taylor & Francis

The right of Mads Bjelke Damgaard to be identified as author of this work has been asserted by him
in accordance with sections 77 and 78 of the Copyright, Designs and Patents Act 1988.

All rights reserved. No part of this book may be reprinted or reproduced or utilized in any form or by
any electronic, mechanical, or other means, now known or hereafter invented, including
photocopying and recording, or in any information storage or retrieval system, without permission in
writing from the publishers.

Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are
used only for identification and explanation without intent to infringe.

Library of Congress Cataloging-in-Publication Data

A catalog record for this title has been requested

ISBN: 978-1-138-48548-8 (hbk)

ISBN: 978-1-351-04930-6 (ebk)

Typeset in Times New Roman

by Wearset Ltd, Boldon, Tyne and Wear
Contents

 
List of Illustrations
List of Abbreviations

1 Introduction

2 The Media System and Political Journalism of Brazil

3 Leaks and Their Effect in Media and on Politics

4 Lawfare and the Judiciary-Political Relations during the Lava-Jato

Corruption Scandal

5 The Informational Cascades of Impeachment and the Lava-Jato

Scandal

6 Coup d’état or Constitutional Act?

Afterword: Lula and Lava-Jato

Index
Illustrations

Figures

5.1 The chain of information in a denunciation of corruption

5.2 The informational cascade between media outlets
5.3 Number of headlines and news items on Eduardo Cunha’s
corruption cases
5.4 Number of headlines and news items on Lula’s corruption cases
5.5 Number of front-page headlines in Globo, Folha, and Estado on
corruption cases
5.6 Comparison of number of news items on corruption cases between
Folha and Estado
5.7 Approval ratings of government and coverage of scandals, 2014–
2016
5.8 Bovespa index, November 2015–March 2016
5.9 Bovespa index, April 18–May 12
5.10 Exchange rates BR$–US$, August 2015–August 2016

Tables

2.1 Print runs of Brazilian newspapers, 2005–2016

5.1 Number of front-page headlines in Globo, Folha, and Estado on
corruption cases
Abbreviations

Brazilian Accountability Institutions

MPF Ministério Público Federal (Public Prosecutors’ Office)

PF Polícia Federal (Federal Police)
PGR Procurador-Geral da República (Prosecutor-General)
STF Supremo Tribunal Federal (Supreme Court)
STJ Supremo Tribunal da Justiça (Supreme Tribunal of Justice)
TCU Tribunal das Contas da União (Tribunal for the Accounts of the
Union)
TSE Tribunal Superior Eleitoral (Superior Electoral Court)

Brazilian Party Abbreviations in 2016

DEM Democratas*
PCB Partido Comunista Brasileiro
PCdoB Partido Comunista do Brasil
PDT Partido Democrático Trabalhista
PEN Partido Ecológico Nacional*
PFL Partido da Frente Liberal
PHS Partido Humanista da Solidariedade
PMDB Partido do Movimento Democrático Brasileiro*
PMN Partido da Mobilização Nacional
PP Partido Progressista*
PPS Partido Popular Socialista
PR Partido da República
PRB Partido Republicano Brasileiro
PROS Partido Republicano da Ordem Social
PRP Partido Republicano Progressista
PRTB Partido Renovador Trabalhista Brasileiro
PSB Partido Socialista Brasileiro
PSC Partido Social Cristão
PSD Partido Social Democrático
PSDB Partido da Social-Democracia Brasileira
PSDC Partido Social-Democrata Cristão*
PSL Partido Social Liberal*
PSOL Partido Socialismo e Liberdade
PT Partido dos Trabalhadores
PTB Partido Trabalhista Brasileiro
PTC Partido Trabalhista Cristão
PTN Partido Trabalhista Nacional*
PTdoB Partido Trabalhista do Brasil*
PV Partido Verde
SD Solidariedade

Note
* In 2017, several parties were discussing or had filed official name changes: PMDB → MDB; PP
→ Progressistas; PSDC → Democracia Cristã; PSL → Livres; PEN → Patriotas; PTN →
Podemos; PTdoB → Avante and DEM → Centro Democrático.
1 Introduction

 
In March 2014, a case of corruption was uncovered in Brazil that would go
on to shake the country’s democratic institutions to their very core. Through
blind luck, a few arrests snowballed, and the case became the largest
investigation of political corruption ever in Brazil – a country already
infamous for dirty politics. However, the corruption investigation, known as
Operation Lava-Jato, did not merely result in trials and sentences for the
culprits. The Lava-Jato case had many impacts – but the most profound
impacts went well beyond the individuals investigated in the case.
Spinning off from the explosive political atmosphere developing along
with the Lava-Jato probe, the elected President Dilma Rousseff became the
target of dozens of petitions for impeachment, and was ousted in 2016, six
months after the installment of impeachment proceedings. In this
atmosphere, much-needed political and economic reform was obstructed
indefinitely, and the difficulties of constructing governable majorities in
Congress kept the state’s fiscal policies in disarray for three years in a row.
Meanwhile, scores of federal representatives, senators, and governors
became targets in an avalanche of new investigations precipitated by the
Lava-Jato probe. Rousseff’s vice-president stepped in and picked a cabinet
from the ranks of these congressmen, but four of the minister were forced to
resign on corruption charges, with two of these in prison at the time of
writing.
With no elections held, and despite the absence of criminal charges
against the impeached president, the political program of the Brazilian
government could change completely overnight, disrupting social conquests
of the last decade and manifesting the prevalent Latin-American trend
toward (or back to) neo-liberal economic politics. The Lava-Jato scandal
not only made this sea change possible, but also precipitated the fall of a
whole generation of politicians, and obliterated the citizens’ already
negligible trust in political authorities. Furthermore, targeting the top
construction companies of the country as well as the continent-leading state
oil company Petrobras, the scandal brought investments and public works in
an already retracting economy to a standstill.
Since 2014, evidence and material of the Lava-Jato probe have
continuously been made public to journalists by the prosecutors and the
regional court of Curitiba, but in addition to this, the same journalists have
been fed morsels (and sometimes torrents) of information in the form of
leaks. Like the WikiLeaks and the leaks of the Panama and Paradise Papers,
the Lava-Jato leaks have been continuously curated by journalists, and like
these other famous leaks, the sources remain undisclosed. The accumulated
data material, over the course of three years, is by now enormous, much
like the internationally known leaks, reaching terabytes of data at the time
of writing. In contrast to the international leak cases, however, the
journalists of the Brazilian news media have not been in complete control
of the material pertaining to the case.
The unauthorized leaks of key information in the case were distributed to
different media outlets, probably for a variety of strategic reasons and from
various sources with access to the investigation. The timing of the
disclosure had a great impact on the political atmosphere. In contrast to
instantaneously publicized data dumps such as the Panama Papers, the
Lava-Jato case unfolded and leaked over the course of several years, adding
a new episode to a crescendo of political drama each time new leaks
appeared. In 2016 alone, the leaks of the Lava-Jato operation had profound
consequences.
The consequences, however, were far more profound than, say, the
disclosure of the Panama Papers in the same period, involving political and
business elites of many other countries, including elites in the rest of the so-
called BRIC countries. While the Panama Papers, detailing international tax
evasion schemes, also made international headlines and meant public
disgrace and interrupted careers for some European bank executives as well
as the Icelandic prime minister, it barely piqued the interest of Russian and
Chinese political authorities, and the systemic issues revealed with the
leaked documents were virtually ignored around the world. The disclosure
of corruption in the Brazilian political and economic elite, on the other
hand, had critical consequences for the whole political paradigm of the
nation, for the state apparatus and economy, as well as the balance between
the branches of government. The ousting of the president, and ascension of
the vice-president with congressional support from the former opposition,
especially, was a surprising and monumental turnover: Surprising because
erstwhile President Dilma Rousseff was not at that point investigated in the
Lava-Jato probe, but instead was impeached on charges of being personally
responsible for delays in transfers of funds between the National Treasury
and public banks. Yet, in the crucial vote for her impeachment, held in the
lower House of Congress on April 17, scores of federal representatives
spoke of Rousseff’s corruption. This belied the fact that the Rousseff
administration had early on refused to interfere in the Lava-Jato
investigations, even as it reached the political and business elite of Brazil.
The impeachment was a monumental event, because it resoundingly
broke the governance model of welfare capitalism in Brazil, spearheaded
for 13 years by the Worker’s Party. By indicting, charging and sentencing
key figures of the Worker’s Party cadre, including the ex-President Lula, the
party image was severely tarnished and the left wing of Brazilian politics
was in tatters.
Finally, the probe and consequent trials broke with the established
pattern of impunity common in Brazilian political corruption cases. The
operation came to herald – perhaps prematurely – the end of one of the
mechanics that made the particular Brazilian hybrid regime of coalitional
presidentialism work in practice. Grafting and directing contracts within
public procurements to political allies may have been the oil that greased
the wheels and made the government coalitions work. Now, with a host of
politicians under scrutiny and with an empowered judiciary and prosecuting
branch of government, such a mode of striking political deals may have
been curbed. This, to be sure, is good for democracy, but the chances of
future consensus among the key players in the political arena looks very
uncertain at the time of writing, six months before the 2018 general
elections. What kind of majority will govern Brazil in the years to come?
What are the tools that can ensure stable coalitions in the future? Questions
such as these are still unanswered, in consequence of the turmoil launched
with the Lava-Jato case. In this political chaos of corruption scandals,
recession, and paralysis of important reform measures, it is pertinent to
consider the role played by scandals in catalyzing political transformations.

The Conundrums of the Corruption Probe

Social and political scientists are normally keen on explaining societal
change, and have developed large repertoires of theory to account for the
changes seen in contemporary societies. Every once in a while, these
repertoires fail to deliver answers. One such moment is arguably the Lava-
Jato case. In a period of less than two years, so many events conspired to
change the political scene that established academic perceptions about
Brazilian society and politics couldn’t keep up. In this book, I will focus on
one of the theoretical challenges posed by this surprising turns of events:
The democratic problem of too much visibility. The accountability overload
and excessive corruption disclosure produced cascades of public signals in
the news media – and with these cascades, news media ended up
disregarding relevant information about the corruption of the political actors
grabbing the reins of the federal government. With this preliminary
statement of the theoretical challenge, I have already introduced one of the
concepts that could increase the scientific understanding of the conundrums
posed by the Lava-Jato case.
After the media storm of the Lava-Jato scandal and the impeachment of
Dilma Rousseff, several accepted truths about Brazilian democracy
capsized, posing the following questions to researchers: First, how could
political actors in a seemingly well-functioning democracy quickly override
checks and balances, ousting the president Rousseff and putting a corrupt
vice-president in her place? Second, how could the nation’s very active
media, while they ostensibly were performing the role of the watchdog, still
fail to deliver media accountability to the public? Third, political scientists
had more or less agreed that the executive branch of the Brazilian system
held too much power relative to the other branches. That notion was
completely undermined with the impeachment proceedings. How could an
elected president not maintain even a third of the congressmen as allies,
given the budgetary and formal powers of the executive branch? Fourth, it
is generally supposed that the independence of the judiciary is a bedrock of
the rule of law, but in this case, it seems that independence of the judiciary
is apt to mutate into political interests. Could it be that there is such a thing
as too much independence, or in other words, that the Brazilian system of
checks and balances is, in itself, neither balanced nor checked properly?
These tendencies – the bypassing of democracy, the accountability
failure in plain sight, the impotence of the presidency, and the imbalance of
the judiciary – are now appearing as issues on the agenda in Brazil. At the
same time, each of these tendencies poses questions relevant to democracies
across the globe, especially in the age of virally spreading information, 24/7
news, and media personalities grabbing the positions as leaders of the
world’s largest democracies. Therefore, the case of Brazil is a valuable
lesson, not just to scholars, but to anyone interested in the state of
democracy today.
Finally, the Lava-Jato investigation came as a big surprise in itself,
because corruption scandals have been the norm since the military
dictatorship ceded the reins of the country to civilians in the 1980s. Since
then, barrages of scandals and disclosure of corruption have reached front
pages of newspapers, the airwaves, and the evening TV news, but accused
politicians were rarely convicted. The de facto law of impunity was
notoriously hard to break, even as public prosecutors, little by little, gained
political independence from the 1990s and onwards. But the astounding
success (which to many spectators should be interpreted as the astounding
excesses) of Lava-Jato has broken that vicious cycle, it seems. In order for
readers not familiar with the recent Brazilian history of political corruption,
I will present a small tour of the environment of media and politics since
the country’s transition, in the mid and late 1980s, to democracy.

25 Years of Media Exposés

The censorship institution of the military dictatorship, which had repressed

free journalism in Brazil in the 1970s (Kucinski 1991), petered out with the
gradual opening of democracy in the 1980s. As the generals stepped down,
a new era was inaugurated in press-state relations, and this era was
characterized by an explosion of corruption scandals during the civilian
presidency of José Sarney, attributable both to high levels of corruption and
the newfound freedom of press (Lattman-Weltman and Abreu 2001: 12ff.).
Arguably, the heyday of Brazilian investigative journalism was the 1992
disclosure of corruption within the government of Fernando Collor de
Mello. The young president was the first to be elected in free elections after
the military left the presidential palace in Brasília. Collor moved into the
Palácio de Planalto in 1990 on a wave of support for his image as the
“hunter of maharajas,” the broom that could sweep out the overpaid and
corrupt public servants still lingering in the country’s bureaucracy.
However, Collor was exposed himself as corrupt by his brother in the
weekly magazine Veja in 1992. He resigned before impeachment
proceedings were finalized, in face of a massive wave of popular protests in
the streets (Figueiredo 2010). His vice-president, Itamar Franco, stepped in,
but he struggled to find a cure for the four-digit inflation that Collor had left
behind, hollowing out the economy.
Only in the middle of the 1990s did the Brazilian government manage to
stop the hyperinflation marring the period after the return to the democracy.
Fernando Henrique Cardoso (former Minister of Finances in the Itamar
Franco administration) was the name of the new president who headed the
team of technocrats behind the successful Plano Real that ended the spiral
of inflation (Skidmore et al. 2010: 343). Cardoso’s team worked within the
boundaries of orthodox economics and the so-called Washington consensus
until his second term was up in 2002. Under Cardoso, many corruption
scandals surfaced, but his appointed prosecutor-general of the republic
decided to shelve the vast majority, and only let 60 out of 600 cases of
grand political corruption pass into court rooms.
At the end of his second period, Cardoso was unable to end a recession,
and this gave the leader of the opposition, the union boss and socialist Lula,
his first win in a presidential election, despite running in every election
since the first free election in 1989. Lula represented (and had founded) the
Worker’s Party PT, and as a former steel worker and union leader he had an
image of being close to the common people. However, in his successful bid
for the presidency he still had to convince the economic elites, both in
Brazil and in the international organizations such as IMF and the World
Bank, that he would maintain Brazil on the path of orthodox economics.
Lula managed to combine that pressure from the international financial
markets with his socialist ideologies, inaugurating several very successful
cash transfer programs and lifting millions of citizens out of poverty
without draining the public coffers (ibid.: 346). The neo-developmental
model of the Lula administrations (Bresser-Pereira et al. 2014) and the
transfer programs were lauded across the world, as the country joined with
other developing giants under the abbreviation of “BRICS.” The middle
class expanded enormously in Lula’s two terms as president. However, he
and his party came under severe fire in a corruption scandal known as the
Mensalão case.
 The Mensalão scandal (named for the neologism used to describe the big
monthly payments to parliamentarians) drew headlines throughout the
country in 2005 and 2006. Money was allegedly pulled from slush funds of
Lula’s PT and dealt out to congressmen in order to sway votes. While Lula
remained unscathed, and succeeded in getting re-elected in 2006, central
members of his inner circle in government and the party were indicted and
went on trial, after many delays, in 2012. Sectors of the electorate became
disappointed with the corruption in PT, a party used to defending ethical
politics loudly. However, in contrast to the Cardoso period, federal
prosecutors had by then obtained independence to pursue investigations into
national political elites. The laws that made this possible were signed by
Lula, and – undeniably – constituted a great step in constructing a
functioning set of checks and balances in the Brazilian democracy.
The trial of the Mensalão case before the Supremo Tribunal Federal (the
Supreme Federal Court) was big news: The exceptional number of
defendants, their political positions close to the presidency, the vast amount
of documents, and the potential to seriously harm the governing party
fueled the intense interest of the media. As the trial progressed, the adamant
will of the majority of judges to condemn political corruption severely were
also hailed as a historical event in the media, a milestone in the country’s
continuous struggle against corruption (Damgaard 2015, Michener and
Pereira 2016, Power and Taylor 2011: 33). The trial happened to be the first
time after the transition to democracy that a minister was sentenced for
corruption in the Supreme Court – even the case of Collor had ultimately
been filed away.
Lula had more or less left the political scene by the time of the Mensalão
trial, recovering from cancer, and left the presidency to a former minister of
Mines and Energy, later chief of staff in the cabinet, Dilma Rousseff.
Rousseff had initiated her presidential period in 2011 with a grand gesture
of cleaning out the ranks, by sacking all ministers denounced in the press.
Combined with impressive growth rates in the economy and a continuously
flourishing middle class, Rousseff managed to remain relatively popular,
even as her party was the target of scathing critique in the media coverage
of the Mensalão trial.
Two international milestones dominated the political arena in the years
following the Mensalão case. Brazil prepared to host the World Cup in
2014, and the Chinese economy began to slow down after decades of rapid
growth. This was gradually felt in the Brazilian sectors of commodities,
especially soy and mineral exports, and eventually meant reduced tax
income and empty public coffers. On the domestic scene, a wave of protests
against a bus fare increase picked up steam and erupted into massive street
protests in June 2013. The protests saw millions in the streets of the large
capitals, numbers not seen since the 1980s and early 1990s. The protesters
were not united under any single banner, as the multitude in the streets
branched off from the topic of public transportation fares (Cardoso and di
Fatima 2013, Saad-Filho 2013). However, the mainstream media quickly
cast the protests as being focused on corruption and governmental
overspending, especially on the World Cup prestige construction projects
(Avritzer 2016). Somehow, in the media discourse interpreting the June
protests, the supposed milestone of the Mensalão trial was not proof that the
law of impunity for political corruption had been broken.

The Initial Phase of the Operation Lava-Jato and the Chain of Plea
Bargains
In 2014, on March 17, another exception to the law of impunity appeared:
The first phase of the Lava-Jato operation caught a big fish with ties that
went to the top of the political scene. That morning, a money launderer (a
so-called doleiro) named Alberto Youssef, working out of Curitiba, Paraná,
was apprehended by Federal Police officers. Youssef had been identified as
a key actor in a criminal network linked to the gas station Posto da Torre in
the hotel sector of central Brasília. The police operation was code-named
Lava-Jato, literally “Jet Wash,” referencing both the car wash (lava a jato)
in that gas station and implying that the amounts of money laundered
surpassed the value of a jet (Netto 2016: 28).
The operation that apprehended Youssef was not what caught the eye of
Brazil’s media, however. While the case was covered briefly on March 18,
it was not linked to Petrobras before it became known that a former director
of Petrobras, Paulo Roberto Costa, was in cahoots with Youssef. Costa, who
headed the oil giant’s Supply Department until 2014, happened to be
arrested in the same case a few days later, and that became the first tremor
that set off the avalanche. On March 20, Costa was apprehended by the
Federal Police, because he, on the morning of Youssef’s arrest, had called
on his family to destroy evidence in his office and remove large amounts of
money. The ex-director’s family members were nearly caught red-handed
by the police, who rapidly accessed surveillance tapes from the very same
morning and eventually stopped the destruction of evidence (ibid.). Costa,
being one among the select group of people being called by politicians to
lead the large state companies, had much to tell the police once he decided
for a plea bargain.
Plea bargains (in Portuguese: delação premiada) are a relatively new
measure in criminal investigations in Brazil (importing it recently, like other
countries, from the common law system of the United States). Plea bargains
were only effectively introduced into the category of legal evidence in
probes into criminal organizations with the law 12.850 of 2013. In
corruption cases, the evidence emerging from plea bargains can be vital,
because corrupt deals are struck, in their very nature, between very few
people, producing as little material evidence as possible. With this legal
measure in hand, the team of police agents and prosecutors that conducted
the Lava-Jato investigations reminded the Petrobras director that he could
obtain a reduced sentence by spilling the beans. Costa, used to the luxury of
the federal capital and up-scale neighborhoods of Rio de Janeiro, took the
bait and revealed a number of things about the procurement processes of
Petrobras. He also mentioned the structure of payments that followed
subcontracts, and when the doleiro Youssef also opted to plea bargain, the
prosecutors were stunned as they gradually unraveled the practice and
extent of corruption in the state company.
There is another reason for mentioning plea bargains at this point, beside
the fact that it was the plea bargains of Costa and Youssef that really started
the avalanche of the operation. As the investigations developed, it became a
firm principle of the Lava-Jato task force and the Prosecutor-General
(directing inquiries in cases pertaining to parliamentarians, ministers and
governors) that plea bargains can only be negotiated when the testimony
incriminates individuals on higher organizational rungs. That is, each
defendant admitting to corruption charges and negotiating a plea bargain
must deliver names and evidence not only of those who moved the money,
but also those who asked for it, and those who paid. This is sensible from
the perspective of the investigations, because that is the only way to get to
the top of the chain of command – and if a defendant keeps certain things to
himself, a reduction in sentence would harm the spirit of the Rule of Law.
However, when this desire to get to the top of the chain of command
becomes manifest, defendants will have incentives to lie, because they are
more likely to obtain a plea bargain if they implicate business leaders and
political leaders above them.
Despite this potential problem for the quality of information in
investigations, the plea bargains of Costa and Youssef went on to trigger a
cascade of new bargains. Youssef implicated the former Samsung employee
Julio Camargo, and once he was arrested, Camargo pointed to an executive
by the name of Augusto Mendonça, who had been his colleague in the
construction company Toya Setal. Mendonça implicated, in turn, the
Petrobras manager Pedro Barusco, who also decided to strike a plea
bargain. Barusco identified more subcontractors of Petrobras that had paid
kickbacks for contracts, and the executives and managers of these
companies named more directors of Petrobras, who then implicated more
construction company leaders with direct or indirect ties to the most
influential Brazilian politicians. Some (but not all) of the plea bargains were
negotiated while the defendant was being held indefinitely in prison under
the law of “preventive arrests,” that is, imprisonment without sentence
aiming at avoiding destruction of evidence or even new crimes. To some
experts, the indefinite span of preventive arrests could be construed as a
way of forcing persons into admission of guilt. At the time of writing, the
former minister of Finances in the Lula administration, Antônio Palocci,
and the former President of the Lower House of Congress, Eduardo Cunha,
have both been held captive in Curitiba for a year, and both of them are
attempting to settle plea bargains. The stakes of this chain of plea bargains
are now so high that only implicating a president or an ex-president is
sufficient to make a deal go through. Hopefully, there will be ample
evidence, and not merely allegations, to back up such revelations about
masterminds in the heart of the political system.
The point, in sum, is that allegations about corruption have been feeding
the investigations, for good and for worse. Allegations have pointed police
agents to new targets and caches of evidence, and insofar as this has been
the case, it has surely been for the better. However, the incentive to
denounce others has had consequences outside of the courts, creating a very
peculiar atmosphere of news. Plea bargains are constructed through the
negotiation between a defendant, his lawyers, and the public prosecutors.
Many individuals had access to such plea bargains and the documentation
surrounding it, even before a judge validated the material as suitable for
trials and investigations. In a number of high-profiled cases of the Lava-
Jato probe, the material has been leaked to the press before the ink was dry
on the paper containing the terms of the plea bargain. This plea bargain law
is thus not merely a great innovation in the judiciary system, but the
successful use of the law in the unfolding Lava-Jato scandal also had
ramifications for the media system and the way investigative journalism
works today in Brazilian news rooms. During the course of this book, I will
present some of the many consequences of plea bargains and media leaks in
the political system.
The plea bargains laid bare a system of political negotiation that went
beyond the system of distributing coalition goods and “pork” (amendment
expenditure) to parties supporting the president (Raile et al. 2011). Not only
can Brazilian presidents wield extended powers to allocate state
expenditures to specific projects in exchange for votes in Congress, but an
astonishing amount of top-level public office positions are also governed,
ultimately, by the executive branch of the State. It is estimated that more
than 20,000 leading positions in public companies, agencies and other state
entities are appointed by the political system. In the case of Petrobras, Paulo
Roberto Costa and other directors admitted that they were politically
appointed (and later maintained in their positions) by government coalition
partners. In this way, each directorate was allocated to a political party, and
the contracts in the business areas of these directorates yielded kickbacks
for the parties. Earlier suspicions of the “auctioning off” of state enterprises
and private-public operations (Power and Taylor 2011: 264) were now
documented, as the Lava-Jato investigations revealed this practice not only
in Petrobras, but also in the state energy company Eletrobras, in nuclear
plants, and in the health sector.
Colluding in tenders, a cartel of 16 companies had informally distributed
sub-contracts of Petrobras and other state companies, fixing prices and
planning kickbacks for directors and for the politicians backing them. In
November 2014, CEOs and presidents of seven of these companies were
arrested, based on the information provided by the chain of testimonies
initiated by Paulo Roberto Costa. That became the start of Operation Lava-
Jato, and in the course of this book, the various consequences of this
investigation will be described.
Theorizing Information Flows and Media Leaks in Accountability
Processes
Little did the federal police agents that arrested Costa in March 2014 know
that three years hence, the operation would have changed the political
landscape of Brazil. Neither could they foresee how they themselves would
be hailed as heroes, and then cast as anti-heroes, by media, segments of the
population, and even ex-presidents and presidents. As pointed out above,
the Lava-Jato case, from 2014 to 2017, had any number of unforeseen and
surprising results. To reiterate, some of those results prompt the central
questions of this book: How could the nation’s media fail to produce
accountability to the public despite extensive coverage of the political
machinations and maneuvers? And how could political actors in a
seemingly well-functioning democracy quickly override checks and
balances, ousting the president Rousseff and putting a corrupt vice president
in her place?
These questions require crossing the disciplinary boundaries of media
studies, accountability research, and institutional theory. The deficits of
Brazilian accountability institutions are well known (for an overview, see
Macaulay 2011 and Power and Taylor 2011), and concerns about media’s
problematic influence on public opinion and politics have been voiced for a
century (e.g., Lippman 1922). The particular problems of the Brazilian
media system in relation to democracy have been articulated and discussed
by many influential authors (Avritzer 2016, Chaia and Teixeira 2000, 2001,
Lattman-Weltman and Abreu 2001, Lima 2004, 2006, 2009, Mesquita 2014,
Miguel 1999, 2003, Porto 2011). The democratic problems arising from the
Lava-Jato scandal cannot be reduced to either the field of media discourse
or an institutional field, however. Rather, the consequences appeared
precisely in the interstices of the media sphere and the web of
accountability mechanisms. The deficits of accountability, transparency, and
public visibility were nested within both organizational fields,
compounding the democratic problems and at the same time imbuing the
Lava-Jato investigations with unusual political potential.
Therefore, this book will establish a theoretical framework that can
address the intersection of media studies and accountability research, taking
into consideration the complex interplay between these social and
organizational fields. I will argue shortly that both the study of political
scandals and accountability theory can benefit from incorporating the
concept of informational cascades. By viewing interactions between media
organizations, the public agenda, accountability agencies and political
actors as essentially informational processes, evolving in dynamic
sequences, the framework established here aims at improving the
theoretical understanding of processes wherein media scandals and fragile
accountability institutions are used instrumentally.
Accountability has been conceived along two axes: the horizontal axis
and the two vertical dimensions of electoral accountability and societal
accountability (Mainwaring and Welna 2003). The horizontal axis
encompasses the constitutionally configured checks and balances of
democratic systems, such as veto powers of executives, judicial review of
courts, investigative and overseeing capacities of various government
agencies and courts, as well as the legislative instruments of oversight,
investigative commissions, and, last but not least, the impeachment
instrument (Smulovitz and Peruzzotti 2000: 153). Accountability actors on
the vertical axis include all citizens, in principle, but in particular, various
collective forms of non-governmental actors, such as civic organizations
and NGOs. The organizations of civil society that can mobilize citizens’
demands for accountability are important actors, as are political parties in
their representative capacity, and media actors. The main electoral
accountability body, included under the umbrella of vertical accountability,
is the constituencies of elected governmental actors. Together, the mesh of
institutions and actors constitute a “web” of accountability institutions
(Mainwaring and Welna 2003: 29) or “accountability networks” (Waisbord
2000: 210). In the Brazilian web of institutions, bodies, and agencies, each
actor might be more or less tuned into the processes of the others, and
problems may arise in both the lacunas and the overlapping areas of the
web (Power and Taylor 2011).
The processes whereby information circulates between the horizontal
accountability institutions, on one hand, and media organizations, on the
other, are particularly interesting in the Lava-Jato case. Here, the main
problem in the web of accountability arises from an institutional relation
crossing the vertical-horizontal divide. The extant literature on media
exposés in the Latin American media has emphasized how media can (and
should) activate accountability mechanisms (e.g., Porto 2012: 40,
Smulovitz and Peruzzotti 2000: 152, Waisbord 2000: 211). Knowing that
exposés and disclosure of corruption might activate horizontal
accountability mechanisms, both political actors and civic actors have
incentives for pushing information to media actors. As Peruzzotti and
Smulovitz say:

once the media was revealed as an effective mechanism for controlling

and accelerating public decisions, civil society organizations used this
discovery to gain access to an alternative route to justice, to get
attention from the public authorities, and to informally judge presumed
illegal activities. In some countries, this new role has led to the
emergence of a strong and sometimes threatening investigative
journalism.
(2006: 18)

However, Peruzzotti also warned that exposés and scandals could be

considered a tool in struggles between political elites, with Latin American
media covertly taking sides in such struggles:

given their privileged access to inside information [the press and the
opposition could choose what to disclose]. Political scandals were thus
largely seen as intraelite games, the outcome in each case
fundamentally depending on the postures and maneuvers adopted by
media and political elites. The citizenry and society at large only
played a passive role as “public opinion.… Can the Latin American
media be considered independent? Is the image of a fourth estate
adequate to describe the workings of journalism in the region? Do
media exposés and denunciations help to strengthen democracy and
governmental accountability? Or are they an instrument of the “politics
by other means” that ultimately feed political cynicism and apathy?
(Peruzzotti 2006: 255–256)

Peruzzotti’s suspicion of “politics by other means” was already voiced in

Latin American media studies of the 1980s and 1990s:

Against popular and media-centric approaches that conclude that the

news media have exceptional powers in bringing down ministers or
presidents, it is necessary to analyze the balance of congressional
forces, the configuration of the judiciary system, and the response of
 organized interests to gauge the consequences of media exposés.
Especially in countries where the news media have not historically kept
a distance vis-à-vis powerful interests but offer channels for doing
politics by other means, it is a mistake to examine the relations
between press and accountability independently from larger political
dynamics.
(Waisbord 2000: 222–223, emphasis added)

The critical perspective of Waisbord (and other related critiques

concerning unsubstantiated media accusations, e.g., Smulovitz and
Peruzzotti 2000: 154) casts doubt on the effectiveness of Latin American
media in creating accountability. These authors highlighted that exposés, in
fact, sometimes express the vested interests of elites rather than the ideals of
watchdog journalism. Though Waisbord analyzed several of the problems
of these horizontal-to-vertical information flows and the sub-politics of
sources and journalists (Waisbord 2000: 93), the matter largely stands as it
did in the 1990s. Then, Waisbord observed that “closeness between
journalists and official sources is also indispensable for the media to delve
into official wrongdoing” and that this proximity “results from the
organization of news-making and a journalistic culture that highly prizes
official news” (ibid.). A decade later, Porto echoed this, and also called to
attention the problem of “[c]ollusion between journalists and officials [that]
frequently prevents watchdogs from barking even when evidence of
wrongdoing is available” (Porto 2011: 111). Moreover, Porto noted the
related problems of counter-denunciations stemming from such closeness:
“Journalists’ dependence on officials also allows some sources to avoid
denunciation, while at the same time ensuring that the media will publicize
accusations against rivals” (ibid.).
In a nutshell, these problems cannot be grasped within the paradigm of
accountability theory, as this theory is not equipped to analyze exactly how
media systems are invested – politically and economically – in the
production or omission of scandals. Furthermore, the fundamental element
here is actually not the accountability mechanisms and institutions, but
rather the flows of information that are generated and disseminated by those
institutions. For these reasons, I will establish a line of analysis centered on
the flows of information. Theoretically, the reason for highlighting
information flows of scandals in relation to accountability is the following:
The production of accountability, in at least two senses of that word,
depends on distribution of information. The vertical accountability of
political actors, whether in the sense of electoral accountability or societal
accountability, depends on constituencies and civic actors having access to
accurate information about politicians and public office holders. Horizontal
accountability, meanwhile, depends on information distribution in another
sense: Interventions made through horizontal mechanisms of checks and
balances (e.g., overseeing or auditing governmental agencies) require an
element of transparency and publicness, and thus of distribution of
information. If the results of horizontal accountability mechanisms are not
made sufficiently public, actors caught in improbity may simply continue in
other areas of public life, or even the very same office, while waiting for a
final verdict or while dragging out trial and audit processes. Massive delays
(Falcão et al. 2014: 57, 81) of investigations, verdicts, and appeals at the top
level of politics are commonplace.
Because of the many opportunities for delaying and annulling legal
processes, voters and other social actors need media or other forms of
information distribution in order to become aware that horizontal
accountability mechanisms have been activated or stonewalled. The
effective and just application of horizontal accountability (such as trials or
impeachment) is thus theoretically linked to the vertical dimension
(electoral processes especially) through information distribution.
The work of prominent scholars analyzing scandals (such as Allern and
Pollack 2012, Maier 2011, Pujas 2002, Thompson 2000, Tumber and
Waisbord 2004) fails as starting points for analyzing the Lava-Jato case,
although these works predict, like the accountability theory cited, that
scandals generally have reputational and symbolic costs for the denounced
(e.g., Balán 2011, Peruzzotti and Smulovitz 2006: 11, Seligson 2002,
Thompson 2000: 245). However, while these authors observe that elites
trigger scandals to inflict this reputational harm on political opponents, that
does not systematically explain the ways in which accountability
institutions and media work to inflict such damage. The academic literature
on scandals has a lot to offer, but it rarely provides models of the
entanglement of judiciary and political institutions, nor of the information
flows in public spheres that set these institutions in motion. Furthermore,
most of this literature has been conceived in times where digital and viral
around-the-clock recycling of news and large-scale leaks were still in their
infancy.
The literature on scandals suggests that the usual route taken is the
“Indirect activation of horizontal mechanisms” where “claimants organize
and mobilize but also […] reach the media or the media reaches them”
(Peruzzotti and Smulovitz 2006: 18). That is clearly not the pattern of many
cases emerging from the Lava-Jato investigations. Here, the accountability
institutions – prosecutors, courts, audit agencies – have fed media with
information so that horizontal mechanisms have activated vertical
accountability actors, which is the opposite of the relation theorized in
much research. Moreover, the reputational costs predicted are only inflicted
when information about corruption is reproduced and acted upon in very
specific ways – but that reproduction does not necessarily lead to action in
accountability institutions, nor extensive coverage in the media.
In sum, the critique of exposés, though two decades old, has stopped
short of investigating the systematic links between legitimate and leaky
sources of accountability institutions, journalists, and editors. Furthermore,
the intra-elite perspective needs to be unpacked, as it glosses over the
important information flows between media and accountability actors,
which might be just as problematic as corrupt political elites denouncing
each other. The trajectory of information in public spheres is a crucial factor
in the evolution of the scandal, but hitherto, models of scandals have not
systematically treated information flows as analytical components. The
same goes for studies of accountability. To remedy the shortcomings of
accountability theory, scandal studies and the intra-elite struggle
perspective, I will present an approach tailored to examining how diverse
representations of corruption in the media came to impact upon the political
agenda, coalition formation and judicial processes. Information disclosure
in scandals and accountability processes together form sequences of
information flows, or informational cascades, defined in the following.

Informational Cascades
Recent theoretical work on information sequences in public spheres has
uncovered the pitfalls inherent in certain sequences of signals. This
structure is called informational cascades, and according to Hendricks and
Hansen (2016), such cascades emerge in groups of actors acting upon their
knowledge of the sequence of public signals rather than private knowledge
of information. I will briefly show how this theory of information and social
psychology can generate insights into the dissemination of corruption news
in Brazil. When deciding how to act,

agents may choose to base their action not only on their private
information, but also on the information extracted from their peers –
social proof.… An informational cascade … may easily occur when a
number of people doubt the accuracy of their own information, and
subsequently observe the preceding decisions and actions of others in
the hope that it can lessen the doubt. This course of action then leads to
each individual making the rational choice of following in his or her
predecessors’ footsteps, independent of his own preceding information.
(Hendricks and Hansen 2016: 52, 62)

Doubt about course of action is not the only reason for choosing to act
upon public signals, however. Scarcity of private information is another
reason for people to pay close attention to social proof. This also goes for
media actors. As described above, information on corruption publicized or
leaked from accountability institutions can pass into journalistic production
in one newsroom and, if it is treated as highly important there, it might then
spread quickly to all other news media. A scoop might spread simply
because one media actor signals the importance of the disclosed
information. In other words, a scoop on a front page will likely be
replicated by competitors in the same news market. The information might
then be recycled many times over in the sphere of media, where different
media outlets monitor and circulate news from competitors and reproduce
and respond to other outlets’ exposés to the degree that such news items are
considered relevant to their own target audience. When response-coverage
add new layers to an exposé, this can in turn be recycled by other outlets,
multiplying the coverage.
The information disclosed by one or more media outlets can be distorted
or misrepresented in various ways (potentially leading to states of dis- or
misinformation, which is defined in Chapter 5). Some information,
furthermore, might bear the connotation of corruption, contextually
signaling political transgression, but not actually spell it out in so many
words. So, in the public spheres of modern media, “informational cascades
may turn out to be extremely powerful instruments if controlled in such a
way as to manipulate public signals, so false information is spread instead
of true knowledge” (ibid.: 22).
Due to organizational constraints in media (deadlines and labor
resources, as well as physical attributes of newspaper space or broadcast
news program duration), some information will be included while other
may be disregarded at each broadcast or deadline. Some aspects might be
highlighted, at the news room or in the editing process, producing particular
frames interpreting the information concerning corruption (Entman 1993).
The same goes for the following instances where information about
corruption is disseminated – meaning is added or subtracted, texts are
interpreted, framed and modified in innumerable ways. But crucially, as
long as the basic bit of information – about political transgression – remains
intact and is consistently present across media and the subsequent sites of
dissemination, members of the various media audiences might consider this
information as multiple instances of the same public signal. In other words,
when there is a preponderance of one frame, signals are accumulating and
that might lead audiences (as well as media professionals) to disregard
information speaking to the contrary. Because of the evolving dynamics of
scandals, disclosure of corruption is not a one-shot public signal, and the
scandal coverage thus produces accumulated signals. This has
consequences for how new information is likely to be treated later on:

Information keeps accumulating until a preponderance of evidence

supports one action or the other by just enough to outweigh one
individual’s private signal. At this point a cascade starts and new
information stops accumulating.
(Bikhchandani et al. 1998: 158)

In situations where private knowledge is very limited for the majority of

actors, such as a national corruption scandal, these actors must make
decisions based on public signals. Looking to peers or the media actors who
supposedly holds privileged information, a “preponderance of evidence”
might then sway these actors.
Thus, at the level of the audience, the effects of the media system’s
circulation and shared coverage are reinforced because media consumers
often have a range of media outlets (and technological platforms) at their
disposal. The everyday experience of (re)-circulated coverage (especially
salient in moments of political scandal) is one aspect of our living media-
saturated lives (occurring while we are also often influenced by local
opinion leaders) (Lazarsfeld et al. 1944). If people recurrently glean
information about a scandal from one or several media – be it printed news,
broadcasting platforms, web-based news portals, weekly magazines, Twitter
or Facebook – then their opinion will not be formed based on a single news
text (see Miguel 2015 for a critique of the ideological weight of news
items), but rather the aggregate stack or sequence of information. So, in so
far as the analysis of cascades is situated at this level of aggregate news
coverage, it is thus hardly compatible with framing studies (in the vein of
Entman 1993 or Iyengar 1990). As the informational cascade accumulates
signals, news items with divergent interpretive frames, for all the potential
to sway individual opinion, are submerged in the wave of news repeating
the initial signal crashing upon the cognitive shore of the audience.
Because each informational cascade is unidirectional and creates herd-
like behavior, the phenomenon has been likened to financial bubbles, where
information-driven markets come to over-valorize certain bits of
information. In relation to the field of theories of media studies, the
informational cascade approach is linked to agenda-setting theory
(Wiewiura and Hendricks 2017). As predicted by agenda-setting theory
(Ghanem 1997, McCombs 2005, McCombs and Shaw 1972), the salience
given to news topics by media outlets will influence audience members’
perceptions of the relative importance of those topics. To this, the
informational cascade approach adds a framework for analyzing the
networked sequences of information flows, and the power of specific
sequences to attract the attention of an entire media sphere – or conversely,
the potential to generate pluralistic ignorance of information (Hendricks and
Hansen 2016). Because informational cascades can emerge in the audiences
of news media just as well as between media actors, the pluralistic
ignorance phenomenon also bears on how audiences are apt to interpret,
recycle or reproduce information, based on the sequence of signals
observed in news media.

Sampling of Material
To investigate the cascade, it is necessary to start where information
emerges: The newsrooms with privileged access to the national political
arena. This point of departure means that the media material sampled for
this inquiry is not exactly representative of all Brazilian media, nor of the
average Brazilian’s news consumption. More likely, the material of this
book reflects what the members of the nation’s political elite are reading
and watching. This focus on what have been termed legacy media – such as
newspapers and weeklies – has more to do with the special access of
information that journalists of the traditional news media continue to
sustain, than with the tradition inherited from earlier studies of scandals.
The newsrooms of the prestigious legacy media of Brazil unite
considerable agenda-setting power due to their privileged position at the top
of the information food chain, and in consequence of their vertical
integration on the media market (Matos 2008, 2012). In the perspective of
information distribution, vertical integration of media outlets means
recycling of the same news in a range of media platforms: The largest
media conglomerate, the Grupo Globo, controls much of the regional media
in both TV, radio, and print, as well as the most-viewed national channel,
the second-most viewed news channel, the third-largest newspaper (O
Globo) and weekly magazine (Época) and the national newspaper of
economic issues (Valor), as well as the most visited online news site (Porto
2012, Newman et al. 2015: 40). The Marinho brothers, who inherited the
executive positions from their father, control the Globo conglomerate. The
media business empire of the Marinhos is only challenged on the market for
news production by the evangelical bishop Edir Macedo (who owns the
broadcasting network Rede Record), by the Saad family running the
Bandeirantes network, by the Mesquita family (owning the newspaper
Estado de S. Paulo and the news agency Broadcast), by the Frias brothers
(controlling the newspaper Folha de S. Paulo and the UOL online
platform), and by the Civita family managing the Abril group with the
flagship weekly Veja (which is traditionally the most-read weekly
magazine, ahead of IstoÉ and Grupo Globo’s Época). The rest of the
Brazilian regional and metropolitan newspapers and broadcasts get their
national political news mainly from the news agencies and mother-
organizations behind those news outlets. According to Abramo:
 Not only the penetration of [Rio- and São Paulo-based] newspapers is
much higher than the press from other regions, but they also tend to
dominate what the other newspapers publish. Upon analyzing the
credited sources of stories published by the newspapers (about 14% of
stories are credited to news agencies), it is found that the news agencies
associated with two newspapers (Folha and Estado) are responsible for
90% of all agency-credited news published by the newspapers …
Adding the Globo agency, the percentage is 95%. This does not count
news items that are simply appropriated by a newspaper without credit,
a very common phenomenon.
(2007: 99)

What Abramo shows is that the news selection criteria are determined by
the media organization structure, and also in many cases by the regional
news values. That is, news emerges in the regional and local spheres of
media according to the organizational links to national media organizations
and news agencies, but, crucially, the coverage is then modified by local
relevance and political ties. This means that regional media will publish
material on national corruption if a local politician is implicated, or, in
many cases, will actively omit such information because of personal ties to
media owners (ibid.: 101), a tendency which is especially salient in the area
of radio broadcasting (Boas and Hidalgo 2011: 871).
The vertical integration of news media organizations is also apparent
online. The major content providers of political news (both newspapers and
broadcasters) each maintain online platforms, and the newsrooms produce
content for several media types, frequently recycling material to their sites.
The result is a media system that is highly concentrated in the hands of a
few owners, and tightly controlled by these. Journalists working with
national political news obtain access to information about political
corruption through a limited number of sources; sources that are often
involved in investigations.
Because these newsrooms of legacy media are usually first to report on
disclosure of corruption, and their news are subsequently filtered down to
affiliated regional media and spill over to broadcast media, I consider these
newsrooms the first potential instances in an informational cascade. Due to
this position among Brazilian news media, it is worthwhile to analyze the
overall distribution of these media outlets’ coverage of the scandals in the
Lava-Jato case and the impeachment. Their coverage patterns are likely to
be echoed in the vertically integrated media ecology. Such a quantitative
research approach is operationalized in the fifth chapter of this book where
I sample media material that, in total, forms the largest corpus of news on
corruption that has yet been analyzed in the Brazilian context: A
quantitative dataset of every news item and political commentary from the
web and the printed versions of Folha de S. Paulo and Estado de S. Paulo
dealing with national corruption cases during a seven-month period (from
October 2015 until the ousting of Rousseff in May 2016). The dataset
comprises 8,200 news items and metadata about each item. On top of this,
the fifth chapter also analyzes a dataset comprising six months of front
pages of both Folha, Estado and O Globo (October 2015 till April 2016),
with more than 1,300 items describing corruption and political
transgression.
Combined, these two datasets serve as proxy for the entire landscape of
Brazilian news discourse, and they cover with great granularity the period
until the impeachment of President Rousseff. The first dataset dwarfs the
many quantitative studies of Brazilian news media’s treatment of corruption
and politics produced in recent years, including those by Biroli and
Mantovani (2014) (n = 236), Makhoul (2009) (n = 252), Lemos and de
Barros (2016) (n = 351), de Paiva et al. (2017) (n = 932), and even Feres
Júnior and Sassara (2016) (n = 6,032). The size of this corpus compares to
the annual corpus in the monumental ten-year study (n = 107,248) by
Mancini et al. (2017) that sampled news items on corruption from three
different European media contexts. In other words, the sample of the two
São Paulo dailies in this study’s seven-month period outsizes the average
corruption coverage of three large European countries combined.

Plan of the Book

The book can be read as a tour of the different stages in the informational
cascade emerging in the public spheres of Brazil, triggered by the Lava-Jato
investigations. As the chapters progress, I move from the initial moments
where corruption is disclosed, via the particular accountability processes
that sustain the media attention, and on to the aggregated macro-level of
months of scandal and media exposés. In the final chapter, the empirical
consequences of the cascade, as well the theoretical contribution of this
movement, are summed up. For the sake of orientation, I will briefly
present the arguments of each chapter.
The next chapter provides an overview of Brazil’s media system,
historically and organizationally, with special attention paid to the vertical
concentration of media conglomerates, and the continuing political
importance of television and a few prestigious newspapers. The chapter
then briefly reviews the style of political journalism called denuncismo,
detected by researchers since the 1990s. This scandalizing and often
criticized off-the-records journalism is now gradually giving way to another
style, because the newsrooms covering the national political arena are
attuned to the workings of the accountability processes and official
investigations. Information flows from these processes provide the base for
a new way of writing news about political corruption, in which prosecutors,
magistrates, and police agents are positioned as protagonists in scandal
narratives. Bearing this in mind, I argue that corruption is a malleable
discursive construct, rather than merely a legal or objective fact, and that
the news frames are particularly important in defining what corruption
means. Framing theory, however, is not useful for the purposes of this book,
because salience transfer effects are unlikely to be encountered in this
context.
The third chapter analyzes the political repercussions of leaks of the
Lava-Jato investigations, by process-tracing the contribution of leaks to the
disintegration of the coalition government led by Dilma Rousseff. In this
chapter, I argue that the disclosure of corruption constitutes an especially
powerful and dramatic style of interaction between judiciary, political and
media systems. Leaks of information can be harnessed by political and
institutional actors to great effect, and in a fast-paced media environment,
leaks fundamentally erode journalistic boundary control and paradoxically
provide powerful (and potentially anti-democratic) access to the media for
those institutional actors with inside knowledge. These actors are
concealed, despite having vested political interests, while the media outlets
represent disclosure as ostensibly increasing transparency.
The fourth chapter describes the accountability institutions in Brazil
relevant in the Lava-Jato probe. On the basis of this, five case studies
demonstrate the complex interplay of political actors, prosecuting
authorities, and judicial institutions in court trials during the period from
2015 to 2017. The outcome of those accountability processes was not
purely the punishment of political transgression, as several trials ended in
impunity. Rather, several macro-level changes and various side-effects
surfaced in the political arena due to legal processes spawned during the
Lava-Jato investigations. In a discussion of the outcomes of the legal
processes, the chapter argues that political actors use legal action and
corruption denunciations strategically to wage political war through the
institutions of law and the repressive power of the state, which has been
termed “lawfare.” Lawfare encompasses the ways that legal (or quasi-legal)
discourse and processes are abused as instruments for subjugating political
subjects. Crucially, the mechanisms of the State, in this case checks and
balances if not accountability institutions, also drive the media agenda.
The fifth chapter introduces the concept of informational cascades and
adapts it to corruption scandals. An informational cascade is an aggregate
phenomenon that coalesces out of actions of political, economic, and media
actors and institutions. Based on a large-n sample of news items from
mainstream print and web media, a content analysis demonstrates how
information about corruption was shaped and passed on in cascades in the
public spheres of Brazil. From political elites, through mainstream media
and on to the stock markets of Brazil, the informational cascade repeated
the signal of crisis and instability and created the conditions for the ousting
of Rousseff.
The final chapter concludes the book by joining the strands of the
preceding chapters: Despite the principally distinct discursive fields
concerning political transgression, and the fact that Rousseff was not
investigated for corruption then, the media climate of the Lava-Jato scandal
created the conditions for replacing Rousseff with her vice-President,
Michel Temer. At that point, Temer had been implicated by five different
individuals in testimonies of the Lava-Jato probe. While the legality of the
proceedings was observed, I argue that the legitimacy of the impeachment
was thoroughly impaired as Temer and his party were participants in the
graft of the state companies and benefitted from the corruption. In the
conclusion, I answer the oft-asked question whether the impeachment could
be considered a coup d’état or a legit constitutional act, contending that the
legitimacy of the process was merely façade. Finally, I consider how the
lessons of the Lava-Jato probe, the leaks, and the lawfare may inform
current academic discussions on accountability, post-politics, and media
systems.

References
Abramo, C.W. (2007) “Brazil: A Portrait of Disparities” Brazilian Journalism Research, vol. 3(1),
93–107.
Allern, S. and Pollack, E. (eds.) (2012) Scandalous! The Mediated Construction of Political Scandals
in Four Nordic Countries. Göteborg: Nordicom.
Avritzer, L. (2016) Impasses da democracia no Brasil. Rio de Janeiro: Civilização Brasileira.
Balán, M. (2011) “Competition by Denunciation: The Political Dynamics of Corruption Scandals in
Argentina and Chile” Comparative Politics vol. 43(4), 459–478.
Bikhchandani, S., Hirshleifer, D., and Welch, I. (1998) “Learning from the Behavior of Others:
Conformity, Fads, and Informational Cascades” Journal of Economic Perspectives, vol. 12(3),
151–170.
Biroli, F. and Mantovani, D. (2014) “A parte que me cabe nesse julgamento: A Folha de S. Paulo na
cobertura ao processo do Mensalão” Opinião Pública, vol. 20(2), 204–218.
Boas, T. and Hidalgo, D. (2011) “Controlling the Airwaves: Incumbency Advantage and Community
Radio in Brazil” American Journal of Political Science, vol. 55(4), 869–885.
Bresser-Pereira, L.C., Oreiro, J.L., and Marconi, N. (eds.) (2014) Developmental Macro-economics.
New Developmentalism as a Growth Strategy. New York: Routledge.
Cardoso, G. and Fatima, B. di (2013) “Movimento em rede e protestos no Brasil. Qual gigante
acordou?” Dossiê Mídia, Intelectuais e Política, vol. 16(2), 143–176.
Chaia, V. and Teixeira, M. (2000) “Máfia dos Fiscais e as estrelas de cidadania” Observatório da
imprensa, ed. 63, Mídia e política, 31–37. Retrieved from
http://observatoriodaimprensa.com.br/primeiras-edicoes/mfia-dos-fiscais-eas-estrelas-da-
cidadania/.
Chaia, V. and Teixeira, M. (2001) “Democracia e Escândalos Políticos” São Paulo Perspectivas, vol.
15(4), 62–75.
Damgaard, M. (2015) “Narrating the Mensalão Trial: Configurations of Corruption” Brasiliana, vol.
3(2), 197–234.
Entman, R. (1993) “Framing: Toward Clarification of a Fractured Paradigm” Journal of
Communication, vol. 43, 51–58.
Falcão, J., Hartmann, I.A., and Chaves, V. (2014) III Relatório Supremo em Números: o Supremo e o
tempo. Rio de Janeiro: Escola de Direito do Rio de Janeiro da Fundação Getulio Vargas.
Feres Júnior, J. and Sassara, L. (2016) “Corrupção, Escândalos e a Cobertura Midiática da Política”
Novos Estudos, vol. 35(2), 205–225.
Figueiredo, A.C. (2010) “The Collor Impeachment and Presidential Government in Brazil.” In M.
Llanos and L. Marsteintredet (eds.), Presidential Breakdowns in Latin America (111–127). New
York: Palgrave Macmillan.
Ghanem, S. (1997) “Filling in the Tapestry. The Second Level of Agenda Setting.” In M. McCombs,
D. Shaw, and D. Weaver (eds.), Communication and Democracy: Exploring the Intellectual
Frontiers in Agenda Setting (3–14). Mahwah, NJ: Lawrence Erlbaum Associates.
Hendricks, V. and Hansen, P.G. (2016) Infostorms. Why Do We “Like”? Explaining Individual
Behavior on the Social Net. 2nd edn. London: Springer.
Iyengar, S. (1990) “Framing Responsibility for Political Issues: The Case of Poverty” Political
Behavior, vol. 12(1), 19–40.
Kucinski, B. (1991) Jornalistas e Revolucionários nos tempos da imprensa alternativa. São Paulo:
Editora Pagina Aberta.
Lattman-Weltman, F. and Abreu, A.A. de (2001) “Controles Midiáticos: Investigação e Denuncismo
na Construção da Visibilidade Pública Democrática.” Paper presented at ANPOCS 25th
conference at Caxambu, Minas Gerais, October 17, 2001.
Lazarsfeld, P., Berelson, B., and Gaudet, H. (1944). The People’s Choice: How the Voter Makes Up
His Mind in a Presidential Campaign. New York: Columbia University Press.
Lemos, C. and Barros, A. de (2016) “Lutas simbólicas na arena midiática: o poder de agência do
Ministério Público e as controvérsias sobre a PEC 37” Opinião Pública, vol. 22 (3), 702–738.
Lima, V.A. da (2004) “Sete teses sobre mídia e política no Brasil” Revista USP, São Paulo, no. 61,
48–57.
Lima, V.A. da (2006) Mídia. Crise Política e Poder no Brasil. São Paulo: Perseu Abramo.
Lima, V.A. da (2009) “Revisitando as sete teses sobre mídia e política no Brasil” in Comunicação &
Sociedade, vol. 30(51), 13–37.
Macaulay, F. (2011) “Federalism and State Criminal Justice Systems.” In T. Power and M. Taylor
(eds.), Corruption and Democracy in Brazil. The Struggle for Accountability (218–249). South
Bend, IN: University of Notre Dame Press.
McCombs, M. (2005) “A Look at Agenda-Setting: Past, Present and Future” Journalism Studies, vol.
6(4), 543–557.
McCombs, M. and Shaw, D. (1972) “The Agenda-Setting Function of Mass Media” Public Opinion
Quarterly vol. 36, 176–187.
Maier, J. (2011) “The Impact of Political Scandals on Political Support: An Experimental Test of
Two Theories” International Political Science, vol. 32(3), 283–302.
Mainwaring, S. and Welna, C. (eds.) (2003) Democratic Accountability in Latin America. Oxford:
Oxford University Press.
Makhoul, F. (2009) A cobertura da revista Veja no primeiro mandato do presidente Lula. Master’s
thesis, São Paulo: PUC-SP.
Matos, C. (2008) Journalism and Political Democracy in Brazil. Lanham, MD: Lexington Books.
Matos, C. (2012) Media and Politics in Latin America. Globalization, Democracy and Identity.
London: I.B. Tauris.
Mesquita, N.C. (2014) “Media and the Quality of Democracy: The Different Impacts of the Media on
Regime Support in Brazil.” In R. Figueiras, P. Espírito Santo, and I.F. Cunha (eds.), Democracy at
Work: Pressure and Propaganda in Portugal and Brazil (13–38). Coimbra: University of Coimbra
Press.
Michener, G. and Pereira, C. (2016) “A Great Leap Forward for Democracy and the Rule of Law?
Brazil’s Mensalão Trial” Journal of Latin American Studies, vol. 48, 477–507.
Miguel, L.F. (1999) “Mídia e Eleições: A Campanha de 1998 na Rede Globo” Dados, vol. 42(2),
253–276.
Miguel, L.F. (2003) “A Eleição Visível: A Rede Globo Descobre a Política em 2002” Dados, vol.
46(2), 289–310.
Miguel, L.F. (2015) “Quanto vale uma valência?” Revista Brasileira de Ciência Política, no. 17,
165–178.
Netto, V. (2016) Lava-Jato: O juiz Sergio Moro e os bastidores da operação que abalou o Brasil. Rio
de Janeiro: Editora Sextante.
Newman, N., Levy, D., and Nielsen, R.K. (2015) Reuters Institute Digital News Report 2015.
Oxford: Reuters Institute for the Study of Journalism, University of Oxford.
Paiva, A.L. de, Garcia, A., and Alcântara, V. (2017) “Disputas Discursivas sobre Corrupção no
Brasil: Uma Análise Discursivo-Crítica no Twitter” Revista Administração Contemporânea, vol.
21(5), 627–647.
Peruzzotti, E. (2006) “Media Scandals and Social Accountability. Assessing the Role of Scandals in
Argentina.” In E. Peruzzotti and C. Smulovitz (eds.), Enforcing the Rule of Law. Social
Accountability in the New Latin American Democracies (249–271). Pittsburgh, PA: University of
Pittsburgh Press.
Peruzzotti, E. and Smulovitz, C. (eds.) (2006) Enforcing the Rule of Law. Social Accountability in the
New Latin American Democracies. Pittsburgh, PA: University of Pittsburgh Press.
Porto, M. (2011) “The Media and Political Accountability.” In T. Power and M. Taylor (eds.),
Corruption and Democracy in Brazil (103–126). South Bend, IN: University of Notre Dame Press.
Porto, M. (2012) Media Power and Democratization in Brazil. TV Globo and the Dilemmas of
Political Accountability. London: Routledge.
Power, T. and Taylor, M. (eds.) (2011) Corruption and Democracy in Brazil. South Bend, IN:
University of Notre Dame Press.
Pujas, V. (2002) “Explaining the Wave of Scandal: The Exposure of Corruption in Italy, France, and
Spain.” In R. Kuhn and E. Neven (eds.), Political Journalism: New Challenges, New Practices
(149–167). London: Routledge.
Raile, E., Pereira, C., and Power, T. (2011) “The Executive Toolbox: Building Legislative Support in
a Multiparty Presidential Regime” Political Research Quarterly, vol. 64(2), 323–364.
Saad-Filho, A. (2013) “Mass Protests Under ‘Left Neoliberalism’: Brazil, June–July 2013” Critical
Sociology, vol. 39(5), 657–669.
Seligson, M. (2002) “The Impact of Corruption on Regime Legitimacy: A Comparative Study of
Four Latin American Countries” The Journal of Politics, vol. 64(2), 408–433.
Skidmore, T., Smith, P., and Green, J. (2010) Modern Latin America. Oxford: Oxford University
Press.
Smulovitz, C. and Peruzzotti, E. (2000) “Social Accountability in Latin America” Journal of
Democracy, vol. 11(4), 147–158.
Thompson, J.P. (2000) Political Scandal: Power and Visibility in the Media Age. Cambridge: Polity
Press.
Tumber, H. and Waisbord, S. (2004) “Introduction” American Behavioral Scientist, vol. 47, 1143–
1152.
Waisbord, S. (2000) Watchdog Journalism in South America. News, Accountability, and Democracy.
New York: Columbia University Press.
Wiewiura, J. and Hendricks, V. (2017) “Informational Pathologies and Interest Bubbles: Exploring
the Structural Mobilization of Knowledge, Ignorance, and Slack” New Media and Society, vol.
20(3), 1123–1138.
2 The Media System and Political Journalism
of Brazil

 
This chapter provides background knowledge for readers unfamiliar with
the Brazilian media system. First, the chapter sketches out the general
market structure of news media and the national media ecology. Such an
outline of the media system (Hallin and Mancini 2004, 2012) and its
relation to the Brazilian state cannot, however, sufficiently explain the
particular way that scandals emerge or come to be narrated in the sphere of
media, and I will therefore review the literature concerning Brazilian
political journalism and, in particular, the type of scandalizing coverage
known as denuncismo. On the basis of this review, I will furthermore
discuss the common framings of corruption found in scandal coverage and
argue that the denuncismo in political journalism is supplanted in the Lava-
Jato case by coverage attuned to official sources. This case and the
impeachment process were both hotly contested issues and were framed and
interpreted in a range of ways by various political actors. The largest news
media, however, all displayed a propensity for framing the corruption
scandals through the same set of frames, based on the prosecutor’s
disclosure of information. Because of that monolithic framing, this chapter
argues that framing is ultimately unhelpful as a methodological approach
for the study of this particular media system and scandal context. The
argument and the media system outline prepares the ground for the
following chapters.
I will briefly sketch out how the different genres and technologies are
located in the landscape of Brazilian news media, in order to show which
media are most important in political scandals, and to what extent these
media reach the population. Before turning to online news media and
television, I will start with the historical emergence of print media and the
limited national reach of newspapers.
 Brazil has had a considerably shorter history of a print media-based
public sphere than many other nation-states, because of the nation’s colonial
origins and because the Portuguese crown banned printing until 1808
(Azevedo 2006: 92, Sodré 1977). In consequence, a public sphere, in the
sense originally and powerfully analyzed by Habermas (1989 [1962]), did
not develop in Brazil in a way comparable to European public spheres.
Furthermore, slavery, poverty, and illiteracy stunted the growth of a market-
based system of mass print media for decades, and even in post-colonial
times, Brazil saw limited circulation of newspapers compared to other
Western media systems (Albuquerque 2010, Hallin and Papathanassopoulos
2002).
Perhaps it is not meaningful to even speak of a “mass press” in the
Brazilian case, given the low circulation relative to other nations. However,
television eventually emerged as mass media in Brazil in the 1960s, and its
reach overlapped and intersected with that of the Brazilian state, ruled by
the military dictatorship. When the television technology was introduced in
the 1950s, few had access and broadcasting did not significantly impact
political events. Following the coup d’état of 1964, the military leaders
came to see television as an instrument fit to edify and modernize the
nation, and placed the cornerstones of the modern media system by
investing in broadcasting infrastructure across the country. Thereby, the
generals also paved the way for the first TV moguls of Brazil. When
democracy was reintroduced in the 1980s, the Brazilian media system was
dominated by the broadcasting conglomerates that prospered during the
dictatorship, and the political elite, to this day, continues to have strong ties
to media conglomerate owners.
The alliances between media and groups within the political elite is
characteristic of the Brazilian media system’s history. Newspapers were tied
to individual political projects in the first half of the nineteenth century, and
the crown maintained prerogatives for restricting freedom of press (Sodré
1977). In the second half of the nineteenth century, (relatively) large-scale
industrial print runs of newspapers emerged (such as Estado de São Paulo,
founded in 1875 as A Província de São Paulo; Jornal do Brasil, founded in
Rio de Janeiro 1891; and Correio do Povo, founded in Porto Alegre in
1895). O Globo and Folha de São Paulo (then Folha da Manhã) were both
founded in 1925 (Azevedo 2006: 93). The relative circulation of such
newspapers reached an apogee in the 1950s but declined soon after, so that
“[i]n the early fifties, the newspaper industry in Brazil was selling 110.6
copies per 1,000 people, a penetration figure that, by the end of the 1960’s,
had declined to 45.4 copies” (Carro 2016: 4).
The period of the military dictatorship (1964–1985) had other important
newspapers than the above-mentioned, such as Jornal do Brasil, Jornal do
Comercio, A Notícia, and Correio da Manhã (Sodré 1977), and weeklies
such as Revista da Semana, Fon-Fon, O Cruzeiro, and Manchete. In face of
censorship, an alternative press also developed, with O Pasquim leading the
pack ahead of Opinião, Repórter, Movimento, and Pif-Paf, among others
(Kucinski 1991). While a few survived the 1980s (O Pasquim until 1991,
Manchete until 2000, and Jornal do Brasil until 2010 as print, continuing as
a digital publication), the structural conditions of the media system during
the dictatorship were obviously different than the period after the transition
to democracy.
Simultaneous with the transition to democracy, the market of news
shifted decisively toward television. UNESCO figures of newspaper
circulation in 1996 were 40 per thousand Brazilians, the lowest of the
countries considered by Hallin and Papathanassopoulos (2002: 177). By
2000, it was up to 60.6, but in 2009 down to 45.3 newspapers sold per
1,000 adult citizens (Albuquerque 2012: 78). Such a penetration ratio of
print media of course depends on the population size, which complicates
the picture. In the two decades from 1990–2010, while the economy, the
middle-class, and the literacy rates grew, and while the newspapers doubled
their print runs, the booming Brazilian population outgrew the newspapers:

[f]uelled in part by this new-found readership, paid circulation

expanded at an average year-on-year growth rate of 4.42 percent
between 1991 and 2000. However, the relatively fast pace of expansion
achieved during the 1990’s, would decrease significantly through the
following decades. Through the 2000’s and 2010’s (so far), the
country’s population increased at a much faster pace than the paid-for
dailies circulation. In fact, between 2011 and 2014, circulation shrunk
at an average rate of 4.31 percent year-on-year, while the number of
inhabitants in the country continued to rise. Consequently, newspaper
readers are becoming an increasingly smaller share of the Brazilian
population … [Circulation figures] went further down to 36.7 copies in
2014.
 (Carro 2016: 4)

Ratios between population and newspaper print runs notwithstanding,

the recent absolute decline of print is visible when comparing figures from
IVC (the Brazilian Institute for Verification of Circulation), in five-year
intervals (2005, 2010 and 2015). The latest yearly figures have also been
included in the table below.
The latest years of decline of already low-circulation papers have left
Brazil with three traditional, prestigious, quality newspapers with national
reach: Folha de São Paulo and Estado de São Paulo (located in the
ideological space between liberal and conservative viewpoints), and the
Rio-based Jornal O Globo, culturally liberal but orthodox in its economic
editorial opinions (Fonseca 2005, Lattman-Weltman and Chagas 2016). In
practice, relatively few newsstands in Rio de Janeiro offer the São Paulo
papers and vice-versa. The economy-oriented Valor Econômico, created in
2000 by Folha and the Globo group (but recently bought out by Globo),
accompanies this trio nationally, with a distinct appeal to the business elite
(thus, it has considerably lower circulation of 40,000 printed copies). In an
already small market by global standards, the print runs of the national
newspapers have in effect been cut in half in the last two decades. The São
Paulo-based newspapers have also cut staff lately in their newsrooms in
consequence of the revenue loss from direct sales and the accompanying
declining revenue from advertisement sales (Becker and Waltz 2017: 119,
note 9).
Table 2.1 Print runs of Brazilian newspapers, 2005–2016
Newspaper 2005 2010 2015 2016
Folha de S. Paulo 309,383 295,558 175,441 154,700
O Globo 276,385 257,262 183,404 181,000
Extra 267,225 248,119 136,000 128,000
Estado de S. 231,165 212,844 149,241 146,000
Paulo
Zero Hora 174,617 No data 144,191 122,365
available

Sources: 2005 figures reported in Albuquerque 2012: 78; 2010 figures

reported in Porto 2010: 113; 2015 and 2016 figures reported in Instituto
Verificador de Circulação 2016. Figures for Extra in 2015 and 2016
reported in Rocha Filho 2016.

In addition to the trio of nationally distributed newspapers, many daily

newspapers are printed and circulated in regional capitals or states. These
newspapers are often purely advertisement-financed, or very subsidized,
and are referred to as “popular papers” by most scholars, because the focus
of these papers are sports and celebrity content (Lattman-Weltman and
Chagas 2016: 341). The landscape of regional newspapers changes
frequently, and in 2016 comprised Extra, Super Notícia, Meia Hora, Zero
Hora, Aqui, Destak, Agora São Paulo, Diário Gaucho, Expresso and Metro
São Paulo. Though these papers account for a large part of the total national
circulation, they rarely produce content about national politics themselves,
but instead reproduce material from the news agencies of Folha, Estado and
Globo (Abramo 2007). For that reason, the regional and popular papers are
not included in the analyses of this book.
Despite the relatively large print runs of the market leaders among the
popular newspapers, the weekly magazines of Brazil are probably the only
print media which could be properly termed “mass” media. While the daily
newspapers of Brazil in their (relatively late) heyday reached hundreds of
thousand readers, the weekly magazines (especially Veja) have, at various
points of modern media history, seen print runs of more than one million
copies. Veja was founded in 1968, published by the Abril group, in the
format of the US weekly magazine Time. The Veja editor and founder Mino
Carta left the magazine in 1976 to launch IstoÉ (publicized by the Editora
Trés [Porto 2010: 114]) as a competitor to Veja, and the Globo
conglomerate launched a weekly magazine (Época) as another rival to Veja
in 1998. For the last decades, the three have been the dominant weeklies,
with political content inclining toward anti-PT views (Biroli and Miguel
2012: 39ff.).
On the opposite wing of the political spectrum, a few weekly and
monthly magazines provide a counter-weight to Veja and the others. Mino
Carta left IstoÉ in 1981 and a decade later founded yet another magazine,
the CartaCapital. CartaCapital started as a monthly magazine in 1994 but
has been weekly since 2001 and caters to a left-leaning audience in contrast
to Veja, Época, and IstoÉ. In 2014–2016, CartaCapital had only a few
competitors for that target audience, most prominently the printed
magazines Caros Amigos and Revista Piauí, and the purely digital news
sites with prominent blogs such as Brasil 247, Jornal-GGN, and
CartaMaior (Carvalho and Albuquerque 2017).
Both the weekly magazines, the large national newspapers, and the main
broadcasters also mirror their printed content online, albeit to different
degrees. Currently, the news sites based on print newspapers and magazines
all employ more or less the same business model: Reading a limited number
of articles each month is free of charge for readers, but beyond that, each
site maintains a paywall to avoid cannibalizing on their print products. The
number of print subscriptions has recently been declining (since 2014 this
trend has become visible across the board, see Instituto Verificador de
Circulação 2016), and Folha paywall subscribers are now more numerous,
for the first time ever, than the print subscribers (Folha de S. Paulo 2016).
The current market leader of the online news media is, if one tallies up
visitors across websites, Globo. The conglomerate hosts the online news
service G1 and the web version of GloboNews (the conglomerate’s 24-hour
news channel). Both reproduce, or “shovel” (a term defined as
“repackaging content produced for other media” [Paterson and Domingo
2008: 7]) material from their news broadcasts onto the online platforms.
The journal O Globo also shovels most material onto the site
oglobo.globo.com, and the conglomerate’s weekly magazine Época does
the same. The most-visited Globo platform is the G1 site (Newman et al.
2015, 2017) which mostly relies on content about politics produced in
Brasília and affiliated regional newsrooms. Many political news items on
G1, especially commentary spots, are simply edited material from the
Globo news program Jornal das Dez or from the evening broadcast Jornal
Nacional.
Likewise, the main competitor of Globo, the broadcaster Rede Record
produces shovelware on the R7 online platform by recycling material from
their repertoire of programs. A similar pattern is true of the national market
leader, UOL. The UOL platform has been the most visited Brazilian site in a
decade (only surpassed in visits by Google, Facebook, and YouTube).
Founded back in 1996 by the Grupo Folha, UOL was an early player on the
stage. The political news of the UOL site is provided by direct links to the
online version of Folha de S. Paulo. Valor Econômico and Estado de S.
Paulo, like the rest of the national papers, maintain news sites with material
mirrored from their print versions, but to some degree extend this by
providing additional material that gets cut in the print version.
Possibly because of the links to UOL, Folha also boasted the largest
number of page views in the peak period dealt with in this book, March
2016. According to IVC research, Folha reached just below 100 million
page views in each week of March, with an apogee of 155 million weekly
page views for the third week. The closest competitors that week were
Estado’s news page Estadão, at 47 million page views, ClicRBS at 42
million, and O Globo at 39 million. After UOL and Folha, the most-
accessed internet news sites were ones tied to the Globo conglomerate (in
addition to the O Globo site, the news site G1, that used to be market-
leader) (Newman et al. 2017: 107).
There are few born-online news media platforms in Brazil with access to
the circles of political power, but a number of journalists have set up
smaller news sites for political news, gossip, and punditry. This kind of
entrepreneur journalists are usually former employees of the above-
mentioned mainstream papers. Most notably, two former leading Veja
journalists started the blog O Antagonista with specific focus on the Lava-
Jato case, and they have managed to break news ahead of mainstream news
a few times. The left-wing allied blogsites Tijolaço and Jornal GGN,
through their political ties, also have access to sources within the political
elite. In addition to these journalistic entrepreneurs, a few online media sites
aggregate and produce news with political content that can claim national
outreach: The prolific alternative news page Midia NINJA excel at
grassroots reporting, commentary, and networked journalism, but although
important in the younger audience bracket, this alt-news channel has not
been in the leak-loop and does not get access to any political scoops
(Becker and Waltz 2017). The American web media outlet BuzzFeed has
lately been investing in the Brazilian market, setting up a local newsroom
and working with the aim of getting traffic to the site through original
journalistic content. What is true of other media contexts (Harder et al.
2017, Welbers et al. 2018) is also true of Brazil: legacy media outlets are
able to influence the online news agenda to a considerable degree.
Radio is a less-used and less-researched news source in Brazil, but it still
survives as a “niche” in certain segments and rural areas of Brazil
(Meneguello 2010). Compared to other countries in a survey conducted by
Reuters in 2014 (Newman et al. 2015), the Brazilian consumption of radio
as a news source was neither high nor low: 39 percent of urban Brazilians
say they get some news from the radio. Because of the regional limitations
of the radio industry, radio as a news media has not been considered
important in the literature dealing with contemporary politics (with
exceptions such as Chaia 2001). Historically, radio stations were linked to
regional political leaders, earning them the nick-name of coroneis
eletronicos (Lima and Lopes 2007, Nunes Leal 1997, Motter 1994),
although this term also refers to the owners of broadcast media.
As mentioned, the audience penetration of televised news surpassed that
of printed news. As television, locally launched in 1950, grew into the mass
media of the 1960s and 1970s, the military authorities actively subsidized
and groomed the growing television industry as a nation-building,
ideological instrument, and broadcast news became the primary source of
popular knowledge about the political arena (Biroli et al. 2011, Biroli and
Miguel 2013, Lima 2004, 2009, Miguel 1999).
TV Globo only had its hegemony challenged in the early 1980s (Matos
2012a, Porto 2012: 70) with Silvio Santos’s Sistema Brasileiro de Televisão
(SBT) and eventually, in the 1990s the Rede Record entered the scene,
gradually conquering larger shares of viewership (from 4 percent of the
evening slot audience in 1996 to 17 percent in 2008 [ibid.: 72]). Globo has
remained comfortable at the top of the broadcast market for decades, while
TV Record and SBT have been vying for second place (Porto 2010: 111).
But despite this continued dominance of Globo after the transition to
democracy, in the 25 years from 1989 to 2014, the media market has
changed substantially: The broadcast market grew both in terms of
producers and diversity, with satellite and cable services entering the market
after the Cable Law of 1995. Furthermore, all of the main Brazilian outlets
allied with multinational media groups in this period: Grupo Abril (that
publishes Veja and a range of other magazines) teamed up with MTV,
Rupert Murdoch’s News Corporation partnered with Globo, and Folha’s
online portal UOL partnered with BBC (Azevedo 2006: 1010, Matos
2012b: 866).
Apart from the three national broadcasters, five groups are dominant in
the five regions of Brazil: RBS (Rede Brasil Sul) to the South;
Organizações Jaime Câmara and the regional TV Anhanguera in the Central
Brazil and Eastern regions; the network Rede Amazônica de Rádio e
Televisão in the Northern part of the country; the Grupo Zahran is
predominant in the large states of Mato Grosso and Mato Grosso do Sul;
and, finally, the media market of northeastern Brazil is dominated by the
Grupo Verdes Mares (Cabral 2017: 51).
In comparison to other media types, television today still maintains the
position as central news source for the population. This is (surprisingly) the
case even for the urban-dwelling Brazilians with relatively good access to
internet: In the survey from 2014 (n = 2033), 91 percent of the respondents
stated that they accessed news online in the past week, but 81 percent also
accessed news from TV, while radio and print news was mentioned by 39
percent and 33 percent of the respondents, respectively (Newman et al.
2015: 51). In this group, 54 percent of the respondents self-identify as more
digital news consumers than consumers of traditional radio/TV/print, while
12 percent define themselves as mostly relying on traditional news sources
(which puts Brazil as the country with fewest traditional news consumers in
the group of countries surveyed by Reuters). So, even in urban Brazil,
which compared to other surveyed populations contain more “digital” than
“traditional” news consumers, reliance on TV is still almost the world’s
highest. Among the surveyed countries in the 2014 survey, Brazil occupied
the last position in relation to internet access, having 54 percent penetration
of internet for the entire population in 2014, which grew to 68 percent by
2017 (Newman et al. 2017: 106).
Even today, with this rapidly rising rate of access to internet, television
reaches outskirts of Brazil where few broadband cables run and no cellular
operator has yet put up an antenna. Because of the geographical challenges,
reliance on TV for news can thus be assumed to be even higher in the vast
rural areas, since broadband coverage is scarce, print media circulation is
low and, in some places, hard to come by. Partly for these reasons, 60
percent of the total advertisement expenditure in Brazil is still allocated in
television, and for similar reasons, researchers such as Porto have
previously concluded that “broadcast television remains the dominant
medium” (Porto 2010: 110). Though this situation may change in response
to the growing consumer demand for internet access and government
policies to promote it, media studies of Western countries with already
high-choice media environments have shown (e.g., Harder et al. 2017,
Shehata and Strömbäck 2013) that internet media platforms still attend to
the agenda put forth by legacy media, and that this agenda thus maintains
strong links to the public agenda of surveyed users, despite individual
consumption of media being more fragmented than in earlier decades.
The evening news Jornal Nacional of TV Globo has remained the major
newscast of the national television market (Biroli and Miguel 2013: 88),
usually with two to three times the audience of Jornal da Record and about
five times the audience of Jornal do SBT (Ribeiro 2017: 113). Jornal
Nacional, in 2015–2016, was usually the most watched or the second-most
watched program among all Brazilian programs, in a field dominated by
popular (semi-global) entertainment formats (e.g., Big Brother) and the
Brazilian mainstay genre of telenovelas. Furthermore, TV Globo reigns in
the domain of 24-hour news service with their GloboNews channel.
GloboNews is paid cable news, however, and because the competing
Record News is available on open frequencies, Record reaches a larger
audience amongst the dedicated news channels.
The complex and non-linear interplay of the technological and
institutional developments of the Brazilian media since the transition to
democracy has thus yielded a system of news media dominated by a few
powerful conglomerates, with much vertical as well as horizontal
integration between different regional outlets and different media types. The
political arena is systemically closed off to all but a few dozen journalists,
working in the Congress, in the Presidential palace, or in the political beat
of the major state capitals such as Rio de Janeiro and São Paulo. Because of
this, only a few newsrooms control the gates of the political system, and the
information admitted through these editorial processes propagates through
networks of news agencies and inter-media shoveling. The limited range of
news agencies and organizations are owned by only a handful families, in
turn, and this certainly poses barriers to the kind of material that can be
published or broadcast by these conglomerates, especially in cases where
owners are aligned with specific advertisers or political elites (and here, the
examples are many; see Lima 2004, Matos 2012a: 52ff.). Such
considerations also restrict editors, whether by direct order or by indirect
pressure. In sum, Brazilian journalists refer to limited number of editors,
and, ultimately, the editors of the largest media outlets have only five
masters. The concentration of media is an important background condition
for the emergence of intra-elite strategies that utilize media exposés and
scandals for political advantage.

Denuncismo and Scandals in the Political Journalism of Brazil

The end of censorship was without a doubt a major step toward a
democratic public sphere in Brazil. However, accountability is not
accomplished simply because of press freedom. In this section, I will
discuss the contribution of media exposés in the production of
accountability in Brazil, and review the problems identified in the literature
on media disclosure of corruption. The production of scandals as a political
weapon is introduced here, and the term denuncismo and the style of
political reporting associated with denuncismo is discussed.
Due to the military censorship, corruption scandals were often subdued
in the political news of daily papers and radio in the 1960s and 1970s. Even
so, large newspapers sporadically “criticized the authoritarian regime and
even investigated cases that involved government officials, but watchdog
journalism remained the exception” (Waisbord 2000: 25). Some of the
notable exceptions to the rule of silence in this period included exposés
concerning the construction of Brasília and the Rio-Niteroi bridge. As João
Figueiredo, the last general-president, relinquished control, and José Sarney
stepped up as the first civilian president (taking over from Tancredo Neves,
who died as president-elect, on the verge of being inaugurated), scandals
began to fill front pages of Brazilian newspapers. This did not necessarily
mean a greater commitment to democratic ideals, however; with the gradual
Abertura (opening), denunciations in the press served as a political
instrument rather than social accountability mechanism. Thus,
 Watchdog journalism implied a commitment to determined political
causes rather than to the democratic ideal of checks and balances. It
responded to the political orientations of owners and reporters and
functioned as a strategy to advance specific agendas. It did not intend
to wrap itself around the flag of balanced reporting and the defense of
public good.
(Ibid.: 26)

The first free and direct presidential elections after the military rule, in
1989, highlighted the theme of anti-corruption. The winning candidate,
Fernando Collor de Mello, had projected an image as a “Hunter of
Maharajas,” fighting the fat cats of the corrupt public bureaucracy during
his campaign, but, ironically, the Collor administration fell victim to
investigative journalism through disclosure of corruption. Exposés in Veja,
IstoÉ, and Folha eventually brought enough scathing evidence to light to
trigger impeachment proceedings against Collor. Matos refers to this era as
the “heyday of Brazilian investigative journalism,” when newsroom began
establishing structures for investigations (Matos 2012a: 108), although
journalists still relied to a great extent on “promiscuous relationships” with
inside sources with political motivation.
Such promiscuous relationships between political sources and journalists
in the political beat has been termed denuncismo by several authors
(Lattman-Weltman and Abreu 2001, Waisbord 2000: 103ff.). This particular
style of producing political scandals is distinct from investigative
journalism as it

puts in evidence the proximity of the press to government sources. The

political clout of a few sources, rarely quoted or only cryptically
mentioned, is often sufficient to print stories, making it unnecessary to
comb other potentially knowledgeable parties or to search for
alternative sources of information.… [Critique alludes] to news
organizations that claim to practice investigative journalism based on
groundless accusations.… The lack of solid evidence undermines the
credibility of denunciations.
(Waisbord 2000: 103–104)
 In its most extreme variant, this style of political journalism is based on
off-the-record statements from sources who often have considerable interest
in the production of scandalous news items. Denuncismo journalism may be
dressed up to look like investigative journalism, but without the in-depth
research and the critical eye to the vested interests of sources. Thus,
denuncismo coverage may simply be the product of hasty “desk
journalism,” such as a single rumor or a newsroom phone call turned into
an article without much substantiation.
The weight of denuncismo, on one hand, and investigative journalism,
on the other, varied historically and according to the segments of the
Brazilian media. The weekly press has increasingly given space to
denuncismo over the last decades (Nascimento 2013), but the use of off-the-
record source material is certainly not restricted to that sector of the media.
As demonstrated amply in the Lava-Jato case, a denunciation may gain
much traction with the media when it is transformed and recycled in various
institutional forms, especially when a denunciation initiates legal processes
in the accountability institutions. In this way, slander and unsubstantiated
denunciations may be perceived to get reinforcement through legitimate
institutional processes, which allows journalists to continue reproducing the
original denunciation without questioning the underlying motives of the
denunciation.
After the heyday of scandals under Collor, Brazil witnessed a decline in
the force of scandals relative to the political system:

Political changes would only come about with “scandals,” which were
difficult to produce as they required a continued exposition of the
denounced facts, with the constant addition of revelations and new
elements. Above all, the changes required the national media, highly
oligarchic, to act as a filter. The emblematic episode of 1993 was the
Budget scandal, in which public funds were siphoned off through
parliamentary amendments to the state budget. The scandal resulted in
cassations of political mandates, renunciations as well as acquittals. It
also resulted in a reform of the state budget process and the law of
budget amendments. In time, even this already limited tool of
denunciation lost force. As the opposition leader, PT reoriented its
strategy and came to prioritize … electing Lula as president over social
mobilization.… In the Cardoso administration, few denunciations
 really lived to become scandals that could provoke significant change
in the system. The production of scandals began to depend on open
confrontation between allied political forces. Denunciations only
prospered and had institutional consequences when utilized by allies in
public disputes.
(Nobre 2013: 9, translation mine)

For example, the “Senhor X” scoop of Folha in 1997 revealed bribes in

the passing of the re-election Constitutional amendment but did not result in
charges or trials. The amendment paved the way for Cardoso’s second
mandate as president (Waisbord 2000: 42, 136). Like other scandals of the
mid- and late-1990s, it is thus not included in the list of prominent political
corruption scandals at the federal level compiled by Power and Taylor
(2011: 2–3), perhaps because the denunciation was obstructed and
eventually filed away. Nobre, in the quote above, suggests that this has to
do with the role of PT in the political system at this moment, and the
strategy of the party. In this period, PT grew to become the largest party in
Congress, and while PT had usually engaged in denouncing political
opponents, upon conquering federal power in Lula’s fourth bid for the
presidency in 2002, the tables turned against the party.
Though Lula, in the beginning of his first term as president, managed to
establish a truce with most mainstream media (Kitzberger 2014: 26), the
weekly Veja remained staunchly opposed and chose to radically scandalize
the PT. With the onset of the Mensalão scandal in 2005 (Damgaard 2015a,
Michener and Pereira 2016), a landslide of media scandals hit PT and the
negative coverage continued throughout the 2006 presidential campaign, as
TV Globo predominantly framed the campaign in terms of PT’s scandals
(Mesquita 2014, Porto 2012: 95). Nascimento (2013) reviewed the
denunciations publicized in Veja and the two competing weekly magazines
IstoÉ and Época in the six months preceding each campaign season of
presidential elections (in the period 1989–2010), and he concluded “that the
average number of stories containing denunciations almost tripled since
2002” (Nascimento 2013: 68), going from a weekly total average of 0.3
denunciation items to 1.3 and 1.5 in the 2002–2010 period (ibid.: 72). The
focus on corruption scandals after the Mensalão scandal was not restricted
to Veja or Globo, as other journals such as Folha eventually joined the
bandwagon (Biroli and Mantovani 2014: 207).
 The charge of denuncismo is still heard in Brazil today, but the problem
of the source-journalist relations and facile accusations has been relocated,
in a sense. Since the Mensalão scandal, another type of textual device than
the anonymized source has been taking center stage in corruption coverage:
the official document. In the prestigious newspapers’ coverage of the Lava-
Jato cases, institutional sources and the paper trail of official investigations
are exceedingly prominent. This is perhaps the media professionals’
institutional response to the charge of denuncismo, or perhaps it is a
tendency fostered because of accountability institutions’ communication
strategies, as discussed in the next chapter.
The increased focus on corruption scandals, especially in the PT
administration, has possibly come about as a strategy to maintain
competitiveness for the Brazilian news media. As described above, the few
national newspapers left have faced pressure by web media and various
popular and free newspapers in metropolitan areas. To counter the
economic pressures on this changing news media market, the three large
daily papers have strengthened their political profile in the last decade or so,
bringing about what Lattman-Weltman has termed repartidarização – the
resurgence of the party press (Lattman-Weltman and Chagas 2016). With a
stronger focus on political news, Estado, Folha and O Globo have tried to
distinguish themselves from newspapers with more focus on celebrity,
sports and quotidian news. This move has been to secure the share of the
news consumers actually willing to pay for quality news, since the
advertisement revenue of quality newspapers is moving toward internet and
ad-financed free papers (Carro 2016: 20). In this media ecology, the
production of scandals has become quintessential, and I will now briefly
explore how denunciations of corruption are shaped and framed textually in
Brazilian scandals.

Writing and Framing Corruption Scandals

If investigative journalism and denuncismo, at different times, became
distinct but real forces in the Brazilian political landscape, it is necessary to
have a closer look at the way such denunciations of corruption play out in
media texts. We need to ask what actions are denounced as corrupt, and
what kinds of acts are excluded from the definition of corruption? These
questions take us into the territory of media studies known as framing
theory, because corruption cannot be defined a priori, outside of the cultural
context – nor, in the study of a scandal, outside of the media context.
Information about political action is not simply information about either
legitimate politics or corruption. Rather, some aspects of political action
come to signify transgression of norms through particular textual strategies
that highlight some perspective upon the issue. This means that
“corruption” is not a static term but continuously negotiated discursively
and legally.
The distinctions of what political corruption “is” differs from society to
society. Indeed, the concept of corruption has evolved for millennia: In
antique and classic formulations (such as Polybius or Machiavelli)
corruption was conceived as particular state forms or regimes. More
recently, political theory of the last four decades (originating with Nye,
1967; see Lennerfors, 2008) frequently defines corruption through a
distinction between public and private roles. But the precise delimitation of
public interest, collective interests, and private interests is always
negotiated and contested, even in the discourse of apparently hegemonic
media outlets. Thus, the concept of “corruption,” depending on cultural
norms, values, and negotiation models, cannot easily be subtracted from its
context. Political corruption is not simply “out there,” it is something
defined and negotiated by a multitude of actors – by the law, but also by
convention, practice, pragmatic concerns, language, and socialized norms
of both the media system and the political system in question. Whether or
not an act is “really” corrupt is of course a relevant question to some
disciplines, law foremost of these. However, legal definitions of corruption,
important as they may be, are often eclipsed in the modes of critique
springing from intense coverage that characterizes scandals (Damgaard
2015). Power and Taylor (2011: 13) have similarly argued that research into
corruption and illegality walks a fine line, especially in Brazil, and that it is
necessary to account for both formally corrupt behavior as well as
questionable, informal practices.
In this book, I set out from the premise that processes of negotiating
meaning in scandals start with media’s disclosure of corruption. When
disclosed information indicates corruption, a journalist and his superiors
(copy editors, headline and web editors, etc.) have to make a number of
choices concerning their framing. They must consider the veracity of the
information, the reliability of the source, the possible defense put up by the
accused; in other words, the gatekeepers of the media consider varying
versions of a contested reality. Media professionals consider disclosed
information about corruption on the basis of background assumptions about
what corruption “is” and the particular framework or understanding of
politics underlining the media system into which they are embedded. Thus,
news texts in corruption scandals normally work by drawing upon a well-
established discursive formation wherein certain kinds of political action
are considered transgressive or corrupt. By invoking these existing
discourses of corruption, the coverage of a scandal also produces the reality
of corruption, through “hard” news as well as public debate, public
condemnations of corruption, and defenses of political actions. Together, all
of this assembles a media event, which in and of itself can influence
political reality (Breit 2011). As stated by John Fiske,

in a postmodern world we can no longer rely on a stable relationship or

clear distinction between a “real” event and its mediated
representation. Consequently, we can no longer work with the idea that
the “real” is more important, significant, or even “true” than the
representation. A media event, then, is not a mere representation of
what happened, but it has its own reality, which gathers up into itself
the reality of the event that may or may not have preceded it.
(Fiske 1994: 2)

Although I will refrain from using the concept of media event here, the
insight of Fiske is an important condition of this book. Like many other
studies (Anders and Nuijten 2009, Breit 2011, Damgaard 2015, Gupta
1995, Kajsiu 2014, Koechlin 2013, Lennerfors 2008), this study of
corruption scandals is essentially socially-constructivist, and even relativist
in its approach to corruption. This means that I will refrain from a priori
judging whether one or the other act was corrupt, but instead focus on the
allegation of corruption, on the denunciation, and the system-level results of
this discursive activity. That move locates this book within the current of
social constructivism, although it deals with political events from a realist
position. This constructivist perspective means that there is more sense in
registering the particular cultural repertoire of actions encompassing
corruption in the actual discourse than in philosophical or political theory,
thereby avoiding embedded academic ethnocentrism in the delimitation of
the research object. Within this constructivist perspective, the corruption
definition used here is entirely pragmatic and relational, rather than
universal and substantial.
I will argue that a news text may establish various frames that convey an
assessment of some political action, but if the assessment is to be
understood as scandalous by audiences, the news text must indicate a
transgression. Even when not explicit, such labelling can be detected
intertextually or in the interpretations offered by sources presented in the
news text. This could be prosecutors, whistle-blowers, or commentators of
various kinds, calling the transgressive action into question, representing it
as morally debased and detrimental to one or more societal sub-systems
(which could be political-institutional systems, the economy, the rule of
law, or national development) (Damgaard 2015: 414). So, drawing upon the
specific local discursive formation, which among other things outlines what
concepts such as “politics,” “public,” and “private” signify, news texts then
label a particular action as corrupt.
Labelling actions as corrupt brings me to the concept of framing. In
newsrooms as well as in journalism studies, the intentional selectivity
involved in packaging information for a media text is conventionally
termed “framing.” That is, both researchers and practitioners explicitly
work with frames, albeit in distinct ways. A news media frame can be
considered as both a “cognitive device used in information encoding,
interpreting, and retrieving” and as a “strategy of constructing and
processing news discourse” (Pan and Kosicki 1993: 57). In a classic
definition proposed by Robert Entman, framing means to “select some
aspects of a perceived reality and make them more salient in a
communicating text, in such a way as to promote a particular problem
definition, causal interpretation, moral evaluation and/or treatment
recommendation for the item described” (Entman 1993: 52, emphasis
removed). The academic perspectives on framing range from cognitive to
constructivist approaches across discourse and symbolic environments.
Some scholars therefore broaden Entman’s definition, viewing frames more
broadly as “structures that draw boundaries, set up categories, define some
ideas as out and others in, and generally operate to snag related ideas in
their net in an active process” (Reese 2007: 150). Entman and other
researchers (e.g., Sniderman and Theriault 2004) predict the relationship
between media frames and audience opinions. I leave the question of
frames and opinion for Chapter 5, in order to continue considerations of the
discursive representations of corruption in Brazilian news texts, especially
the definition and attributes made salient.
In the Lava-Jato scandal, the standard reporting pattern found in the
sampled Brazilian newspapers (as well as in broadcasting news) puts an
emphasis on legal processes. Starting with the lead of the article, or the first
sentence presented by a news anchor, the average news item assembles a
story from three elements: a politician under suspicion, an accountability
process in a legal arena, and a judicial or investigating actor. The basic lead
or headline reads “Politician X has been indicted by the public prosecutor Y
before Court Z in the Lava-Jato.” While the pattern has innumerable
variations, it underpins a significant number of news items.
In terms of information flows, what lies behind the emergence of this
standard framing of corruption is not exactly the personal intervention of
prosecutor Y, but rather that various accountability actors have begun
publicizing much information about investigations and legal processes and
feeding this directly to the inbox of journalists. Transforming tedious
processual information into readable news stories, journalists will often
utilize the metonymic relationship between specific institutional actors (say,
Prosecutor-General Rodrigo Janot) and the accountability institution
moving along some legal process or other. Individual political actors are
normally the targets of such processes in consequence of the criminal code,
while only a few processes charge the political parties of Brazil. Since the
coverage of corruption is largely based on legal processes, the majority of
the coverage frames corruption as an individual, rather than systemic
problem. This also holds true for news items concerning denunciations that
have cropped up in testimonies and plea bargains.
The journalistic framing replicates the indictment strategy of the
prosecuting authorities. A similar pattern was found in Folha’s coverage of
the Mensalão trial, where Biroli and Mantovani (2014) concluded that
although the newspaper included scattered points of view opposed to the
charges of the prosecutors, such as the defendants’ reactions and the
defense lawyers’ arguments, Folha only presented the events of the trial
through the framing of PT as the perpetrator and mastermind behind the
scandal – mirroring the indictment.
 The narrative projected by Brazilian news on corruption can often be
reduced to this simplistic schema of suspicion or denouncement being
contested in various arenas – courts and politics foremost – and this yields
what can be termed a responsibility frame in the sense proposed by Iyengar
(1990). Within Iyengar’s variation of framing theory, one would predict that
this kind of news is likely to be construed by audience members as
signaling the individual culpability of the mentioned politician. In theory,
this kind of framing would make the personal responsibility of political
actors salient, rather than the systemic problems of impunity and incentives
for rent-seeking in the political systems.
Opposing frames exist, of course, in the entrepreneurial media outlets
and even in the mainstream media, but they are rarely found here as the
main framing device. The political importance and use of such conflicting
frames (or counter-frames) has been explored in non-scandal contexts (e.g.,
Chong and Druckman 2013). Classic experimental studies have shown how
framings can be used to radically alter audience responses to messages with
the same informational content (Kahneman and Tversky 1979). Although it
would be possible to study the relative weight of frames in the sample texts,
researchers can hardly gauge the effect of the different frames found in the
media, because it is not possible to artificially isolate the audience
(Sniderman and Theriault 2004).
Summing up, the frames deployed by journalists and editors of the
mainstream Brazilian media treated in this book are structured around the
prosecutor’s perspective. This means that what is defined as corrupt and
what is left out of that category mostly hinges upon the legal processes and
official documents. While opinion pieces and brief spurts of investigative
journalism may sometimes expand the vision of the political arena,
counterframes are less frequent. Because of this, it would be difficult to
structure research around the effects of frames, since audiences only
infrequently encounter such frames in the mainstream media. Furthermore,
in the case of the Lava-Jato probe and the impeachment process, it is highly
unlikely that any one journalistic framing would produce any effect at all,
since the entwined scandals were so central to the contemporary public
debate of Brazil, and audience members would be likely to have formed
opinions before encountering specific texts. So, due to the stereotypical
pattern of reporting, which is arguably founded upon the information-
gathering procedure of the journalists in the political beat of Brasília, I
contend that it is not fruitful to use framing theory independently to
examine or explain reception effects of the Lava-Jato coverage. Still, the
emergence of a hegemonic, unidirectional frame, common to several
influent media actors, is interesting in itself. The following two chapters
will explore two of the background conditions for the emergence of this
frame, namely the leaking of information from investigations and the
continuous reproduction of corruption denunciations through the steady
grinding of the various accountability institutions of Brazil. These
explorations can, in part, explain why the frames tend to converge and
uniformly mirror the prosecutors’ indictments.

References
Albuquerque, A. de (2010) “A modernização autoritária do jornalismo Brasileiro” Alceu, vol. 10(20),
100–115.
Albuquerque, A. de (2012) “On Models and Margins.” In D. Hallin and P. Mancini (eds.),
Comparing Media Systems beyond the Western World (72–95). Cambridge: Cambridge University
Press.
Anders, G. and Nuijten, M. (eds.) (2009) Corruption and the Secret of Law: A Legal Anthropological
Perspective. Aldershot: Ashgate.
Azevedo, F.A. (2006) “Mídia e democracia no Brasil: relações entre o sistema de mídia e o sistema
politico” Opinião Pública, vol. 12(1), 88–113.
Becker, B. and Waltz, I. (2017) “Mapping Journalistic Startups in Brazil: An Exploratory Study.” In
L. Robinson, J. Schulz, and A. Williams (eds.), Brazil: Media from the Country of the Future
(113–135). Bingley: Emerald Publishing.
Biroli, F. and Mantovani, D. (2014) “A parte que me cabe nesse julgamento: A Folha de S. Paulo na
cobertura ao processo do Mensalão” Opinião Pública, vol. 20(2), 204–218.
Biroli, F., Miguel, L.F., and Mota, F. (2011) “Mídia, eleições e pesquisa de opinião no Brasil (1989–
2010): um mapeamento da presença das pesquisas na cobertura eleitoral” Revista Compolítica,
vol. 1(1), 67–89.
Biroli, F. and Miguel, L.F. (2012) “Orgulho e preconceito: a ‘objetividade’ como media-dora entre o
jornalismo e seu público” Opinião Pública, vol. 18(1), 22–43.
Biroli, F. and Miguel, L.F. (2013) “Meios de comunicação, voto e conflito político no Brasil” Revista
Brasileira das Ciências Sociais, vol. 28(81), 77–95.
Breit, E. (2011) On the Discursive Construction of Corruption. A Critical Analysis of Media Texts.
Publication #227. Helsinki: Svenska Handelshögskolan and Hanken School of Economics.
Cabral, E.D.T. (2017) “Mídia Concentrada no Brasil: Até Quando?” Revista LatinoAmericana de
Ciencias de la Comunicación, vol. 13(24), 49–59.
Carro, R. (2016) “Brazilian Newspapers: The Risk of Becoming Irrelevant,” Research paper, Reuters
Institute for the study of Journalism. Retrieved from
https://reutersinstitute.politics.ox.ac.uk/people/rodrigo-berndt-carro-independent-journalist.
Carvalho, E.M. and Albuquerque, A. de (2017) “A Blogosfera Progressista e a releitura do modelo de
jornalismo independente no Brasil.” In C.J. Napolitano, M.M. Vicente, and M.C. Soares (eds.),
Comunicação e cidadania política (305–323). São Paulo: Cultura Acadêmica.
Chaia, V. and Teixeira, M. (2001) “Democracia e Escândalos Políticos” São Paulo Perspectivas, vol.
15(4), 62–75.
Chong, D. and Druckman, J. (2013) “Counterframing Effects” The Journal of Politics, vol. 75(1), 1–
16.
Damgaard, M. (2015) “Multiple Margins and Mediatized Transgression” Ephemera – Theory and
Politics in Organization, vol. 15(2), 411–434.
Entman, R. (1993) “Framing: Toward Clarification of a Fractured Paradigm” Journal of
Communication, vol. 43, 51–58.
Fiske, J. (1994) Media Matters: Everyday Culture and Political Change. Minneapolis: University of
Minnesota Press.
Folha de S. Paulo (2016) “Folha é o 1º jornal brasileiro a ter circulação digital maior do que a
impressa” Folha de S. Paulo, published on September 25, 2016. Retrieved from
www1.folha.uol.com.br/mercado/2016/09/1816633-folha-e-o-1-jornal-do-pais-a-ter-circulacao-
digital-maior-do-que-a-impressa.shtml.
Gupta, A. (1995) “Blurred Boundaries: The Discourse of Corruption” American Ethnologist, vol.
22(2), 375–402.
Habermas, J. (1989 [1962]) The Structural Transformation of the Public Sphere. Cambridge, MA:
MIT Press.
Hallin, D. and Mancini, P. (2004) Comparing Media Systems. Three Models of Media and Politics.
Cambridge: Cambridge University Press.
Hallin, D. and Mancini, P. (eds.) (2012) Comparing Media Systems beyond the Western World.
Cambridge: Cambridge University Press.
Hallin, D. and Papathanassopoulos, S. (2002) “Political Clientelism and the Media: Southern Europe
and Latin America in Comparative Perspective” Media, Culture & Society, vol. 24, 175–195.
Harder, R., Sevenans, J., and Van Aelst, P. (2017) “Intermedia Agenda Setting in the Social Media
Age: How Traditional Players Dominate the News Agenda in Election Times” The International
Journal of Press/Politics, vol. 22(3), 275–293.
Instituto Verificador de Circulação (2016) Circulação de jornais diarias, 2015, 2016. Retrieved from
www.ivcbrasil.org.br – reproduced in Poder360, retrieved from www.poder360.com.br/wp-
content/uploads/2017/04/circulacao-IVC-dez2015-dez2016.png.
Iyengar, S. (1990) “Framing Responsibility for Political Issues: The Case of Poverty” Political
Behavior, vol. 12(1), 19–40.
Kahneman, D. and Tversky, A. (1979) “Prospect Theory: An Analysis of Decision under Risk”
Econometrica, 47(2), 263–292.
Kajsiu, B. (2014) A Discourse Analysis of Corruption: Instituting Neoliberalism against Corruption
in Albania, 1998–2005. New York: Routledge.
Kitzberger, P. (2014) “Demands for Media Democratisation and the Latin American ‘New Left’.”
Hamburg: GIGA German Institute of Global and Area Studies.
Koechlin, L. (2013) Corruption as an Empty Signifier. Leiden: Brill.
Kucinski, B. (1991) Jornalistas e Revolucionários nos tempos da imprensa alternativa. São Paulo:
Editora Pagina Aberta.
Lattman-Weltman, F. and Abreu, A.A. de (2001) “Controles Midiáticos: Investigação e Denuncismo
na Construção da Visibilidade Pública Democrática.” Paper presented at ANPOCS 25th
conference at Caxambu, Minas Gerais, October 17, 2001.
Lattman-Weltman, F. and Chagas, V. (2016) “Mercado Futuro: A Economia Política da
(Re)Partidarização da Imprensa no Brasil” Dados, Revista de Ciências Sociais, vol. 59(2), 323–
356.
Lennerfors, T.T. (2008) The Vicissitudes of Corruption. Degeneration – Transgression – Jouissance.
PhD thesis, Royal Institute of Technology, Stockholm.
Lima, V.A. da (2004) “Sete teses sobre mídia e política no Brasil” Revista USP, São Paulo, no. 61,
48–57.
Lima, V.A. da (2009) “Revisitando as sete teses sobre mídia e política no Brasil” Comunicação &
Sociedade, vol. 30(51), 13–37.
Lima, V.A. de and Lopes, C.A. (2007) “Rádios Comunitárias: Coronelismo Eletrônico de novo tipo.”
Observatorio de Imprensa. Retrieved from
www.observatoriodaimprensa.com.br/download/Coronelismo_eletronico_de_novo_tipo.pdf.
Matos, C. (2012a) Media and Politics in Latin America. Globalization, Democracy and Identity.
London: I.B. Tauris.
Matos, C. (2012b) “Media Democratization in Brazil: Achievements and Future Challenges” Critical
Sociology, vol. 38(6), 863–876.
Meneguello, R. (2010) “Aspectos do desempenho democrático: Estudo sobre a adesão à democracia
e avaliação do regime.” In J. Moisés (ed.), Democracia e confiança: Por que os cidadãos
desconfiam das instituições públicas? (201–236). São Paulo: Edusp.
Michener, G. and Pereira, C. (2016) “A Great Leap Forward for Democracy and the Rule of Law?
Brazil’s Mensalão Trial” Journal of Latin American Studies, vol. 48, 477–507.
Miguel, L.F. (1999) “Mídia e Eleições: A Campanha de 1998 na Rede Globo” Dados, vol. 42(2),
253–276.
Motter, P. (1994) A batalha invisível da constituinte: Interesses privados versus caráter público da
radiodifusão no Brasil. Master’s thesis. Brasília: Universidade de Brasília.
Nascimento, S. (2013) “Reportagens com denúncias na imprensa brasileira: análise de duas décadas
da predileção por mostrar problemas” Verso e Reverso, vol. 65, 68–76.
Newman, N., Levy, D., and Nielsen, R.K. (2015) Reuters Institute Digital News Report 2015.
Oxford: Reuters Institute for the Study of Journalism, University of Oxford.
Newman, N., Fletcher, R., Kalogeropoulos, A., Levy, D., and Nielsen, R.K. (2017) Reuters Institute
Digital News Report 2017. Oxford: Reuters Institute for the Study of Journalism, University of
Oxford.
Nobre, M. (2013) Choque da Democracia. Razões da Revolta. São Paulo: Companhia das Letras.
Nunes Leal, V. (1997) Coronelismo, Enxada e voto. Rio de Janeiro: Nova Fronteira.
Nye, J. (1967) “Corruption and Political Development: A Cost-Benefit Analysis” American Political
Science Review, vol. 61(2), 417–427.
Pan, Z. and Kosicki, G. (1993) “Framing as Strategic Action in Public Deliberation.” In S. Reese, O.
Gandy, and A. Grant (eds.), Framing Public Life: Perspectives on Media and Our Understanding
of the Social World (35–65). London: Routledge.
Paterson, C. and Domingo, D. (eds.) (2008) Making Online News: The Ethnography of New Media
Production. New York: Peter Lang.
Porto, M. (2010) “The Changing Landscape of Brazil’s News Media.” In D. Levy and R.K. Nielsen
(eds.), The Changing Business of Journalism and its Implications for Democracy (107–124).
Oxford: Reuters Institute for the Study of Journalism, University of Oxford.
Porto, M. (2012) Media Power and Democratization in Brazil. TV Globo and the Dilemmas of
Political Accountability. London: Routledge.
Power, T. and Taylor, M. (eds.) (2011) Corruption and Democracy in Brazil. South Bend, IN:
University of Notre Dame Press.
Reese, S. (2007) “The Framing Project: A Bridging Model for Media Research Revisited” Journal of
Communication, vol. 57, 148–157.
Ribeiro, E.F. (2017) “A redução da audiência da televisão aberta brasileira” Temática, vol. 13(6),
109–126.
Rocha Filho, A. (2016) “Jornalismo popular na era digital: a experiência do Extra,” Paper presented
at II Simpósio Internacional Jornalismo em Ambientes Multiplataforma, November 21, 2016,
ESPM, São Paulo.
Shehata, A. and Strömbäck, J. (2013) “Not (Yet) a New Era of Minimal Effects: A Study of Agenda
Setting at the Aggregate and Individual Levels” The International Journal of Press/Politics, vol.
18(2), 234–255.
Sniderman, P. and Theriault, S. (2004) “The Structure of Political Argument and the Logic of Issue
Framing.” In W. Saris and P. Sniderman (eds.), Studies in Public Opinion: Attitudes, Nonattitudes,
Measurement Error, and Change (133–165). Princeton, NJ: Princeton University Press.
Sodré, N.W. (1977) História da imprensa brasileira. São Paulo: Graal.
Waisbord, S. (2000) Watchdog Journalism in South America. News, Accountability, and Democracy.
New York: Columbia University Press.
Welbers, K., Atteveldt, W. van, Kleinnijenhuis, J., and Ruigrok, N. (2018) “A Gatekeeper among
Gatekeepers” Journalism Studies, vol. 19(3), 315–333.
3 Leaks and Their Effect in Media and on
Politics

 
Disclosure of corruption among political leaders is sure to catch the
attention of any independent media system. Corruption scandals are not
only excellent media material, but also the focus of intense interest in
international governance. The fight against corruption has thrust the term
transparency to the top of the priority lists of many international
organizations and corporations, and global charters, frameworks and
compliance policies are proliferating. This global focus on corruption as the
evil twin of good governance does not always result in successful anti-
corruption initiatives and efficient court cases that bring culprits to justice,
however. Even if denounced for corruption, individuals holding positions of
trust and public authority, as appointed or elected representative of citizens,
can still in many cases count on impunity, due to political protection or
judicial sluggishness. In other cases, scandalizing information is disclosed
to great effect, when public officeholders are forced to face the pressure of
courts, their own constituencies and the electorate in general.
Whether resulting in sanctions or not, the disclosure of information
about corruption can be a political game-changer, even if that information
turns out to be inaccurate or false. Scandals may thus wreak havoc in
consolidated as well as emergent democracies, and the effects of corruption
scandals, of disclosing transgressions and alleging moral misconduct, go
beyond the immediate horizons of individual punishment or impunity. In
this chapter, the way that such disclosure of scathing information plays out
is explored. I will argue that the disclosure of corruption constitutes an
especially powerful and dramatic style of interaction between judiciary,
political, and media systems. Leaks of information, in particular, can be
harnessed by political and institutional actors to great effect. In a fast-paced
media environment, leaks fundamentally erode journalistic and editorial
control and, paradoxically, provide powerful (and potentially anti-
democratic) access to the media for those institutional actors with inside
knowledge. These actors are shrouded as sources with vested interests,
while the media outlets frame disclosure as ostensibly increasing
transparency.
The chapter will proceed by first defining leaks and discussing the
control of information flows in Brazilian public spheres. Subsequently, the
chapter provides a description of the particular leaks that flowed from the
investigations of Brazil’s Lava-Jato case into the sphere of national media.
The effects of these leaks in the political and judicial system are then
tracked through a simplified process-tracing methodology. The outcomes
traced with this methodology are relevant to research into various aspects of
political systems, including the coalition formation (Balán 2011, 2014) and
congressional veto-players in presidential systems (Pérez-Liñán and
Rodríguez-Raga 2009), judicial intervention in the political domain (Hirschl
2008, Tate and Vallinder 1995), presidential breakdowns (Llanos and
Marsteintredet 2010) as well as Latin American media-government
relations (Kitzberger 2014).
The goal is not to measure how much scandalizing leaks influence
political support for specific actors (Maier 2011) or overall regime
legitimacy (Seligson 2002), but to identify the specific contributions of the
leaks to the regime change in Brazil in 2016, including the impeachment of
Dilma Rousseff and the various legal processes against her successor,
Michel Temer. Thereby, this chapter expands the insights of Arantes (2002,
2011) and others (Chaia and Teixeira 2000, Porto 2011) concerning the
media-savvy Brazilian prosecutors’ dissemination strategies by exploring
what consequences the disclosure of corruption generated besides
investigations and trials.
On the background of this material, several theoretical questions emerge:
How is it that disclosed information is turned into knowledge in the public
sphere? How is truth determined and influenced by the mode of disclosure?
What is the relationship between leaks on political corruption and the
proliferation of transparency discourses in recent decades? The discussion
of these questions, in the concluding sections, help explain how leaks
provoke institutional rupture and possibly destabilize democratic
institutions.
Existing Research on International Leaks and Brazilian Vazamentos

In this section, I will introduce some of the existing academic views about
the potential democratic benefits of leaks. This means looking at both the
current international research into leaks (in the sense of large amounts of
data) and the research on politics of sources and information control. This
section provides a background for understanding why and how leaks of
information in corruption “skip the gates,” that is, quickly reach the
headlines of national media. The journalistic mechanisms of boundary
control, which requires not only selecting the information that enters public
spheres, but also fact-checking and source-checking, are challenged by the
rush of the scoop when secret and scathing information is disclosed. Leaks
thereby relocate some of the power of information control normally held by
media, and this means that watchdogs may turn into lapdogs (Porto 2011).
The current international literature about leaks is increasingly tuned into
the disclosure of very large amounts of documents; a trend probably pushed
by the emergence of data storage and sharing technologies. During the past
decade, leaks of vast amounts of scathing information have reached
international public spheres with increasing frequency, visibility, and
impact. Targeting national elites, diplomacy, and war efforts, leaks such as
the Panama Papers, the Paradise Papers, and especially the prominent
WikiLeaks have been lauded by scholars for fostering accountability
(Russell and Waisbord 2017), for providing a corrective for imperfect
institutions, a “critical safety valve” (Benkler 2014: 303) in governmental
branches, and for championing free speech (Wahl-Jørgensen 2014).
The NSA documents leaked by Edward Snowden have recently been
discussed by Heikkilä and Kunelius (2017), as well as Waisbord and
Russell (2017), while the history of the Wikileaks platform is detailed and
discussed by Benkler (2014), Chadwick (2013), Coleman (2014), Fuchs
(2011), Munro (2016), and Wahl-Jørgensen (2014). Apart from the praise of
the democratic potential, many of these authors also acknowledge that such
a complex and hybrid assemblage of new and old media technologies and
logics has very uncertain consequences in addition to the obvious potential
for transparency and resistance. I will highlight the perspective of Russell
and Waisbord, Benkler, and Wahl-Jørgensen, who view leaks of large
amounts of data as an important new phenomenon in democracies. This
kind of leak represents a novel way of doing investigative journalism, as
well as a new integration between politics and media – an integration that
potentially can remedy some of the flaws of accountability in modern
politics (Benkler 2014).
The novelty, first off, lies in the new organizational and institutional
structures emerging in order to accommodate the scale of leaks: In April
2016, the consortium behind the Panama Papers revamped the WikiLeaks
model of the networked workload between freelance journalists, national
quality papers and the technological aspects of servers, mass storage and
encryption (Cabra and Kissane 2016, The International Consortium of
Investigative Journalists 2016). The consortium of journalists disclosed the
financial actions of scores of banks and hundreds of powerful and rich
individuals (Drezner 2016), while the related disclosure of the Paradise
Papers, publicized in November 2017, turned public the tax evasion
practices of the offshore law firm Appleby. A similar combination of
journalism, data leaks, and cross-media collaborations saw the light in the
Football Leaks driven by the European Collaborative Investigations in late
2016. Large data dumps, in this way, have recently been assimilated by
journalistic communities and editorial practices, with an organizational
setup emerging along with the leaks. High-profiled Western media actors
(such as the Guardian, Der Spiegel, and El País) have converged upon a
kind of consensus and modified the constitutive framework of journalism
and the production of news to manage not only the data leaks and the
economic distribution of working hours, but also the economic gains of
scoops strategically placed in breaking-news cycles.
Second, large-scale leaks challenge theoretical distinctions between
journalistic material and, for a lack of a better word, data. Leaks exist at the
border between “raw data,” if there ever was such a thing, and a more
classic, journalistically curated disclosure of previously hidden information.
The border is blurring partly because of the divergent ideals of instantly
leaking big data dumps contra journalistically curating the material as news.
WikiLeaks (e.g., in the leaks of the 251,287 diplomatic cable message)
originally preached radical openness built upon an open-source mindset,
while other, later leak cases, such as the Panama and Paradise Papers, were
disseminated in accordance with a journalistic logic. This was already a
tension in the early days of WikiLeaks (Chadwick 2013), when WikiLeaks
produced a video (the Collateral Murder video, in two versions), which is
more of a documentary based on leaked footage than a leak in itself.
Despite the buzz of big and raw data, researchers need to consider where to
draw the line: Does the slow trickle of leads and clues followed by
investigative journalists in the Watergate scandal count as a leak – or was
that merely a source disclosing information and telling the Washington Post
journalists where to look? No yardstick exists to measure leaks.
In the Lava-Jato case, anyway, both small-scale leaks and large data
dumps have been disclosed, and both kinds have had political
consequences. Whether or not off-the-record tips from official inside
sources are theoretically distinct from data dumps of gigabytes worth of
documents, and no matter the degree of journalistic pre-production, the
analysis needs to account for both types, because the principal effect of a
leak in public spheres cannot be predicted by either megabyte size or
editing processes.
The academic literature of Brazil dealing with journalist-source relations
(Waisbord 2000: 93ff.) and the public prosecutors’ media relations (Arantes
2002, Porto 2011) have mostly mentioned small-scale leaks (vazamentos):
Brazilian journalists have not seldom received access to elements of
investigations from official sources, in off-the-record interviews (ibid.: 118)
or even in concerted actions where whistle-blowing testimonies are
recorded and transmitted live by a broadcaster (Arantes 2002: 166, 211).
Investigative journalism of the 1990s (in Brazil and internationally) was
rarely executed on the basis of massive databases of secret documents,
however.
Today, with the introduction of cheap storage devices and processing
power to facilitate searches across large swathes of documents, journalists
have been given new tools to produce investigative journalism, or, rather,
journalism-about-investigations (which is perhaps superseding the former,
e.g., Porto 2011: 118). Journalists may thus document and demonstrate
facts of corruption cases to a degree previously unheard of. A gradual
opening of the judicial system has meant that a number of journalists are
allowed access to a subset of non-classified case files in the same digital
content management systems used by police agents, lawyers and judges, in
order to cover cases. The classification of case material by courts is thus an
extremely important gate, and the decisions about public access vary on a
case-to-case-basis. The task-force of the Lava-Jato case and the regional
court of Curitiba maintain a policy of publicizing any and all material in
cases as long as that publication cannot impair investigations (Albuquerque
2017). This is not common practice in Brazil, however, and a number of
cases in the Supreme Court are not known to anyone but the judges in that
court.
In addition to the transparency initiatives of various Brazilian courts,
various processes of the Lava-Jato probe have fed information directly to
the mainstream media. At the end of this chapter, I will argue (in a more
detailed manner) that the truth claims built into the discourse on leaks and
exposés are strengthened with all this information at the disposal of the
journalists, and I will argue for this. Here, turning to the relationship
between politics and leaks in media, I will argue that the new tools and
dynamics also introduce a new dynamic, without resorting to technological
determinism. The new technologies of dissemination (and obtainment) of
information connect accountability institutions and media in a new way.
Functionally speaking, the integration model of data-dumps and
investigative journalism does not necessarily entail anything new in
political arenas. Leaks of the type of WikiLeaks, Panama Papers, or
Football Leaks are fundamentally disclosure with the interest of weakening
political or economic actors or entire elites, calling them out in
democratically dubious financial or political transactions or decisions.
Quoting Thompson, ‘[s]candals are struggles over symbolic power in which
reputation and trust are at stake’ for individual actors and organizations
(Thompson 2000: 245). Such disclosure was the content of Watergate,
innumerable scandals after it, and many historical examples before that time
(e.g., ibid.: 50). That is, contemporary data leaks are, functionally, one
possible trigger event of scandals relating to transgression of the frontstage
roles expected in public offices or similar (usually media-covered) elite
positions.
While the effects of leaks on political systems are still embedded in the
established ways of handling transgression and holding officials
accountable, the media format of leaks is different: The scope is global, and
the volume and extension of content is of an entirely different scale.
Furthermore, as I will discuss in the following, this content is being pushed
into public spheres in a qualitatively new way. The outcome – for example,
media pressure and popular protests against corrupt leaders – might still be
the same as for old-fashioned scandals, but the raw material and the
journalistic labor of the scandal looks much different now, because of the
technological affordances (Hutchby 2001). The intertwined technologies of
the internet, social media, big data, search capabilities and data
visualization, while not determining the emergence (and certainly not the
effects) of leaks, do in fact provide some of the necessary conditions to
extract and distribute the information of a scandal. Moreover, in the
following sections, dealing with the dynamic of information flows, sources,
and media organizations’ control of information, I argue that leaks
contribute to bypassing the gatekeeping role or boundary control exerted by
legacy media newsrooms (Shoemaker and Vos 2009).

Information Control
I will now unpack the claim that professional journalism no longer
exclusively patrols the borders of public spheres, and that every such sphere
in its very publicness is, in fact, seriously challenged in this “Information
Age” (or whatever epochal catch-phrase the reader might prefer). The
decline of legacy media’s control of the news agenda was foreshadowed in
Bill Clinton-related scandals of the late 1990s (Delli Carpini and Williams
2001), but this trend is hypercharged in contemporary leaks: The
affordances of a wide-ranging network of technologies, actors, and quasi-
ideologies – the whole gamut of server technology, hacker communities,
social media, and popular post-Watergate mythology – results in the erosion
of the journalists’ and editors’ jurisdiction over public spheres. Leaks are
one particular element in this erosion of the boundary control exercised by
media professionals.
In the following, I will briefly define information control and how this
control is usually exercised and negotiated in media spheres, followed by a
discussion in the next section on the typical situations in scandals where
information is disclosed to journalists (sometimes in the form of leaks).
Based on this analysis of information disclosure settings, I can establish
theoretically why leaks seem to be able to get past the gates and quickly
reach the headlines of all national media outlets. I work with the label
“information control” since I deal with information flows in a variety of
settings, but the phenomenon has been researched before in media studies,
under a different name. As early as 1950, under the heading of
“gatekeeping,” David Manning White theorized the relations between, on
one hand, the media organizations’ structures, and, on the other, the events
and information that would survive selection and editing processes and
ultimately make it to the news (White 1950). The primary research object of
this book is not newsroom processes, however, and the following sections
only deal with leaked information at the moment it reaches a media
organization. Therefore, I omit the tradition and the specifics of gatekeeping
theory and instead present a brief description of the obstacles in the
publication of news first, and second a description of the forces that push
leaked information into public spheres.
Before disclosed information can even be made to fit criteria of
newsworthiness and produced as journalistic content, information has to
reach a news gatherer (which is usually also a journalist, be it a freelancer, a
newspaper or broadcaster employee, or a journalist working with news
agency). In a way, the individuals with first-hand or authoritative
knowledge constitute a kind of gate to the public sphere of the news, as
these sources can decide whether to leak or not. Because the status of the
source is important to journalists (Tuchman 1978: 69), some sources have
an easier time pushing information. Having information from an authority
or otherwise legitimated source (ibid.: 92ff.) is important to journalists,
because such sources are “in-the-know” and are reliable producer of truth-
claims, even when that truth is contested by other social actors. Therefore,
press releases from government authorities were the staples of “objective”
news gathering for decades in the United States (Shoemaker and Vos 2009:
55, Tuchman 1978: 93).
Even in the large-scale, vertically integrated, contemporary media
corporations, White’s insights about the forces of selection and rejection
inside news organizations remain relevant; many layers of editorial
selection influence the production outcome, i.e., the final printed or
broadcast news product. A number of external conditions influence the
selection – for example, the properties of particular events – of the news
agency hierarchy, and the construction and pre-packaging of the event as
newsworthy by other media professionals trying to push their texts into
publication. News events themselves, the sources, the news-gathering
journalists, the media organization, ideology and ownership structures all
play different roles in information flows of the public sphere. Corruption
scandals are a special case of information control, however, and not exactly
comparable to the general processes of news selection and production. This
is due to the fact that political actors implicated in a scandal will have every
incentive to reduce media attention, in contrast to everyday politics where
politicians want to attract the spotlight. In scandals, the parallelism of media
organizations’ logic of market optimization and the politician’s interest in
audience optimization (Landerer 2013) is inverted – a denounced politician
will want as little attention as possible, while the media outlet still wants to
maximize the audience.
The control of scandalous information is not just a question of sources or
news objects wanting to minimize attention. The media organizations also
have hierarchies determining the resources allocated to covering a scandal
(or, at times, to eliminate scandalous news), meaning that journalists do not
single-handedly determine what to cover, and neither does any single editor.
Rather, in contemporary newsrooms, a complex set of interactions influence
the process of selecting news for the next issue or the next newscast. Some
of the relevant interactions, constraints, and causes include: The individual
journalist’s personal beliefs, identity, and professional profile position in
relation to specific subsets of the news (regular columns, specific teams and
tasks to which she is assigned, etc.), as well as her location in the newsroom
hierarchy – all of these factors play a role in whether she can produce a
news item that cuts through to publication. From time to time, a journalist
might even choose not to make a news item of some bit of information in
order to preserve professional relations to a source. In general, “since media
trade in information, content decisions can also be driven by the need to
appease suppliers, i.e., sources” (Shoemaker and Vos 2009: 77). In sum, all
of this can cause a news event or bit of information to be rejected as news
before it even reaches the desk of the copy editor.
Next, editors might reject material on many grounds, stemming from
ideological editorial principles, such as the news media’s standards of
documentation or of objectivity, or stop a story on more pragmatic grounds
(space constraints in the print version, or, in extreme cases, the risk of libel
lawsuit). Editors might also reject stories because they operate under certain
constraints set down by their superiors. In delicate political moments,
editors will also consider the current editorial stance and line of opinion
presented by the newspaper, in order to ensure alignment (and, as we shall
see, even manipulate or omit data in order to ensure that alignment). The
desire to emphasize a specific editorial line might lead to the inclusion or
rejection of material handed over by journalists. The editors are also
influenced by the fact that their organization, due to target group
considerations, may incline toward a certain ideology and omit negative
stories related to persons spouting such ideology. In Brazil, the personal
relations of media conglomerate owners to political allies (which could
simply be advertisers, or it could be political connections), have also led
editors to remove material (see Miguel 2003: 292).
Despite the interest of implicated actors for minimizing attention, and
political, editorial, and organizational reasons not to run a scandal story,
leaks may still rapidly enter the public sphere as information in news items.
There are a number of reasons for this: Leaks normally deal with powerful
and influential individuals, and a leak about such an individual’s previously
secret actions is per definition newsworthy, because public office-holders
are not, as a rule with some exceptions, supposed to keep secrets from their
constituencies. Because of this transgression of the ideal frontstage role
(Goffmann 1956) of public office-holders, leaks touch most news values:
relevance, drama, scandal, negativity, conflict, human interest, proximity,
and suspense (discussed originally in Galtung and Ruge 1965). Since leaks
of secret information beg the question of “who else knew this,” a leak
generates speculation and draws in other actors (discussed in the final
section of this chapter). No matter the veracity, a leak can result in
repercussions for political actors and even cause a reconfiguration of the
balance of power between decision-makers, which draws media attention.
In sum, many possible agents and factors contribute to obstructing or
facilitating the flow of information into news at various levels: Individual
levels (sources, news gatherers, and news processors such as editors), at the
professional and organizational level (the routines of journalists and editors
in a particular media organization set-up), the social institutional level
(which at the general level determines what kind of news is no-go, what can
be published and what should be avoided) and a more abstract level of the
social system, culture and ideology (Shoemaker and Vos 2009: 31).
However, the very secrecy of leaked information signals transgression, and
disclosed transgressions are important material for news items because of
the potential political repercussions, provoking speculation and coverage of
reactions to the leak.

Dissemination and Disclosure

I will now explore the levels and the ways in which information is
controlled, disclosed, and passed to media in scandals. Based on the media
material of the Lava-Jato case, I will describe the typical situations where
corruption is disclosed, or where new information about corruption
emerges, and the forces that play into these situations and ensure that this
information is transformed into news.
The settings and situations exhibit varying levels of information control,
that is, the sources act as gates of information in relation to the journalists,
to various degrees and for different purposes. However, that does not
exclusively mean that sources act as obstacles by withholding information;
on the contrary, many situations of corruption disclosure are specifically
designed to maximize the publicity of denunciations and investigations. The
Lava-Jato task force, for instance, normally host press conferences
following larger police operations.
In a scandal, journalists may encounter information and elicit
information from sources in several ways. This can be broken down into
three groups: (1) the moments of “live,” active journalism at scheduled
events or in sudden, on-the-fly situations; (2) pre-packaged material sent to
journalists; and (3) the special kind of source material leaking from
investigations. This last group may give rise to investigative journalism
based partly on insider sources, triangulated with publicly available
information and information from more distant sources who may
corroborate facts.
Live material comprises footage and possibly interviews from press
conferences, court trials, and even investigation phases underway. The
already-packaged material includes press releases from prosecuting,
judging, and investigating authorities and their counterparts, the denounced
or arrested individuals and their lawyers. Leaks usually either emerge from
sources close to corruption, or from within the judiciary system. In the
former cases, the sources are mostly clear to the eventual news consumer,
while in the latter cases, the sources may be completely opaque, or the
source deciding to transmit the information may be hidden. In the last case,
it is usually necessary for the journalist to obscure the identity of the source
for protection.
The pre-packaged information is, naturally, the most tightly controlled,
and the sources act with the deliberate intention of controlling information
and projecting certain framings of the material publicized: Where material
is pre-packaged, the aim of the sources is always to get their version across
to the media, and, possibly, to omit or de-emphasize information that could
turn out to be harmful. When courts publicize material, it is often to comply
with a specific policy set down to regulate and also restrict public access,
meaning that such material frequently omits a lot of information deemed to
be of personal nature, or potentially harmful to parallel investigations, if
made public. Although pre-packaged material, in its nature, is designed to
inform the press, journalists are aware of that, and will resist the framing
projected by the sources, especially if the information is ultimately a denial.
The framings of pre-packaged material are more likely to survive the news
processing if the information comes from sources which are considered
authoritative (Tuchman 1978: 69), or if the frame is consonant with the
ideology and editorial line of the news outlet.
The live situations are, to various degrees, also controlled, just like the
pre-packaged textual material of press releases. At one end of the spectrum,
trials in court rooms are usually very ritualized and highly predictable, with
the possible exception of situations where defendants go beyond the script.
Information about the case will usually be accessible to interested parties
and journalists either a long time ahead or not at all, in the cases of
confidential cases (a small, but interesting subset of cases in the Supreme
Court). Thus, the “liveness” of the trial is restricted to the spectacle of the
trial (and the visual imagery of the court) because information concerning
the alleged facts and accusations of the case will already have passed
through the gates, and as a rule, the defense lawyers will already have
decided upon a strategy of defense and made their arguments available to
the press.
Another “live” setting where journalists may encounter information is
the press conferences hosted by police agents and prosecutors who present
phases of investigations. Such press conferences have been very common in
the Lava-Jato case. They usually follow a planned script, designed to
convey the facts of the police operation and explain the line of inquiry,
providing easily digestible material for journalists. Likewise, the press
conference situations where denounced individuals present their counter-
arguments also feature tight control of information. In both situations, the
speakers will usually stick to their script unless forced to respond to
unexpected questions, which once in a while can turn up a nugget of
information or trap the person rhetorically. The two situations differ
somewhat in another sense of information control, however: When
investigations are triggered and described for the first time by police
authorities and prosecutors, it usually generates enough interest that editors
will want to commit resources and send journalists to cover the press
conference. This increases the likelihood of publication, because the
resource commitment means that editors are hoping to get something in
return for that investment. When a denounced person calls a press
conference, it is assumed that he or she will be denying the allegations,
which is less newsworthy, since it is neither surprising, unusual, nor
(presumably) contributing any new information to the case. The only news
criteria activated by this type of live press conferences are the sense of
conflict, which is heightened by denials, and the availability of visual
footage (possibly of individuals with various emotions emanating).
Admissions of guilt would, of course, be surprising and more probable to
pass gates; but straight admissions are also exceedingly rare (although
possibly quite important for political careers; see Allern and Pollack 2012).
The live situations with the least amount of control of information are
the moments where teams of reporters follow investigations, getting
glimpses of arrests, and perhaps comments from investigated or
investigators which have not been prepared due to the acute circumstances.
Later in this chapter, a notable case of such an investigation phase is
described, wherein the police force leaked the upcoming arrests to the press,
in order for the TV teams to arrive in time to get material for the news. The
images of such investigations and police operations may in themselves be
hugely significant because they engender speculation and commentary from
media pundits. The drama of live footage, where police agents with search
warrants enter private homes and drag out criminals, is visually much more
appealing than the stale press releases of courts and prosecutors, and thus
more likely to pass gates in broadcast media.
Finally, we get to the leaks. It is necessary to distinguish between two
senses of leaks: One type of leaks comes in the form of documentation –
evidence, phone taps, footage, etc. – while another type is testimonies and
witness accounts. What is characteristic of both types is their apparently
“raw” nature, which makes them appear as inherently truthful
representations of reality. Leaks of the document-evidence kind can be
interpreted in many ways, allowing for speculation and extrapolation. Leaks
of testimonies and plea bargains, on the other hand, seem to be of a more
obviously subjective nature. However, since the information disclosed in
plea bargains is revealed under severe penal restrictions, this type of
information is likely assumed to contain more truthful representations of
reality than denunciations and allegations presented without the restraints
imposed by criminal charges and pending sentences.
When discussing leaks, I am not referring to case material publicized
officially by the court in Curitiba or by the STF once dissemination could
not harm the investigation. Rather, the leaks of material and information
discussed here frequently came before the instauration of inquiries, before
plea bargains had been finalized, or before any indictments. Several
important leaks to the weekly press were based on pre-testimonies of the
Lava-Jato probe, in other words, preliminary stages of plea bargains, where
arrested individuals indicate the scope of their knowledge concerning
corruption schemes. Such leaks could only originate with either the
testifying person or the police agents obtaining the statements. In fact, the
office of the Prosecutor-General of the Republic has frequently complained
about the federal police agents being the source of leaks. The cases of plea
bargains where police agents have been shut out of negotiations have also,
however, resulted in leaks. A number of actors handle the information
contained in testimonies and emerging from police investigations, including
public prosecutors or police agents involved in the investigations, and in
some cases (depending on the levels of confidentiality in investigations) the
lawyers defending the investigated politicians and businessmen. These
actors have different incentives to leak.
Lawyers would usually have little incentive to leak the information
unless the media actors offered substantial rewards, which is known to
happen in the UK, but is taboo in Brazilian journalism. A lawyer’s
professional credibility could be jeopardized if he were revealed to leak
confidential information. However, the negotiation situation of plea
bargains probably makes it more likely that some suspects would want to
leak (or have their lawyers leak) information. This is due to the fact that
when leaked information gets enough public attention, it might force the
hand of the prosecutor or police agent negotiating the terms of a plea
bargain: Police investigators can hardly threaten to deny a plea bargain and
discard information when that information is already in the media limelight.
 The judge and prosecutors working on the Lava-Jato case have made it
abundantly clear that public attention to corruption (through media
attention) is central to the efforts of combatting corruption (Albuquerque
2017, Dallagnol and Martelo 2016, Moro 2004). This proactive public
relations strategy is close to the heart of the Lava-Jato task-force’s strategy
for maintaining pressure on the political system and legitimizing lengthy
preventive imprisonment of public officials and businessmen. In interviews
and opinion pieces, the prosecutors have repeatedly argued that the
preventive arrests would seem unjust in the eye of the public if citizens
were not informed of the political transgression. Leaks thus join legally
publicized material as elements of a strategy for gaining sufficient
legitimacy to enact extraordinary judicial measures. For these reasons, it
makes sense that the judiciary and prosecuting actors have little interest in
finding, investigating or punishing any internal sources of leaks.
Furthermore, since many restrictions apply to the disclosure of pending
legal actions, the instances where these restrictions are bypassed by
individuals within the judiciary system are sure to generate interest among
journalists. Therefore, a leak of information, merely in virtue of its
extraordinary disclosure, generates an important thrust to get information
past the initial gate. In particular, the moment and timing of disclosed
evidence is of the utmost importance in the situations we will consider in
the next section of this chapter.
The bottom line of all this is that in the publication of leaks, the sources
and their motives are hidden from view, in contrast to either pre-packaged
material or the “live” situations described above. In the Lava-Jato case, this
meant that the insiders of the corruption cases (lawyers, prosecutors, and
judges) have had the privilege of being able to decide whether they want to
present information to the public (and how much). Because the mere
connection to this scandal warranted front-page and prime-time attention,
journalists were eager to reproduce scandalous allegations from
investigations and testimonies. The tendency to uncritically reproduce
denunciations thus circumvents gatekeeping and changes the dynamics of
the source-journalist relation.
While in theory, leaks contribute in fostering accountability (Benkler
2014, Russell and Waisbord 2017), the occultation of sources shrouds the
enunciation of the leak in necessary shadows, while the media coverage
paradoxically is affirming the dangers of secrecy in public space. The
transparency created by leaks is shallow and one-dimensional, because the
leaking agent remains incognito (Wahl-Jørgensen 2014). The accountability
of politicians to their constituencies and the general public is a democratic
necessity, but the accountability of the sources of news is rarely under
scrutiny. In other leak cases and leak organizations, such as WikiLeaks and
Panama Papers, the various ways of protecting whistle-blowers and other
sources are even built into the information distribution technology and the
organization itself, which completely removes the sources from the
limelight. The concealment of sources is in reality an inversion of the
transparency supposedly created by leaks: While a sudden disclosure of
information may force political actors to account for themselves, the
lightning-strike that forces visibility upon the targets simultaneously blind
the spectators. The visibility is only directed toward the objects of the leaks,
and rarely shines a light upon the position of the enunciator.
Beyond the shrouding of sources, certain institutional actors of Brazil
also gain a privileged position through leaks:

• Prosecutors and police agents (because these, at different times,

apparently control the timing of at least some of the leaks, and control
to whom they leak the information).
• Judiciary actors (gaining legitimacy relative to the political branch of
government, because leaks cast suspicion upon the politicians while
suppressing the question of political agency behind the leaky
investigations and trials).
• Media conglomerates (because only the large mainstream media have
resources to extract and process all the scoops, and the agenda-setting
power attractive to the leaking agents).

Having established how information relating to corruption investigations

is controlled and disclosed in general, I will now describe the specific
sequence of major leaks that dominated the front-pages and newscasts in
Brazil between 2014 and 2017.

Leaks of the Lava-Jato Case

In this section, I will present the chronology of the leaks in the form of a
narrative that also outlines the immediate political impact of the leaks and
the unfolding scandal. The sections that follow this one will provide the
theoretical framework for exploring the effects of the leaks in depth.
The leaks of the Lava-Jato operation began a few months after the initial
arrests, described in the introductory chapter. These early stages of the
investigation targeted different departments and executive offices of the
state oil company Petrobras, and although speculations abounded, few
named politicians were implicated in the case until the presidential
campaign of 2014 was well underway, in August and September 2014.
The first major leak, on October 23, 2014, happened just three days
before the second round of presidential elections. In this election round, the
incumbent Dilma Rousseff of PT was pitted against Aécio Neves, candidate
for the liberal party PSDB. A leak of the testimony given by Youssef, the
main money launderer of the case, was published by the tabloid magazine
Veja, out of the normal publication schedule. The corruption in Petrobras
and the losses to the state were denounced on the front page of this
lightning-strike special issue, and, quoting from the leaked testimony, the
headline, “They knew it all,” was printed on top of the faces of Rousseff
and her mentor, ex-president Lula, both looking pale and desperate against a
black and foreboding background. The election resulted in a historically
close race, with Rousseff barely securing a win.
Costa (the ex-director of Petrobras) and Youssef both negotiated plea
bargains in order to reduce their sentences. This generated a snowballing
chain of evidence, as these two plea bargains incriminated more CEOs,
bosses from sub-contractors, and clandestine financial operators. The
testimony of another Petrobras manager was leaked to Estado on December
14, 2014, targeting João Vaccari Neto, the national treasurer of the
president’s party PT. Several others implicated in the case also pointed to
the treasurer as the man collecting kickbacks, and the media picked up on
various bits from these testimonies during January and February 2015. On
February 5, the treasurer was forced to testify to the police, and Vaccari
Neto was denounced on March 16, and arrested the following month.
In the first days of March, leaks about the Vaccari’s role in the Petrobras
graft were overshadowed by the rumors and leaks about the list being
compiled by the prosecutor-general’s office, awaiting approval of the
Supreme Court. The prosecutor-general’s list, and its many appendices
detailing evidence, requested the authorization of criminal investigations
into scores of parliamentarians. The information seeping from the
prosecutor-general’s office into Folha and then the other major newspapers
formed the second major leak and forced both the president of the Senate
and the president of the Câmara dos Deputados (the Lower House of
Congress) to take political action. The Senate leader, Renan Calheiros of
PMDB, blocked the government’s budget adjustment in retaliation for the
leak, which he perceived to be politically motivated, while the president of
the Câmara, Eduardo Cunha (also of PMDB), went on stand in a
parliamentary investigative committee and denied his involvement in the
case. Both reactions became emblematic, as the government struggled for
the rest of the year in pushing the 2015 budget through Congress, and
because those leaks ostensibly started the antagonism between Cunha and
the government.
The third major leak, publicized in O Globo on July 15, 2015, also
targeted Eduardo Cunha, and Cunha repeated his reaction by blaming the
government for the leak, and augmented this by calling for his party PMDB
to remove their support for the government. Splitting PMDB, the largest
center party in the Congress, in two, would in effect create a sitting-duck
president out of Dilma Rousseff through the lack of this party’s votes. Half
of the congressmen from PMDB followed Cunha’s line, thus making it
extremely difficult for Rousseff to pass legislation or even the state budget
for the rest of 2015.
The top leader of Cunha’s party, Vice-President of the Republic Michel
Temer, was asked by the President to take on political responsibility, but
their relationship, never friendly, soured considerably as the vice-president
did little to contain the discontent of his party. Even after a cabinet reshuffle
in October, giving more minister positions to the PMDB politicians, the
government still had not managed to pass the year’s budget. During these
months, the Swiss prosecutor-general had publicized evidence, confirming
that Cunha indeed had accounts in the country, and a disciplinary process
against Cunha ensued. In November, Cunha tried to secure votes for filing
away this process by promising to block impeachment charges against the
President Rousseff. Her party members in the Ethics Commission of the
Câmara did not yield to blackmail, in the end, and voted to start a process
that would eventually lead to Cunha losing his mandate. A few hours after
this vote in the Ethics Commission, on December 3, Cunha responded by
using his prerogative as president of the Câmara to initiate impeachment
proceedings. President Rousseff had gotten her state budget passed at the
last minute, with an extraordinary deficit, but now faced a political-juridical
process charging her with the ultimate responsibility for delayed transfers
between the State Treasury and a number of public sector banks. Repaying
the interest on those delayed transfers, in order to avoid impeachment
charges for illegal obtainment of loans, increased the state budget deficit to
170 billion Brazilian reais.
A fortnight later, a week before Christmas, the federal police searched
the offices and houses owned or associated with Cunha, confiscating
documents and mobile phones. The text message material of these phones
allowed the police to draw several new connections in their probe. The
fourth major leak, publicized in Estado on January 7, 2016, comprised a
vast amount of text messages extracted from the cell phone of the
construction company boss Leo Pinheiro. Pinheiro, in messages to many
politicians, including Cunha and several other PMDB and PT leaders, had
negotiated economic support for election campaigns in various suspect
ways. The chief of cabinet was also implicated and had to evade the media
in the following weeks. Pinheiro’s company, OAS, was one of the heavy-
weight sub-contractors in the Petrobras case. OAS and Pinheiro in
particular turned out to have connections to the ex-president Lula. This
strengthened the original claim of Veja and the tabloid press, heralded back
in 2014 (and even before the Petrobras case, see Damgaard 2015a), that
Lula “knew all” about the corruption in PT and was behind it all.
This notion – that Lula somehow was behind it all – was also the main
thrust of the fifth major leak comprising 400 pages of testimony from the
imprisoned senator Delcídio do Amaral. The former PT leader and
Petrobras director had tried to help one of the arrested Petrobras CEOs flee
the country and had been caught red-handed. Following three months of
preventive arrest, Amaral opted for a plea bargain, and between the many
pages of information, first presented on March 3 by the tabloid IstoÉ, Lula
and Rousseff’s knowledge of corruption was denounced again. The senator
had also denounced a number of other notable politicians, including the ex-
president Cardoso and the 2014 presidential runner-up Neves.
Notwithstanding this, pressure mounted on Rousseff, as PMDB threatened
to follow Cunha by leaving the government altogether. In addition to this,
the marketing boss of the PT was arrested on corruption charges linked to
Petrobras.
 The ties between Lula and the construction company OAS had been
pursued by journalists of all the main media in January and February,
gradually turning up pieces of documentary evidence, witness accounts and
archive files that showed how OAS had worked on two real estate sites
linked to Lula. None of this was leaked from investigations, but it turned up
through investigative journalism. However, the federal police conducted
parallel investigations, and since the real estate case had already been
mentioned in Folha and Veja in 2015, it is reasonable to assume that the
media interest is not merely coincidental with the police investigation. In
any case, more evidence of relations between Lula and OAS accumulated;
evidence that arguably needed to be linked to actual crimes, however. Lula
was forced to testify on March 4, being escorted from his home in São
Paulo at dawn. The Globo news team was ready with cameras at the ex-
president’s house before the police arrived, which suggests a leak in the
local police force.
The leaks of Delcídio do Amaral and the much-covered testimony of
Lula also provided the backdrop for huge protests in the capitals of Brazil
on March 13. While many opposition leaders cast these protests as anti-PT,
not many succeeded in gaining political capital from these protests, as many
protesters were disenchanted with politics in general. A few opposition
politicians were threatened violently as they made their appearance in
protests in São Paulo.
The sixth major leak followed on March 16, where phone taps of Lula
were publicized and heavily covered. The legality of the taps was in
question, as was the publication. The federal judge presiding over the Lava-
Jato case, Sérgio Moro, had made the recordings of Lula’s conversations
accessible to journalists working on the case, and one recording in
particular was extraordinary: In this conversation, Lula’s interlocutor was
the president herself. Secretly taping the president and making the recording
public was, no doubt, a radical move. The reason given by judge Moro,
however, was to expose the implicit motivation for the conversation
between President Rousseff and the ex-president. Rousseff said she wanted
Lula to enter her straggling government, reinforcing her team with his
considerable political capital. Moro and the prosecutors argued that this was
only a decoy explanation, because a minister position would also give Lula
the benefit of having his case judged before the Supreme Court, which was
notoriously slow and lenient to politicians. In the recorded conversation,
Rousseff suggests that Lula could use the official document that would
effectively install him as minister “in case of necessity.” This phrase
became the object of much controversy in the following days. The
recording was breaking news in all media, triggering spontaneous protests
in front of the Presidential Palace, and the conversation was quoted ad
verbatim on the newspaper front pages the following morning.
With the steadily mounting pressure of leaks and arrests, PT could not
prevent allied parties from jumping ship. Cunha, in yet two more leaks, had
been linked to corruption in the reconstruction of Rio de Janeiro sites prior
to the Olympics, but he succeeded in slowing down the congressional
disciplinary process against him. Still in command of the Câmara, he
presided over the impeachment proceedings and scheduled the first vote for
April 17. Scores of congressmen declared their support for impeachment of
Rousseff, and several parties left the government and their minister seats
(discussed in depth below). The vote resulted in a two-thirds majority for
impeachment in the Câmara, and the impeachment proceeded to the Senate,
who would act as a final jury in another vote.
In the midst of the turmoil of impeachment, the seventh major leak was
mostly disregarded by the TV Globo, although it received some coverage in
the rest of the mainstream media. Based on apprehended documents from
the country’s largest construction company, Odebrecht, no less than 25
different political parties and hundreds of politicians were suspected of
taking bribes during campaigns. This included scores of the more important
parliamentarians that had voted for impeaching Rousseff the week before.
The bribe files of Odebrecht became the straw that bent the back of the heir
to the Odebrecht dynasty, Marcelo Odebrecht, who had been arrested
preventively and kept in prison since June 2015. Resisting up until this
point, he finally admitted to corruption charges and opted for a plea bargain,
drawing 76 managers of the many branches of the Odebrecht holding group
with him into a monumental testimony. True to form, once finalized, this
mega-plea bargain also had numerous parts leaked, but only after the
political elite had finished ousting Dilma Rousseff from the presidency.
Suspended temporarily by the Senate on May 12, Rousseff left the
Presidency to Vice-President Michel Temer. Having appointed an all-male,
all-white cabinet, Temer had to let three ministers go within three weeks as
new secret recordings were published by Folha in the eighth major leak.
Most prominently, the minister of Planning and Budgets, Romero Jucá of
the PMDB, had to take his leave after 12 days as minister. In the recorded
conversations with Sérgio Machado (a state company director and potential
target of the Lava-Jato probe), Jucá discussed how to stop the corruption
probe from reaching himself and his allies. A number of PMDB bosses,
including ex-President José Sarney, argued likewise in the conversations
that were recorded. The conversations were taped by the state company
director to be used as bargaining chip if he should be arrested and
eventually wished to negotiate a plea bargain. The very same Jucá, upon
stepping down, returned to his seat in the Senate, and participated in the
Senate’s final impeachment verdict against Rousseff on August 31, when
Michel Temer was permanently proclaimed President.
A minor, but interesting leak was provided by the independent, leftist
blog Tijolaço. The journalist Fernando Brito, who runs the blog, showed
that the supposedly independent survey company DataFolha (part of the
Folha group) had omitted crucial aspects of a publicized survey, providing
Folha information about the new president’s approval ratings. The
information of the poll was erroneous if not downright manipulative. On
July 17, Folha ran the results of the recent DataFolha survey as a graphical
element on the front page, showing that 50 percent of the respondents
would prefer Temer to Rousseff as president (Folha de S. Paulo 2016). The
actual percentage of affirmative responses to the question of whether they
wanted Temer to stay was only 3. However, the curiously restricted
question, “If you had to choose between the return of Dilma Rousseff as
president and Michel Temer staying in office, what would you prefer?” had
50 percent of respondents answering that they preferred Temer, with 32
percent preferring the return of Rousseff. Brito had discovered that
Datafolha had produced a second version of the survey to support the front
page of Folha, where questions implying hostility to Temer were omitted.
The original survey was, however, still available and Brito quickly
downloaded it, published it on his blog and had the American journalist
Glenn Greenwald (of Snowden and WikiLeaks fame) mirror the
manipulation on his site The Intercept. The situation was remarkable
because the political influence was obvious (although not a unique situation
for DataFolha; see Biroli et al. 2011), but the Folha editor Sérgio D’Avila
denied this, even though his newspaper not many months before published
the editorial, “Neither Rousseff nor Temer,” indicating the editor’s
preference for calling new elections instead of inaugurating Temer as
President.
In the latter half of 2016, the ninth major leak gradually dripped into
various media (Veja, O Globo and Folha most prominently among these).
The information derived from the plea bargain of the Odebrecht group, and
it continued to rock the boat of the new PMDB-led government. More
ministers stepped down, and the new president experienced a national
record of all-time low, single-digit approval ratings during the second half
of 2016. The plea bargain became known as the End-Of-The-World
testimony (Delação do Fim do Mundo – a term that emerged in a column
with that title in Folha, see Rossi 2016) and was heralded by the tabloid and
quality press to be of “apocalyptic proportions.” The Supreme Court judge
authorizing the plea bargain unexpectedly died in an airplane crash in
January 2017, but the Supreme Court president pushed the plea bargain
through and initiated 76 investigations against congressmen and ministers,
while redistributing 201 cases to other courts, including probes into 12 out
of the federation’s 27 state governors. A range of these cases had already
been leaked to the press, so the information of the Odebrecht material
constitutes the ninth major leak in the Lava-Jato case.
Many weeks before the Odebrecht material was actually distributed to
the various Brazilian courts, on March 15, 2016, 16 politicians were named
on the front pages of the newspapers, implicated in the corruption scheme.
The names comprised the top of yet another swathe of inquiries, for which
Prosecutor-General Rodrigo Janot requested authorization. All of the three
main newspapers had gotten access to this list and featured that information
prominently on their front pages, likening the situation to the first list of
names, back in March 2015. Folha reported that five of Temer’s ministers,
plus Lula, Rousseff, Aécio Neves, José Serra, as well as the PMDB senators
Renan Calheiros, Romero Jucá and Edison Lobão were all implicated in the
testimonies of the Odebrecht bosses. Estado highlighted Lula and Roussef,
along with the PT ex-ministers of Finance Antônio Palocci and Guido
Mantega, then the PSDB senators José Serra and Aécio Neves, as well as
the new presidents of Congress, Rodrigo Maia (DEM) and Eunício Oliveira
(PMDB). Below these names (and faces), the same five ministers
mentioned by Folha were shown. O Globo’s front page named Lula,
Rousseff, Aécio, and Serra at the top, continuing with the same names as
mentioned by the others – ministers, PMDB senators and ex-ministers of
Finance from PT. This constituted the tenth major leak, an exceptional one
because the source of the leak was revealed and discussed in the very same
media the following days.
The obvious overlap of this information became the topic of an editorial
penned by the journalistic ombudsman at Folha, Paula Cesarino Costa. She
claimed that a strictly off-the-record press conference had been held by the
prosecutor-general with a select circle of journalists, leaking exactly these
16 names (from a total of 320 requests for investigations). Furthermore,
Cesarino Costa questioned the motives of the prosecutor-general for
planting these names, omitting the many other names of allegedly corrupt
politicians that surfaced the following days. These subsequent leaks, she
argued, were channeled through the usual, one-on-one relationships
between journalists and their sources. The collective off-the-record
interview was suspect, to the ombudsman of Folha, however, because the
journalists had simply submitted to the selection criteria of the leaking
prosecutors. Cesarino Costa affirmed that the dependence of journalists
upon their sources in the Lava-Jato case was becoming excessive (Cesarino
Costa 2017a). The Estado columnist and political commentator Vera
Magalhães argued that the overlap of the 16 names was simply due to the
existing stock of leaks from Odebrecht, available in the public sphere to all
journalists. Thus, Magalhães argued that the front pages were constructed
from the news about the requests for inquiries and the existing knowledge
of the journalists (Magalhães 2017). The prosecutor-general denied any
leak, while the independent blog O Antagonista argued that federal
prosecutors not working on the Lava-Jato case (pointing, in particular, to
the ex-minister of the Rousseff administration, Eugênio Aragão) were
behind the leak. The Folha ombudsman reiterated the claim and stated that
three independent sources had confirmed the existence of an off-the-record
collective interview (Cesarino Costa 2017b).
The eleventh major leak almost resulted in the removal of Michel Temer.
The O Globo journalist Lauro Jardim, usually in charge of editing a small
column of minor political rumors, struck gold in April 2017. He got access
to information from the prosecutor-general’s office about a number of so-
called “controlled actions,” – covert operations involving secret
surveillance, civil agents, and plea bargaining individuals of the Lava-Jato
case. The operations inculpated both President Michel Temer and Aécio
Neves. O Globo’s site finally published the leak on May 17, and the
journalist stated in interviews afterwards that the timing was chosen to
avoid damaging these investigative operations. That requires knowledge
about the processes of prosecution and suggests that Jardim was told
exactly when he was cleared to go public.
The contents of this leak shocked the political arena to the core. First off,
the president was caught on tape agreeing that paying Eduardo Cunha to
shut up and not negotiate a plea bargain was a good idea. This conversation
had been taped by the owner of the meat production giant JBS, Joesley
Batista, in the presidential palace. Furthermore, a close associate of Temer
was caught with a carry-on luggage filled with money bills – half a million
real in total. Aécio Neves, meanwhile, had been caught negotiating a R$2
million kickback with Batista on the phone, in order to pay his bills for
lawyers – ironically, his defense lawyers in the Lava-Jato case. The money
delivery was handled by Neves’ relative, and delivered to an associate of
Neves’ fellow Minas Gerais senator, Zeze Perella (PMDB). While their
close aides and Neves’ sister went to prison, both Temer and Neves ended
up escaping prosecution altogether in the second half of 2017, which is
described in the next chapter.
On the basis of this narrative of 11 major leaks in the Lava-Jato case, I
will develop theoretically the claim that leaks can have perverse effects that
do not help democracy or provide transparency or accountability to the
public. In the next section, I will prepare this argument by building a brief
theoretical model for assessing the effects of leaks in different institutional
settings.

Tracing Institutional Interactions Triggered by Leaks

In this section, I will go deeper into the ramifications of the many leaks in
the Lava-Jato case. I will argue that leaks, as a subset of information flows
in the public sphere, constitute an especially powerful and dramatic style of
interaction between institutions and organizations of the judiciary system,
political system and media. I use the term “interaction” in a broad sense
here, to capture the various processes that emerge from these different
arenas. Based on the leaks described above, as well as a range of legal and
legislative processes (detailed in the next chapter), I have compiled a
number of cases where the institutions interacted or clashed in response to
the leaks. These cases are then examined with a simplified process-tracing
methodology.
First, I will briefly explain this chapter’s conception of institutions. The
field of media organizations is viewed here as one market-driven, political
institution (Hughes 2006: 47ff.), while the various branches of government
are viewed as separate institutional fields. Scholars have argued that media
organizations, because of their roles in democracies, should be viewed as
parts of a single political institution (Cook 2006, Sparrow 1999). “[T]hese
multi-layered news media institutions are, by and large, mutually
consistent; they cohere into a single actor for most intents and purposes”
(Sparrow 1999: 133). Cook pointed out several caveats of this viewpoint,
emphasizing a transactional perspective and demonstrating the
heterogeneity of media organizations. Since the media are not formally
structured like political institutions, contain much diversity within them,
and constantly communicate and transact with actors outside the
institutional walls of media, media should not be considered an institution
like Congress or a court (Cook 2006: 162). Keeping Cook’s caveats in
mind, later North European theories (e.g., Hjarvard 2013, Krotz 2014) have
cast media as one institution within the broader social and organizational
strand of theory dubbed new institutionalism (Thornton et al. 2012). In this
second wave of institutional theory, a complex concept of institutions was
proposed in order to grasp local agency and dynamics that led to change
and conflict in institutions.
Following these authors, the interactions of media institutions and other
institutions vary locally because particular institutional actors work
strategically within the “socially constructed, historical patterns of material
practices, assumptions, values, beliefs, and rules by which individuals
produce and reproduce their material subsistence, organize time and space,
and provide meaning to their social reality” (Thornton and Ocasio 2008:
101). Organizational action or reform might coalesce around one or
multiple such “patterns” in a local context. This theoretical approach
defines institution as one repertoire of logics, organizing principles, and
vocabularies for articulating and enacting agency and subjectivity.
Because of the attention to variant patterns and logics, new
institutionalism allows for further sub-distinctions between the institutions
of democratic systems, such as the political institutions of parliaments and
governments, the judiciary institutions of courts, prosecuting bodies and
auditing bodies. The sub-distinctions among Brazil’s accountability
institutions are only discussed in the next chapter, however. Admittedly, the
Brazilian judiciary is not a homogeneous actor but rather a many-headed
hydra of hierarchy and multi-layered bureaucracy, just as the legislative
branch of government is (in its representative or delegative nature) a
multiple political body configured to contain within it conflicts of interest.
Furthermore, the public prosecutors play a central role in the web of
accountability institutions, but they are only one among many actors
investigating and processing corruption. Congressional committees and
hybrid budget tribunals also produce horizontal accountability and act as
checks and balances, but they are partly controlled by legislative actors
supporting the executive branch. However, for the purposes of this chapter,
the sub-distinctions of institutional intersections and allegiances need not
occupy us, because the leaks, as we saw above, mostly juxtapose the three
branches of government (the executive, legislative, and judiciary) to each
other, as well as to the media.
Having preliminarily (and deliberately in a simplified manner) outlined
the actors, I will now turn to the situations of the Lava-Jato scandal where
these institutional actors and spheres interact and react to each other in
consequence of leaks. In the following analysis of institutional interactions
emerging in the wake of leaks, I will use a simplified, case-centric variant
of process-tracing (Beach and Brun Pedersen 2013) that merely aims at
explaining the outcomes of a particular case. The idea here is neither to
approach the case deductively, nor try to develop generalizable mechanisms
of leaks, because a number of scope conditions (Falletti and Lynch 2009)
are not generalizable. Scope conditions include the “relevant aspects of a
setting (analytical, temporal, spatial, or institutional) in which a set of initial
conditions leads … to an outcome of a defined scope and meaning via a
specified causal mechanism or set of causal mechanisms” (ibid.: 1152). The
scope conditions of this case, such as the Brazilian media system, the
Constitutional framework, and the legal jurisprudence, to name a few
contingencies, are particular to this national context, and cannot easily be
generalized to leaks and scandals in other contexts. However, the insights of
the case-centric process-tracing may very well be helpful to comparative
research into different settings, and to generate new questions.
The process-tracing methodology adopted here will simply point out the
instances where a leak came to be a necessary, but not in itself sufficient,
cause for an outcome. Trying to establish every necessary condition would
be a huge undertaking, and not helpful to the research scope in this chapter.
This section will instead approach the case more modestly by “crafting an
explanation that accounts for a particular outcome” (Beach and Brun
Pedersen 2013: 11). Here, I divide outcomes into two categories, depending
on the institutional setting, and define these post hoc as changes in the state
of the relevant system.
The terms “interaction” and “conflict” are defined, in turn, on the basis
of such outcomes:

• Political outcome: In the political system, an outcome could be

changes in coalition formation, the passing or rejection of a bill, or
other movements with implications for democratic governance.
• Judicial outcome: In the legal system, an outcome could be changes in
legal status, such as indictments, or pressing charges, or moving a case
from one court to another.
• Institutional interaction will be defined here as an event where the
outcome of one institutional setting or system impacts upon another
(e.g., legal actions against parliamentarians, policy aimed at
obstructing justice, the political appointment of a judge, etc.).
• Conflicts between institutions are more loosely defined here, as a
subset of institutional interactions, usually involving public critique,
protests, appeals, or contestation through either media or institutional
measures.

The Leaks about Eduardo Cunha and Judiciary Intervention in the

Legislative Branch
I will start out by illustrating one process, pertaining to the preventive arrest
and indefinite imprisonment of Eduardo Cunha, former president of the
Câmara, on corruption charges. This outcome was the final result of a
protracted struggle in Congress to remove him from office and from his
mandate, which would also result in Cunha losing the prerogative of trial
before the Supreme Court. In this case, I will argue that a necessary, but not
by itself sufficient, condition for this outcome was the leak that exposed
Cunha’s lie in the CPI of Petrobras.
 As described, Cunha’s name had cropped up in leaks in the early days of
the Lava-Jato probe. As newly elected president of the Câmara, he went on
stand on March 12, 2015, in an investigative commission he himself had
recently opened to examine the Petrobras corruption. There, Cunha calmly
denied a number of charges and openly claimed that he did not have any
overseas bank accounts. Cunha’s defense was covered as an event of
redress and vindication, full of eulogies for Cunha from both opposition
parties and the center-left parts of government (e.g., Mascarenhas and
Talento 2015, in Folha). In the commission meeting, Cunha opened a flank
of attack against the Prosecutor-General Rodrigo Janot, first claiming that
the government had forced Janot to leak information in order to weaken
him in Congress, and later arguing that Janot was persecuting him extra-
officially. The same critique, not coincidentally, was used by Michel
Temer’s lawyers in 2017 as well as the Supreme Judge Gilmar Mendes.
However, the assertion that Cunha made about the bank accounts was
belied by the facts uncovered by the MPF. On September 21, the lobbyist
João Augusto Henriques was apprehended, and he admitted to having
transferred more than R$1 million in kickback money to Cunha’s account in
the Swiss bank Julius Baer. This information spread from the MPF to the
parties in Congress, and the small left-wing party PSOL, on October 1,
demanded official access to that information from the prosecutor-general.
Although not revealing the files themselves, the prosecutor responded in the
affirmative to the question of whether Cunha had bank accounts in
Switzerland. With that response, and the earlier denial of bank accounts in
mind, PSOL filed a disciplinary action against Cunha in the Ethics Council
of the Câmara on October 13. In the meantime, files documenting the bank
account had been leaked to the press. Despite substantial pressure, because
he was in control of the Câmara, Cunha managed to stall even the simple
delivery of that action to the Council for a fortnight, and the opening of the
process was delayed for a full month.
Fast-forwarding to mid-2016: Finally, after a tortuous eight-month
process, the Ethics Council decided (in a very close vote) to approve a
recommendation for the removal of Cunha. At that point, Cunha had
already been removed from the presidency of the Câmara by the Supreme
Judge Teori Zavascki. Cunha appealed, and the action moved to the
Commission of Justice in the Câmara, but after one month, the appeal was
rejected. His substitute at the head of the Câmara, Rodrigo Maia, then had
to schedule a plenary vote for removal of his mandate, and on September
12, 354 days after the Swiss bank account was first detected, Cunha lost his
mandate, 450 votes against 10. The next month, Cunha was arrested and
imprisoned preventively in Curitiba at the order of the judge, Sérgio Moro,
and was sentenced to 15 years in prison, in the first of several indictments
in March 2017.
Cunha’s sentence was the upshot of the leak, in the sense that the other
parliamentarian (Senator and ex-President Fernando Mello de Collor)
indicted by the prosecutor-general on similar charges in the same month as
Cunha (August 2015) remains in Parliament and was only charged before
the Supreme Court in August 2017; but Collor has neither been removed
from office nor convicted. Thus, the processes of the Supreme Court delay
the sentencing substantially. Without the removal of his mandate in
Congress, Cunha would have preserved the prerogative called foro
privilegiado, escaping the jurisdiction of the Curitiba court handling the
Petrobras case. (The foro prerogative means that congressmen and ministers
automatically move to the top of the Brazilian three-tier system of courts.)
Thus, Sérgio Moro would not have been able to order his arrest, and
eventually sentence Cunha. The lie in the investigative commission in
response to leaks, and the leak of documents providing the disciplinary
process with crucial evidence constituted the necessary conditions for the
outcome of his imprisonment and sentence.
The political outcome of the case was brought about with the
intervention of the judiciary into the legislative branch, and the leaks
(probably coming from public prosecutors) that provided PSOL with the
means to initiate a disciplinary process was the triggering cause of this
intervention. The imprisonment of Cunha was the first time an
(ex-)president of the Câmara was imprisoned (but repeated six months later,
in 2017, with the arrest of his ally Henrique Eduardo Alves, also of
PDMB). Cunha’s removal from office in May 2016, ordered by the
Supreme Court, was likewise a historically unique intervention in the
federal legislative body by the judiciary, albeit not connected causally to
any leak. The Supreme Court was compelled to consider the prosecutor-
general’s request for removal with urgency as it became clear that Dilma
Rousseff would be ousted, making Cunha the de facto vice-president
according to the law of presidential succession in the Brazilian presidential
system. This situation is analyzed in the next chapter.
Protecting Supreme Justice by Cancelling a Plea Bargain
Leaks also influenced outcomes within the Brazilian judiciary. A leaked
testimony from the plea bargain of Leo Pinheiro, the president of the OAS
construction company, resulted in the exceptional cancellation of the plea
bargain negotiations by order from the Prosecutor-General. In edition #2492
(August 2016), Veja published leaked material of the testimony, in which
Leo Pinheiro had described how he and his company had helped the
Supreme Justice Dias Toffoli installing security measures in his house.
Pinheiro had directed Toffoli to a security company of Brasília. The bill,
according to Veja, was paid by Toffoli, however (a fact that was omitted in
other media’s accounts of the testimony). But the mere inclusion of the
name of a supreme justice in a testimony about criminal and corrupt activity
grabbed the attention of the media.
The next day, August 22, Prosecutor-General Janot denied that the leak
came from his office, or that this part of the testimony had ever entered his
office. Janot went on to cancel the negotiation of the plea bargain entirely,
claiming that this leaked information seemed to be an attempt at baiting the
agents of the investigation. Pinheiro had already been sentenced to 17 years
of prison by Sérgio Moro at that point, so he had a weighty incentive to
negotiate a plea bargain in order to reduce the sentence. Janot claimed that
OAS, or Pinheiro’s lawyers, had leaked the testimony to Veja because the
public interest in the matter would force the prosecutor’s hand in the
negotiation. One political commentator, Luis Nassif, speculated that the
Lava-Jato task force had leaked this as payback against Toffoli, who had
cancelled a preventive arrest of a former PT minister in one of the
investigation’s many phases (Nassif 2016), and this was also argued by the
Supreme Judge Gilmar Mendes, who over the course of the Lava-Jato
investigations began to critique the methods of the probe. Like Toffoli,
Mendes would go on to release arrested individuals in the Lava-Jato case,
contrary to the pattern established in Curitiba.
The immediate cancellation of the plea bargain negotiation sent a strong
signal to other individuals implicated in the Lava-Jato case: The Supreme
Court Judges should not be mentioned. Naturally, much speculation ensued
about other secret motives for cancelling the plea bargain, given the fact
that a number of plea bargains had leaked without any requests for
cancellation. The commentator Nassif, for instance, speculated that
annulling the plea bargain of Leo Pinheiro would also mean protecting the
PSDB parliamentarians implicated in this testimony (ibid.). The same could
be said, on the other hand, for Lula, who supposedly had been in contact
with Pinheiro on many occasions and allegedly benefitted personally from
various services provided by his company OAS. In any case, the debate
demonstrated that lawyers in plea bargain negotiations seem to assume that
public attention will help defendants – or, at least, that the federal
prosecutors claim that lawyers leak with this motive.

The Phone Tap Case and the Suspension of Lula’s Nomination as

Minister
The leak case with the most far-reaching consequences during the Lava-Jato
investigation was, arguably, the leak of recorded phone conversations
between President Rousseff and the ex-President Lula. To recapitulate:
Sérgio Moro had ordered phone taps of Lula in the days preceding his
testimony. This was only suspended on March 16, and the audio files were
– according to the Curitiba court – by human error publicized to the
journalists following the case. In reality, this meant that every journalist
with access to the public parts of the investigations received a push
notification in their inbox reporting that new material had been added to the
case. That afternoon, word spread rapidly from newsroom to newsroom and
the audio files, despite being taken down from the site of the Curitiba court,
kept circulating.
A number of conversations featuring Lula were of interest to the press,
but the central object of the leak was the conversation where Rousseff
mentions his nomination, and that Lula could use the official document that
would effectively install him as minister “in case of necessity.” Because this
could be interpreted as one way of ensuring the foro privilegiado for Lula,
removing his case from the jurisdiction of Sérgio Moro and relocating the
case to the Supreme Court, it would also effectively mean that the president
and the ex-president had worked toward obstructing justice. If this was the
case, the leak was – arguably – justified. If not, Moro had been out of
bounds by not only ordering the recording of the president (which is the
exclusive jurisdiction of the Supreme Court, but only if he or she is
officially investigated), but also by publicizing this in relation to a criminal
case without authorization. Compounding the problems of jurisdictional
competency and the legality of publicity was the fact that this particular
conversation, in contrast to the rest of the recordings, had occurred after the
order for ending the phone taps had been issued by Moro. Thus, as evidence
in a trial, it would not even be legal on formal grounds.
Nevertheless, the argument for obstruction of justice became the
centerpiece of a number of petitions, filed by diverse independent
individuals (lawyers and parliamentarians among these), requesting various
Brazilian courts for a suspension of Lula’s appointment as minister.
Because of the national character of a minister, all federal courts could
actually effect such a suspension. As the day wore on, rulings came from
several courts, decreeing such a preliminary suspension of Lula’s position
as minister, and a few rulings from other courts annulled those suspensions.
Confusion mounted as to the legality of the petitions, as well as the status of
Lula in relation to a hypothetical arrest order issued by Moro. For the
duration of those minutes in which Lula was minister, he held the
prerogative of the foro privilegiado, and responded to the Supreme Court.
As an ordinary citizen without holding public office, he was in the
jurisdiction of Moro insofar as his case related to the Petrobras case. In
total, 50 petitions were considered by Brazilian courts, and 13 of these were
filed in the Supreme Court, with 9 of these being allocated to the Supreme
Judge Gilmar Mendes.
While protests were gathering in front of the Palacio de Planalto that
afternoon, journalists were working frantically to check out all the material
of the leaked audio files. That night, the Jornal Nacional started by
introducing the leaks as the height of the crisis of the Lava-Jato scandal:
The first words of the anchor William Bonner were “A crise no governo
Dilma Rousseff atinge o ponto mais alto” (“the crisis of the Dilma Rousseff
government peaks”). The peaking crisis was then related to the nomination
of Lula for the cabinet, without mentioning at this point why a nomination
by itself could constitute a crisis. The implicit justification was given in the
third sentence of the newscast, where Bonner repeated the suspicion of the
Lava-Jato task force and the judge of Curitiba: The nomination would let
Lula escape the reach of Sergio Moro and the Lava-Jato investigations.
Thus, the crisis presented in the opening sentence was predicated on the
nomination having this subversive intention. The newscast went on to
report live, on site, concerning the protests in the federal capital. More
protests ignited the same night and continued the following day in the larger
cities of Brazil.
The following morning, the conversation was quoted ad verbatim on the
newspaper front pages. By midday, Rousseff swore Lula in as minister in a
ceremony in the Palácio do Planalto, but shadows of doubt and attrition
were showing on the face of Lula. During the ceremony, Rousseff pointed
out that his appointment as minister (termos de posse) had not been signed
in advance by her – and that this declaration, therefore, would have been
null and void as a legal shield for Lula if Moro had ordered his arrest.
Rousseff asserted that the phone taps were illegal, and the two PT leaders
signed the document and exited the palace soon afterwards.
Two days after the leak, on the 18th, Supreme Judge Gilmar Mendes
ended the controversy between the different courts and judges of diverse
jurisdictions by suspending Lula’s nomination as minister. Historically
affiliated with the main opposition party PSDB (and nominated for the STF
during the last PSDB government), it came as no surprise that Mendes
ratified the suspensions doled out by other judges. In his verdict, Mendes
affirmed the malevolent intentions of Rousseff:

The goal of the lie is clear: Avoiding compliance with an arrest order
from the first-level judge. A kind of amnesty, issued by the President of
the Republic. That is, the issuing demonstrates not only the objective
elements of deviant finality, but also the intent of fraud.
(Mendes 2016: 33, translation mine)

As it turned out, no arrest order would come from Sérgio Moro, but the
STF anyway filed away the remaining counter-petitions in Lula’s favor. As
the impeachment process removed Rousseff from the presidency, all
ministers were discharged, and the judicial proceedings on the matter
ground to a halt.
The case was emblematic, because the leaking actor was clearly
identified, and because it clearly demonstrated the power of leaks combined
with judicial measures. A government powerless to nominate ministers1 is,
arguably, hardly a government. The level of judicial interference in a matter
of the executive branch was extraordinary, based on evidence with only
dubious legality. These leaks – or rather the mediatized spectacle, sparking
protests and spates of speculation – discredited Lula and Rousseff in the
eyes of their opponents, and crucially demonstrated that the government
was fighting an uphill battle, not just against political opponents, but also
actors of the judiciary. In the following chapter, several aspects of this case
are discussed further.

Leaks Against Ex-Presidents

Rousseff and Lula would not be the only ex-presidents to feel the power of
leaks, however. In fact, all of the elected Brazilian presidents since the
return to democracy (that is, all but Itamar Franco) have been implicated by
leaks of the Lava-Jato probe. But these leaks have provoked a varied range
of reactions and responses in the legal arena, from the complete absence of
inquiries to a historical sentence: The first civilian president, José Sarney
(who was only indirectly elected vice-president by the Colégio Eleitoral in
January 1985, and took the mantle from the dying president-elect Tancredo
Neves at the moment of his inauguration) was caught on tape in a leak in
May 2016. In the series of secret recordings of conversations made by the
Transpetro director Sergio Machado, Sarney was caught discussing how he
could help Machado avoid being arrested by the Lava-Jato probe. The
recordings were leaked to Folha immediately after Temer took power, and
in June 2016, the prosecutor-general requested the immediate imprisonment
of Sarney along with the other interlocutors of Machado. The STF denied
the request. Sarney has not been known to be the object of the Lava-Jato
investigations, even though he had retired from the Senate in 2015, and
therefore was not protected by the foro privilegiado nor in the jurisdiction
of the STF.
Another former president, Fernando Collor de Mello (who served in
1990–1992 as the first democratically elected president after the
dictatorship, and who became the first Brazilian president to be removed
through impeachment) was also indicted several times by the prosecutor-
general based on evidence from the Lava-Jato investigations, and Collor is
facing charges in the Supreme Court in one of these cases at the time of
writing. The evidence against Collor emerged from the very first plea
bargains of the case. Along with Eduardo Cunha, he was among the first
politicians to be denounced in the Lava-Jato case, and he is the third senator
who is facing a trial in the Supreme Court case.
 The arrested senator Delcídio do Amaral, ex-PSDB, ex-PT and ex-
Petrobras manager, testified in prison and implicated the next elected
president, Fernando Henrique Cardoso (1995–2002). Amaral’s testimony
depicted corruption as a generalized model of management in Petrobras in
the 1990s, and documents found in his house suggested that Cardoso was
well aware of the fact.
As described in the overview of the leaks, as well as in the section
above, both Lula and Temer were implicated in the Lava-Jato case by leaks
at several points. Temer was, as vice-president and president, shielded from
denunciations of corruption in periods outside of his current mandate, just
as Dilma Rousseff was during her presidency. After the so-called
“controlled actions” of the Lava-Jato investigation in 2017, however, the
prosecutor-general filed two indictments, to be ratified by Congress before
the corruption charges against Temer could be sent to the STF. This was
legal because the evidence collected in those phases of investigations
indicated criminal activity while Temer was functioning as president. On
August 2, 2017, the Congress voted against the first indictment, 263 to 267
votes, protecting the executive from this interference, and on October 25,
251 congressmen voted against and 233 for the indictment. Meanwhile,
Lula, as we saw above, did not gain any such protection despite his
historical influence. Five different corruption cases (resulting in nine trials)
were brought to bear on the ex-president, and on July 12, 2017, Lula
became the first convicted ex-president ever in Brazilian history. Sérgio
Moro sentenced him to nine years and six months in the case of the
beachside apartment refurbished by the construction company OAS, and the
federal appeal court augmented the sentence to 12 years and 1 month. Lula
was incarcerated in Curitiba on April 7, 2018.

Media Outlets under Attack

Latin American governments in the last decade have been highly critical of
the region’s media (Kitzberger 2014), and Brazil is increasingly conforming
to that pattern. The attention given to the Lava-Jato scandals has pushed
Lula, Rousseff, Temer, as well as many other politicians into aggressive
stances vis-à-vis the Brazilian media outlets. Throughout the three years of
the Lava-Jato case, politicians implicated in the case have repeatedly
denounced the Brazilian media for being biased, printing slander,
manipulating, lying and actively interfering in political matters related to
publication of leaks. In Chapter 5, the reader may observe the exact level of
bias of attention in three Brazilian media outlets, while the following
example is merely an illustration. It serves to illustrate that almost no matter
what they do to prevent it, political actors perceive that they are at the
mercy of the media.
The first major leak in the Lava-Jato case, publicized by the weekly
magazine Veja, initiated a two-year period of intense dispute between the
press and the presidency. At the final debate between the two presidential
candidates left in the race, the incumbent Rousseff lashed out against Veja:

I cannot keep silent facing this act of election terrorism, articulated by

Veja and its secret partners. It is a shameful attitude for the press and it
damages our democratic tradition. Without providing any concrete
evidence, and yet again based on supposed declarations made by
persons from the criminal underworld, this magazine tries to involve
me and President Lula in the Petrobras situation under legal
investigation. All of the voters know about the systematic campaign
this magazine has been waging for years against Lula and me, but this
time, Veja trespassed all boundaries.
(Dilma Rousseff, cited in Chaia 2016: 50, translation mine)

At this point, PT, Lula, and Rousseff initiated a range of lawsuits and
libel suits against Veja for calumny. In total, nine different lawsuits were
initiated, but the freedom of expression clause of the Constitution protected
Veja in each of these cases. The same pattern has been true of other weekly
magazines and Globo.
The openly confrontational stance of Rousseff (and, in 2016, Lula)
toward specific media outlets (see Almeida 2016) aligns Brazil with many
of the leftist governments in neighboring countries, most notably Argentina
in the Kirchner period (Kitzberger 2014). However, while Lula in his
second mandate launched a public broadcasting policy (Matos 2012:
161ff.), Rousseff never got around to challenging mainstream media
through reform. In the legal arena, the PT leaders had to accept that their
constitutionally protected right to respond to media coverage only got them
into the very last paragraphs of articles. Other examples of libel suits have
involved Eduardo Cunha, Aécio Neves, and a range of other prominent
politicians, but the conflict initiated by leaks has not – in the Lava-Jato
case, anyway – resulted in substantial changes in the way these media
outlets cover corruption.

The Effect of Leaks upon Coalition Formation

The political scientist Manuel Balán has convincingly showed how that
corruption denunciations have been used strategically in Latin American
democracies during the last decades (Balán 2011, 2014). Denunciations are
useful in two different strategic moves: Either as a convenient way of
leaving a government coalition (which Balán calls “to jump ship”) or as a
way of increasing power within government at the expense of other
coalition partners (which is termed “leap-frog”). Thus, the public
denunciation, or leak to one or more media organization, provides a pretext
for challenging and renegotiating power relations between parties, but also
between Congress and the presidency. In some cases, Balán notes (2011:
462), governments are not under much pressure from a weak or fragmented
opposition in Congress, in which case corruption leaks are more useful for
leap-frogging than for jumping ship. With a strong opposition, there are
more constraints for the leap-frog strategy, while jumping ship becomes
viable if (but only if) there is space in the opposition coalition for another
party.
As seen in the case of Eduardo Cunha’s break with the government (the
third major leak of the case, described above), the leaks and subsequent
denunciations of corruption in Brazil provided exactly that: A pretext for
jumping ship. Eventually, the jump-ship strategy of PMDB was wildly
successful, giving the party control of the presidency and remaining in
control of the Senate. In the following, I will describe how the leaks fed this
jump-ship strategy and gradually undermined the coalition formation of the
Rousseff government. I will heuristically use cabinet positions as a proxy
for the distribution of coalition goods (Raile et al. 2011), as the ideological
coherence of Brazilian parties is often weak. This means that the use of roll-
call votes as a proxy for government coalition yields an unstable picture,
while the cabinet positions are more stable over time.
At the start of her second mandate, Dilma Rousseff counted upon the
following ten parties to govern: PT, PMDB, PSD, PP, PR, PRB, PDT, PTB,
Pros, and PCdoB, which tallied up to 329 members of the Câmara’s 513
seats (distributed in the following way: PT 70, PMDB 66, PSD 37, PP 36,
PR 34, PTB 25, PRB 21, PDT 19, Pros 11, and PCdoB 10). In the Senate,
the coalition numbered 57 parliamentarians out of 81 seats (distributed in
the following way: PMDB 18, PT 14, PDT 8, PP 5, PR 4, PSD 3, PTB 3,
PCdoB 1, Pros 1) (D’Agostino 2014). In both houses of Congress, the
coalition numbered around three-fifths of the congressmen – enough to pass
constitutional amendments, if need be.
However, these numbers covered the fact that a large number of the
coalition parties had elected conservative politicians for their seats. This
created a center-right majority in Congress that throughout 2015 succeeded
in passing an unusually high number of bills authored by the legislators,
rather than the executive branch. The frenetic pace of legislative activity
was directed by the PMDB leaders Eduardo Cunha and Renan Calheiros,
occupying the positions as presidents of the two houses of Congress.
The cabinet positions of the coalition were distributed with a heavy
emphasis on PT leaders (nine slots) and experts from civil society, leaving
six minister slots for PMDB (most of them heavy-weight ministries such as
Agriculture and Energy), while one slot was distributed to each of the minor
coalition partners: Pros, initially, had Cid Gomes as minister of Education,
PSD’s Gilberto Kassab headed the Ministry of Cities, PCdoB’s Aldo Rebelo
was in charge of the Defence Ministry, PRB’s George Hilton headed the
Ministry of Sports, PP had Gilberto Occhi in the Ministry of National
Integration, PR was represented by Antônio Carlos Rodrigues as Minister
of Transports and Manoel Dias of PDT as Labour Minister. The PMDB
politician and former president of the Câmara, Henrique Eduardo Alves
entered the cabinet in April, after a minor reshuffle.
As Eduardo Cunha became the center of attention with the leak
implicating him in the Lava-Jato case, he urged his party PMDB to break
with the government. In response, president Rousseff decided to offer more
cabinet positions to PMDB and closed the deal on October 2, 2015. The
number of ministries was reduced from 39 to 31, but PMDB gained a
cabinet position, going from six to seven ministers.2 However, the
government still struggled to find support in Congress, and the final
approval of the year’s fiscal goal was only passed on December 2.
The minister of Civil Aviation, Eliseu Padilha of PMDB, left the cabinet
on December 4, just after the impeachment petition had been accepted by
his party colleague Cunha. That became the first sign that the government
coalition was ripping at the seams. Over the Christmas holidays, and the
Carnival season in February, political negotiations slowed, but in early
March 2016, a number of events transpired that undermined Rousseff and
PT. Two major leaks (numbers 4 and 5, in the overview of leaks above)
rocked the Rousseff administration, Lula was forced to testify, and protests
dominated the streets of the state capitals on the weekends. From then on,
the parties of the coalition began withdrawing support: On March 16, the
PRB announced that it would leave the government. “This crisis cannot be
endured, or be allowed to continue. The Brazilians are suffering,” said the
PRB president Marcos Pereira, referencing the wave of protests (on March
13) and the negative news about the PT (especially the leak of the
imprisoned Senator Delcídio do Amaral the week before). The party also
announced that its minister of Sports, George Hilton, would leave the
position. In response, Hilton left PRB, announced his affiliation with Pros,
and remained minister of Sports until the impeachment of Rousseff. The
then-president of PTB, Cristiane Brasil, likewise, announced that she would
vote for the impeachment of Rousseff, on March 18, although the party only
officially recommended impeachment on April 13.
PMDB had been deliberating internally when to leave ever since Cunha
urged the party to leave the year before. During the parliamentary recess
and the internal elections for party president, the discussions had
demonstrated increasing hostility toward Rousseff, and a party convention
was called shortly after the 2016 carnival. The agenda was simple: A vote
whether or not to leave the government. Henrique Eduardo Alves left his
minister position on the eve of the convention, on March 28. The following
day, PMDB pulled off the shortest party convention ever seen in Brazil, as
the sole point on the agenda was resolved in three minutes. After a
symbolic vote, the senator Romero Jucá, who would become minister for a
fortnight before resigning after a new leak, announced that the motion for
withdrawing from government was approved. Next to Jucá stood Eduardo
Cunha, and the event ended with the 100 delegates shouting “Brasil pra
Frente, Temer Presidente” (“Go Brazil, Temer for President”). Nonetheless,
the ministers Celso Pansera, Marcelo Castro, and Katia Abreu remained
ministers in the Rousseff administration until she was ousted.
Sensing that a new majority was about to form, the strategy of jumping
ship suddenly became attractive to the rest of the coalition parties, and the
exit of PMDB therefore started a wave of defections: On April 12, PP
announced their decision to abandon government, support impeachment,
and Occhi left the Ministry of National Integration the following day. The
same day, April 13, the PSD group leader in the Congress, Rogerio Rosso,
declared that there was a broad consensus about supporting impeachment,
and that the minister of the party, Kassab, was known to respect the will of
the parliamentary group. Kassab resigned two days later, despite stating that
he “was convinced of the President’s personal integrity.” The
parliamentarians of PR never decided upon an official party line and ended
up mostly voting for impeachment (with the votes 26 for and 10 against),
but the PR minister Rodrigues continued in the cabinet until the final
removal of Rousseff.
The process of coalition breakdown (and subsequent formation) in
2015–2016 was a two-step process, triggered by leaks: First, in July 2015,
the initial conflict between the legislative and executive branches, or rather
between Cunha and Rousseff, and then, the leaks targeting Lula and PT in
March 2016, which influenced and boosted the large-scale protests in the
streets of the capitals on March 13 and 16.
Most of the parties jumping ship went on to adhere to the line of PMDB
and the new president throughout 2016, essentially replacing PT with
PMDB in charge. The new government was thus based on center-party
votes, just like the Rousseff government, but crucially augmented by the
votes of the former opposition-leading party PSDB.
The strategy of jumping ship, however, later came to be used against
PMDB and Michel Temer as the JBS leak turned public. The large coalition
partner PSDB became divided, had one minister (Bruno Araújo) state that
he would leave the cabinet and two smaller coalition parties also left (PTN
and PPS, with the minister Roberto Freire). Araujo only left the cabinet six
months after the scandal, however, as PSDB was preparing to leave the
government en bloc. The rupture of the Temer government coalition
resulted in further delays to the reform bills in Congress in 2017.

Preliminary Conclusion
I will sum up the preceding sections by highlighting several interesting
aspects of leaks from these cases. Media leaks about corruption provide
political actors with opportunities for shifting political allegiances, and
leaks also provide the conditions for a range of exceptional judicial
interventions into politics and even conditions for political interference in
accountability processes. This means that leaks may be harnessed for
political effect through four mechanisms. First, leaks may rapidly bypass
normal gatekeeping and dominate the media agenda, and this will be
discussed further in Chapter 5. Second, by virtue of their extraordinariness,
leaks absorb attention and this also means eclipsing other information on
certain actors; in other words, the leaks work as a distraction-spin, by
removing emphasis from political attempts at active agenda setting. Third,
this locates specific politicians in defensive or reactive positions in media
narratives, as news on leaks frequently assesses the political capital of
denounced and implicated actors, ultimately signaling potential outcomes of
political and judicial processes and the relative probabilities of these
outcomes. Fourth, leaks are crucial for setting the media agenda at timely
moments before previously scheduled public protests and manifestations,
thereby mobilizing and supporting the goals of these protests. Leaks thus
provide legitimation for planned or spontaneous protests, but they may also
legitimate legal actions on the horizon, such as upcoming trials or expected
arrests.
The timing of leaks is also crucial. The visibility conferred by leaks to
certain bits of scandalous information worked at key moments of the
political events that eventually resulted in the impeachment of the president.
With this, I claim that the political effect of these leaks went beyond the
mere sequence of scapegoating and scandalizing parties in the mainstream
media: At critical moments, the leaks shed light on certain elements in the
public sphere (evidence, connections, and allegations) and what they made
visible in this light, then, was a certain knowledge – knowledge of the
moral comportment of politicians, as well as knowledge of positions of
denunciation and defense. This is further elaborated in the final discussions
of this chapter.

Leaks and the Question of Truth, Knowledge, and Visibility in Public

Spheres
In this section, I will move the problem of leaks into more philosophical
territory and explore how leaks relate to knowledge in public spheres. I will
make a general argument about the desire for transparency in contemporary
democracies, and then show that this produces a particular kind of visibility
in the Brazilian mainstream media. First, however, I will consider how
information becomes knowledge and how knowledge can be known as
truthful in public spheres.
When particular news items reporting on the Lava-Jato investigations
and leaks are closely examined, it becomes apparent that such texts at times
play hard and fast with the truth – or at least work in a gray area of
conjecture and conflicting truth claims. Grammatically, many news items
exhibit far more sentences in the conditional and subjunctive than usually
seen in the news. There are many reasons for journalists to produce news
with speculative content, having to do with the availability and quality of
information in leaks, especially leaks of testimonies. Specifically,
information gathered by journalists might be incomplete, based on not
verified (or not verifiable) witness accounts, and it may often come from
obviously biased sources, or in other ways be of debatable credibility. This
leads reporting to revolve around speculation and inference, and, in order to
stay within (mainstream) journalistic ethical bounds, it requires the
journalists to emphasize that statements of facts are merely possibly true.
News texts describing leaked information, like other forms of speech
events, organize information in more or less coherent story arcs through
intertextuality (Wortham and Rhodes 2012). One news text, by association,
connects to earlier news texts dealing with the same case or related
corruption. Therefore, if an earlier story about the corruption of an actor is
well-known, then the new text will work as confirmation (“actor A is
corrupt”) or continuation (“he was corrupt then and still is”) at the level of
reception. For each new piece of evidence, denunciation, or testimony, the
suspicions of political transgression are reinforced. This reinforcement
works through the ongoing link of news media texts, accumulating on top
of earlier texts in the chain of scandals and leaked testimonies, by
suspending the ambiguity or implied uncertainty embedded in each
individual text. This suspension is effected by weight of numbers and
through blunt repetition. In other words, repeated news items about
corruption will seemingly demonstrate that the denunciations are factual,
bringing to mind the old idiom “where there is smoke, there’s fire.”
Furthermore, news texts do not merely talk about the past. In scandal
news texts, it is especially apparent that the question of consequences is
also of great interest, and such texts will usually include one or more
sentences about the reactions (demands for justice, for instance) and the
likely scenarios triggered by the revelations. Thus, through these
intertextual connections and speculative predictions, story arcs emerge,
projecting certain futures for the implicated actors. Thus, grammatically, as
well as conceptually, revelations and corruption stories branch into the
fictional territory, and therefore, when written coherently, scandalous
information turns into “a story” (Bird 1997), in both the journalistic sense
(Tuchman 1978) and the narrative sense.
No matter the grammatical ambiguity and hedging, the information
presented in news items then turns into public knowledge about corruption.
This knowledge (of stories and actors) spins a web of facticity (ibid.: 82),
constructed and mutually legitimized by various media institutions
(Kristensen and Mortensen 2013). Repeatedly recycled information from
leaks winds up as truth because the watchdog role of the media carries
expectations of truth: Audiences (or at least audiences before the age of
post-factual politics) expect media to check facts, to investigate, and to
correct their coverage if new information comes to light. So, a continuous
chain of news texts about a parliamentarian’s corruption will normally be
considered a tell-tale sign that he is in fact corrupt – even if each text
included many conditional clauses and textual hedges. Audiences presume
some kernel of truth despite ambiguity, qualifying clauses, and conjecture
in the grammatical content of the news reports. This is one aspect of the
regime of truth (Foucault 1980) governing modern mediatized societies:
News media coverage represents truth, and that assumption has only lately
been called into question with the increased attention to fake news and post-
truth politics.
Truth and knowledge is always constructed, however. One of the seminal
philosophers of social sciences, Michel Foucault, argued that truth was
produced by institutions, politics, and social practices. He suggested that we
can think of society as constituted in part by certain regimes of truth,
defining and thereby creating what normally passes for accepted truth:

Each society has its regime of truth, its “general politics” of truth: that
is, the types of discourse which it accepts and makes function as true;
the mechanisms and instances which enable one to distinguish true and
false statements, the means by which each is sanctioned; the techniques
 and procedures accorded value in the acquisition of truth; the status of
those who are charged with saying what counts as true.
(Ibid.: 131)

Among other things, Foucault worked with the installment of truth

through mental institutions, medical professions and techniques. The
practices surrounding these defined (and continues to define) what it means
to be sane and insane, thus partly defining truth and erecting the boundaries
for the accepted experience of the world. Media organizations, meanwhile,
install truths about the leaders and decision-makers of our world. The
regime of knowledge relevant to leaks is that which John Thompson has
termed “the new visibility” – a unidirectional, mediated episteme (or
horizon of knowledge), structured by a global framework of media
technologies, experiential and cultural forms, peculiar to the age of
television and internet. In Thompson’s words, the emerging visibility
regime of the modern (if not the post-modern) sphere of media creates “a
society in which it was possible and, indeed, increasingly common for
political leaders and other individuals to appear before distant audiences
and lay bare some aspect of their self or their personal life” (Thompson
2005: 38). Thus, politics across the globe are in part governed by the
techniques and practices that allow our leaders to appear before us – which
is, needless to say, media techniques and practices. The disclosure of the
personal back-region (ibid.: 44, Goffmann 1956) has become the norm of
politics – and a willful manipulation of that disclosure, though perhaps
always present and always expected, is the centerpiece of many scandals in
the late twentieth and early twenty-first century.
The mediated visibility that Thompson speaks of enables the public to
“know” their leaders, and audiences generally suppose that media render
politics and political personalities visible. Leaks shed a new light on this
knowledge in the public sphere at critical moments with disclosure of
evidence, connections, and allegations. What leaks make visible in this
light, then, is a more specific kind of case-based knowledge – a knowledge
of the moral comportment of politicians, as well as knowledge of rhetorical
positions of denunciation and defense. This shows that audiences are also
aware that sometimes, leaders deceitfully provide disinformation or
misinformation (see Chapter 5 concerning this distinction). Tellingly, TV
viewership in Brazil is negatively correlated with support for political
actors such as government, Congress, and parties (Mesquita 2014: 24). The
possibility of deceit by media personalities is always present, and leaks
either confirm suspicions of deceit and disinformation, or, if the allegation
is proven wrong, underscore the truth of the leaks’ possible impact upon a
specific political career. In sum, when performing the role of society’s
watchdog (Waisbord 2000), media organizations act in the role of an arbiter
of truth in the public sphere at the same time, and the power inherent in this
role is tremendous.
A politician in public office is assumed to promote the collective will of
his or her constituency – that is the justification for electoral and
representative democracy. However, there is also the background
assumption that the public office should not be used for personal gain at the
expense of that collective goal. In ordinary politics, that assumption goes
largely unnoticed, but it becomes very visible with leaks. Indeed, leaks are
interesting to the public because they present a rupture in that assumption.
Leaks thus perform a double check on the visibility maxim already in place:
Did this politician present a truthful picture of herself? Did he or she show
the true backstage persona on the screen, in the radio or in text, or at least a
representation in which the supporters could believe? Thompson points out
that this check is a crucial test for politics at large in these conditions of
mediatized visibility:

People become more concerned with the character of the individuals

who are (or might become) their leaders and more concerned about
their trustworthiness, because increasingly these become the principal
means of guaranteeing that political promises will be kept and that
difficult decisions in the face of complexity and uncertainty will be
made on the basis of sound judgement. The politics of trust becomes
increasingly important, not because politicians are inherently less
trustworthy today than they were in the past, but because the social
conditions that had previously underwritten their credibility have been
eroded.
(Thompson 2005: 46)

The kind of public-political individual groomed in our mediatized world

is a subject formed by the virtual subjugation to the regime of visibility.
Thus, the threat of a leak works like a spectral Panopticon – always
looming, always monitoring, but to an unknowable extent. Public
officeholders are punishable in consequence of their visibility, and in
consequence of the traces their actions produce in various public domains,
be it bank accounts, investments, tax information, or entertainment
preferences. The public individual must internalize this visibility in order to
present a clean backstage persona to media and (through media) to
supporters and voters. The emergence of a scandal represents a failed test of
accountability to supporters, constituencies, and the public at large: Trust
has been invested (say, by a constituency) but the trustee has failed to live
up to this trust (Boltanski and Thevenot 2006: 133ff.).
What is at stake in a scandal is not merely the trust in one individual,
however. The scandal is also a process that draws boundaries in social
space, that is, a process that allocates (negative) value to certain kinds of
behavior performed by the representative bodies of the State. With a
scandalizing leak, the news items and commentary will valorize and laud
some actions (the disclosure and distribution of the leaked information, for
the sake of the public), while the disclosed political or economic actions
(embezzling, bribing, money laundering, etc.) are condemned.
In the journalistic discourse on the disclosure of such actions through
leaks, boundaries of the social space are drawn, and certain sets of
knowledge about the society are constituted, and media thereby provides
“publicly available and politically vital spaces for imagining communities”
(Damgaard 2015: 419ff.). The imagined political communities, however,
might be imagined as corrupted (Gupta 1995). The mediatized space for
imagining the State, or a nation, is also contested, as is the knowledge
within these spaces. More insidiously, they are eroded in the sense that the
arbiters of the space are not journalists and editors with public profiles,
ethical standards, and explicit agendas subject to scrutiny. Instead, the space
created by leaks emerges from a vault of information controlled by
unknowable leaking actors, through channels that are inaccessible to the
public.
Summing up, scandals provide knowledge about specific individuals,
their actions, and their trustworthiness, but at the same time, scandals
negotiate meaning concerning the values that govern public life in general.
One of the values of political life that is increasingly becoming important is
transparency, and the next section will discuss the relationship between
transparency, knowledge, and leaks.
Media, Leaks, and the Transparency Creep
The mechanism of leaks in public spheres taps into a larger ideological
current that has been emerging at least since the early 1990s. This current
has consisted in the promotion of transparency and a number of associated
concepts (due diligence, compliance, and to some extent corporate social
responsibility) and was catapulted unto the international scene of
governance when former World Bank employees founded the NGO
Transparency International. This section explores how leaks play into the
discourse and practices of transparency in the global media ecology. First, I
will define the contours of transparency discourse, and then point out how
leaks (especially leaks about corruption) fit into the mold provided by this
discourse.
Although the contemporary discourse concerning transparency is of a
recent vintage, it is a societal value with deep roots. It is the selective
application of a more general moral principle inherited from the
Enlightenment:

As a moral principle, however, transparency, stretched farther than

calls for self-repair of oversight systems. It proposes an overall
“antiseptic treatment” of visibility for the society. Its canonical roots
also date back to Enlightenment political imaginaries, deriving from
Jeremy Bentham’s utilitarianism, which claims that maximum
transparency contributes to maximal happiness … A diluted version of
this legacy is embedded in professional journalism, as the notions of
the Fourth Estate and watchdog function of journalists suggest.
(Heikkilä and Kunelius 2017)

Heikkilä and Kunelius highlight that watchdog journalism and the

political imaginary of transparency are linked. This is particularly evident
now, in this precise historical moment, where leaks enter the scene: A
combination of particular forms of data handling, information distribution,
and oaths and technologies of secrecy, with journalistic collaboration across
national borders and an at times uneasy commitment to traditional news
values. In this way, leaks provide journalists with the raw material of the
scandals of the twenty-first century, quite different in its incipient form
from the pattern of Watergate (two dedicated journalists digging after facts
for several years). The outcome – for instance, media pressure and popular
protests against corrupt leaders – might still be the same, but the raw
material and the labor of the scandal looks much different now, because of
the whole gamut of server technology, hacker communities, social media,
and popular post-Watergate mythology.
It is not just media technologies and their viral potential that make leaks
fit exceedingly well with transparency discourse, however. A keyword in
transparency discourse is accountability, and this is another cross-over
between scandals and transparency efforts. As Michael Power and other
British scholars have argued (Power 1997, 2000, Strathern 2000), a
transparency creep has been globally visible for decades, so that audits and
accountability are invasive concepts in many social and institutional
settings: “The concept of audit in turn has broken loose from its moorings
in finance and accounting; its own expanded presence gives it the power of
a descriptor seemingly applicable to all kinds of reckonings, evaluations
and measurements” (Strathern 2000: 2). Expanding this argument, a decade
later, Hansen and Flyverbom note that

the project of exposing the hidden is important in many contexts,

ranging from business and politics to science, but when put into
practice it often ends up concealing more than is revealed, such as
when an account of something in the past simplifies so much that it
eliminates important, contextual information. In this way, transparency
can easily become part of the problem it was intended to solve,
producing concerns with distortion, concealment and collusion.
Similarly, we need to consider whether highly touted phenomena like
“big data” will make it possible to uncover the unknown while
remaining sensitive to contextual dynamics … or rather lead to new
types of opacity. More generally, the project of transparency elicits a
particular stance towards what counts as truth and certainty. When
transparency is “rolled out” programmatically in organizations or
societies, typically with a view to correcting failures or empowering
employees and citizens, it is understood narrowly as a superior mode of
knowledge, a cultural signifier of unmediated objective information.
(Hansen and Flyverbom 2014: 2)

Leaks are, in an interesting way, connected to the transparency culture

and the desire for ever more disclosure. The technological organization of
WikiLeaks, for instance, is based in the alternative democratic and
informational ideals of hacker communities, where the disclosure of any
and all information is good in principle and practice. Nick Davies, a
journalist at the Guardian, working with WikiLeaks, described the ideology
of the WikiLeaks founder: “The problem is [that Julian Assange is]
basically a computer hacker. He comes from a simplistic ideology, or at that
stage he did, that all information has to be published, that all information is
good” (quoted in Leigh and Harding 2011: Ch. 8, para. 24). In this ideology
lies a tension with principles of news value (Chadwick 2013), and
moreover, as seen above, a tension with the gatekeeping role of journalism
– and the position of the journalists as disclosing agents in public spheres.
The ideology of leaks, furthermore, is interesting in relation to the
question of knowledge. As Hansen and Flyverbom argue above, when
information is exposed, even if it distorts or conceals, it is positioned as
unmediated and objective because of its disclosure. This is logically flawed,
of course, but it works at a symbolic level: Leaks create a new terrain of
knowledge which has been less contested than many other scandals,
because leaks are imagined as rooted in data and facts.
Scandals were, to be sure, always a potential setup – the Dreyfus affair
of France at the turn of the last century being the most prominent, classic
case. Leaks, in contrast, ostensibly enter the public sphere as ready and
transparent information. The information of leaks is apparently spawned as
truth but may not be indicative of any criminal acts or moral wrongdoings.
In the Lava-Jato case, the leaks at times seemed to be quite true but tedious
and non-transgressive, and the power/knowledge relation they enacted was,
instead, that of generally tarnishing the image of an individual politician.
By leaking pieces of information cast as evidence in a trial (although no
trial existed at that time), anything could be made to indicate something
heinous and a secret motive.
Scholars (such as Allern and Pollack 2012, Thompson 2000, Tumber and
Waisbord 2004) have thought of scandals as a dramatization of political or
moral conflict, an occasion for deliberating the limits of public behavior, for
judging transgressive behavior, and expelling, catharsis-style, offensive
moral phenomena from the body politic. Media are ubiquitous in these
processes of fertilizing the ground for scandalous denunciations and then
reaping the rewards of attention through quasi-public processes of sense-
making and extra-judicial trial. And media scandals have always been
contested and objects of critique – media hunts are perceived to be
sensationalizing, dramatizing, and individualizing, or just plain wrong. But
with leaks, informational content is being pushed into public spheres in a
qualitatively new way, in a manner that is rarely questioned on
epistemological grounds. This innovation reflects the wider cultural and
socio-political developments: The episteme and imaginaries of visibility of
modern media worlds (Krotz 2014: 84ff., Thompson 2005) have evolved
since Watergate, and leaks are now forcing political practices into the
limelight, promising transparency and accountability. When they do so with
vested political interests, leaks may circumvent both journalistic pretensions
to un-biased or objective truth, judicial pretensions to fair trials, and societal
pretensions to citizenship under a rule of law. This targeted transparency
can have perverse effects, however; some of these effects have been
demonstrated in this chapter. The ultimate consequences of media
disclosure of corruption for the Brazilian democracy is a thread that runs
through the following chapters and is treated in detail in the final chapter.

Notes
1 The Rousseff administration had, earlier in March, nominated Wellington César Lima for
Minister of Justice, but his nomination was also overturned, because Lima was employed as
prosecutor in the MPF, and this overlap is not permitted under the Constitution.
2 The second cabinet of 2015 was announced on October 2 by Dilma Rousseff. The reshuffle
reduced the number of ministries from 39 to 31, and retained most of the minor coalition
partners’ ministers, but installed several new ministers from the ranks of PMDB: Kátia Abreu,
Leonardo Picciani, Eliseu Padilha, Eduardo Braga, Helder Barbalho, Marcelo Castro, Henrique
Eduardo Alves, and Celso Pansera represented PMDB in the cabinet. PT had nine slots, and the
cabinet included eight ministers not affiliated politically. The other coalition partners with
ministerial positions were PDT (André Figueiredo), PSD (Gilberto Kassab), PTB (Armando
Monteiro), PCdoB (Aldo Rebelo), PRB (George Hilton), PP (Gilberto Occhi), and PR (Antônio
Carlos Rodrigues).

References
Albuquerque, A.L. (2017) “Sem exposição, é impossível avançar contra poderosos, afirma
Dallagnol” Folha de S. Paulo, p. A4, November 24. Retrieved from
www1.folha.uol.com.br/poder/2017/11/1937812-sem-exposicao-e-impossivel-avancar-contra-
poderosos-afirma-dallagnol.shtml.
Allern, S. and Pollack, E. (eds.) (2012) Scandalous! The Mediated Construction of Political Scandals
in Four Nordic Countries. Göteborg: Nordicom.
Almeida, R. de (2016) A Sombra do Poder. Os Bastidores da Crise Que Derrubou Dilma Rousseff.
Rio de Janeiro: Editora LeYa.
Arantes, R. (2002) Ministério Público e Política no Brasil. São Paulo: Sumaré/Fabesp.
Arantes, R. (2011) “The Federal Police and the Ministério Público.” In T. Power and M. Taylor
(eds.), Corruption and Democracy in Brazil (184–217). South Bend, IN: University of Notre
Dame Press.
Balán, M. (2011) “Competition by Denunciation: The Political Dynamics of Corruption Scandals in
Argentina and Chile” Comparative Politics, vol. 43(4), 459–478.
Balán, M. (2014) “Surviving Corruption in Brazil: Lula’s and Dilma’s Success Despite Corruption
Allegations, and Its Consequences” Journal of Politics in Latin America, vol. 6(3), 67–93.
Beach, D. and Pedersen, R.B. (2013) Process-Tracing Methods: Foundations and Guidelines. Ann
Arbor: University of Michigan Press.
Benkler, Y. (2014) “A Public Accountability Defense for National Security Leakers and
Whistleblowers” Harvard Law and Policy Review, vol. 8, 281–326.
Bird, E. (1997) “What a Story! Understanding the Audience for Scandal.” In J. Lull and S. Hinerman
(eds.), Media Scandals: Morality and Desire in the Popular Culture Marketplace (99–121). New
York: Columbia University Press.
Biroli, F., Miguel, L.F., and Mota, F. (2011) “Mídia, eleições e pesquisa de opinião no Brasil (1989–
2010): um mapeamento da presença das pesquisas na cobertura eleitoral” Revista Compolítica,
vol. 1(1), 67–89.
Boltanski, L. and Thévenot, L. (2006) On Justification. Economies of Worth. (C. Porter, Trans.).
Princeton, NJ: Princeton University Press.
Cabra, M. and Kissane, E. (2016) “The People and the Tech Behind the Panama Papers” The Source,
April 11. Retrieved from https://source.opennews.org/en-US/articles/people-and-tech-behind-
panama-papers/.
Cesarino Costa, P. (2017a) “Um jato de água fria” Folha de S. Paulo, March 19. Retrieved from
www1.folha.uol.com.br/colunas/paula-cesarino-costa-ombudsman/2017/03/1867852-um-jato-de-
agua-fria.shtml.
Cesarino Costa, P. (2017b) “Pela transparência e a boa prática jornalística” Folha de S. Paulo, March
26. Retrieved from www1.folha.uol.com.br/colunas/paula-cesarino-costa-
ombudsman/2017/03/1869768-pela-transparencia-e-a-boa-pratica-jornalistica.shtml.
Chadwick, A. (2013) The Hybrid Media System: Politics and Power. Oxford: Oxford University
Press.
Chaia, V. (2016) “O Impeachment da president Dilma Rousseff????” Em Debate, vol. 8(2), 47–54.
Chaia, V. and Teixeira, M. (2000) “Máfia dos Fiscais e as estrelas de cidadania” Observatório da
imprensa, ed. 63, Mídia e política, 31–37. Retrieved from
http://observatoriodaimprensa.com.br/primeiras-edicoes/mfia-dos-fiscais-eas-estrelas-da-
cidadania/.
Coleman, G. (2014) Hacker, Hoaxer, Whistleblower, Spy: The Many Faces of Anonymous. London:
Verso.
Cook, T. (2006) “The News Media as a Political Institution: Looking Backward and Looking
Forward” Political Communication, vol. 23(2), 159–171.
D’Agostino, R. (2014) “PT e PMDB encolhem, mas mantêm maiores bancadas; PSDB cresce” G1
Eleição em Números, special section of G1 website, published October 6, 2014. Retrieved from
http://g1.globo.com/politica/eleicoes/2014/blog/eleicao-em-numeros/post/pt-e-pmdb-encolhem-
mas-mantem-maiores-bancadas-no-congresso-psdb-cresce-na-camara.html.
Dallagnol, D. and Martelo, O. (2016) “Lava Jato, de onde veio e para onde vamos” Folha de S.
Paulo, October 30. Retrieved from www1.folha.uol.com.br/opiniao/2016/10/1827555-lava-jato-
de-onde-veio-e-para-onde-vamos.shtml.
Damgaard, M. (2015) “Multiple Margins and Mediatized Transgression” Ephemera –Theory and
Politics in Organization, vol. 15(2), 411–434.
Delli Carpini, M. and Williams, B. (2001) “Let Us Infotain You: Politics in the New Media Age.” In
W.L. Bennett and R. Entman (eds.), Mediated Politics: Communication in the Future of
Democracy (160–181). Cambridge: Cambridge University Press.
Drezner, D.W. (2016) “Nothing to See Here, Just the Biggest Global Corruption Scandal in History,”
Washington Post. Retrieved from
www.washingtonpost.com/posteverything/wp/2016/04/04/nothing-to-see-here-just-the-biggest-
global-corruption-scandal-in-history/?utm_term=.97e0725c5542.
Falleti, T.G. and Lynch, J.F. (2009) “Context and Causal Mechanisms in Political Analysis”
Comparative Political Studies, vol. 42, 1143–1166.
Folha de S. Paulo (2016) “Cresce otimismo com a economia, diz Datafolha” Folha de S. Paulo, front
page, July 17, 2016.
Foucault, M. (1980) Power/Knowledge: Selected Interviews and Other Writings 1972–1977. New
York: Pantheon.
Fuchs, C. (2011) “WikiLeaks: Power 2.0? Surveillance 2.0? Criticism 2.0? Alternative Media 2.0? A
Political-Economic Analysis” Global Media Journal (Australian Edition), 5(1), 1–17.
Galtung, J. and Ruge, M. (1965) “The Structure of Foreign News” Journal of Peace Research, vol.
2(1), 64–91.
Goffmann, E. (1956) The Presentation of Self in Everyday Life. Edinburgh: University of Edinburgh.
Gupta, A. (1995) “Blurred Boundaries: The Discourse of Corruption” American Ethnologist, vol.
22(2), 375–402.
Hansen, H.K. and Flyverbom, M. (2014) “The Politics of Transparency and the Calibration of
Knowledge in the Digital Age” Organization, vol. 22(6), 872–889.
Heikkilä, H. and Kunelius, R. (2017) “Surveillance and the Structural Transformation of Privacy.
Mapping the Conceptual Landscape of Journalism in the Post-Snowden Era” Digital Journalism
Studies, vol. 5 (3), 262–276.
Hirschl, R. (2008) “The Judicialization of Mega-Politics and the Rise of Political Courts” Annual
Review of Political Science, vol. 11(1), 93–118.
Hjarvard, S. (2013) The Mediatization of Culture and Society. London: Routledge.
Hughes, S. (2006) Newsrooms in Conflict: Journalism and the Democratization of Mexico.
Pittsburgh, PA: University of Pittsburgh Press.
Hutchby, I. (2001) Conversation and Technology: From the Telephone to the Internet. Cambridge:
Polity.
International Consortium of Investigative Journalists (2016) Panama Papers. Retrieved from
https://panamapapers.icij.org/.
Kitzberger, P. (2014) “Demands for Media Democratisation and the Latin American ‘New Left’.”
Hamburg: GIGA German Institute of Global and Area Studies.
Kristensen, N.N. and Mortensen, M. (2015) “Amateur Sources Breaking the News, Metasources
Authorizing the News of Gaddafi’s Death: New Patterns of Journalistic Information Gathering and
Dissemination in the Digital Age” Digital Journalism, vol. 1(3), 352–367.
Krotz, F. (2014) “Media, Mediatization and Mediatized Worlds: A Discussion of the Basic
Concepts.” In A. Hepp and F. Krotz (eds.), Mediatized Worlds (72–87). London: Palgrave
Macmillan.
Landerer, N. (2013) “Rethinking the Logics: A Conceptual Framework for the Mediatization of
Politics” Communication Theory, vol. 23, 239–258.
Leigh, D. and Harding, L. (2011) WikiLeaks: Inside Julian Assange’s War on Secrecy. London:
Guardian Books.
Llanos, M. and Marsteintredet, L. (eds.) (2010) Presidential Breakdowns in Latin America: Causes
and Outcomes of Executive Instability in Developing Democracies. London: Palgrave Macmillan.
Magalhães, V. (2017) “Castelo de Areia 2?” Estado de S. Paulo, March 22. Retrieved from
http://politica.estadao.com.br/noticias/geral,castelo-de-areia-2,70001709012.
Maier, J. (2011) “The Impact of Political Scandals on Political Support: An Experimental Test of
Two Theories” International Political Science 32(3), 283–302.
Matos, C. (2012) Media and Politics in Latin America. Globalization, Democracy and Identity.
London: I.B. Tauris.
Mendes, G. (2016) DJE nº 54. Decisão monocrática. Brasília: Supremo Tribunal Federal. Published
on March 22, 2016. Retrieved from http://portal.stf.jus.br/processos/downloadPeca.asp?
id=308995628&ext=.pdf.
Mesquita, N.C. (2014) “Media and the Quality of Democracy: The Different Impacts of the Media on
Regime Support in Brazil.” In R. Figueiras, P. Espírito Santo, and I.F. Cunha (eds.), Democracy at
Work: Pressure and Propaganda in Portugal and Brazil (13–38). Coimbra: University of Coimbra
Press.
Miguel, L.F. (2003) “A Eleição Visível: A Rede Globo Descobre a Política em 2002” Dados, vol.
46(2), 289–310.
Moro, S. (2004) “Considerações sobre a Operação Mani Pulite” Revista CEJ, vol. 8(26), 56–62.
Munro, I. (2016) “Whistle-Blowing and the Politics of Truth: Mobilizing ‘Truth Games’ in the
WikiLeaks Case” Human Relations, 69(1), 1–25.
Nassif, L. (2016) “A vingança torpe da Lava Jato contra Dias Toffoli” JornalGGN, published August
20, 2016. Retrieved from https://jornalggn.com.br/noticia/a-vinganca-torpe-da-lava-jato-contra-
dias-toffoli.
Pérez-Liñán, A. and Rodríguez-Raga, J.C. (2009) “Veto Players in Presidential Regimes: Institutional
Variables and Policy Change” Revista de Ciencia Política, vol. 29(3), 693–720.
Porto, M. (2011) “The Media and Political Accountability.” In T. Power and M. Taylor (eds.),
Corruption and Democracy in Brazil (103–126). South Bend, IN: University of Notre Dame Press.
Power, M. (1997) Audit Society: Rituals of Verification. Oxford: Oxford University Press.
Power, M. (2000) “The Audit Society – Second Thoughts” International Journal of Auditing, vol. 4,
111–119.
Raile, E., Pereira, C., and Power, T. (2011) “The Executive Toolbox: Building Legislative Support in
a Multiparty Presidential Regime” Political Research Quarterly vol. 64(2), 323–364.
Rossi, C. (2016) “A Delação do Fim do Mundo” in Folha de S. Paulo, published on March 23, 2016.
Retrieved from www1.folha.uol.com.br/poder/2016/03/1753135-a-delacao-do-fim-do-
mundo.shtml.
Russell, A. and Waisbord, S. (2017) “The Snowden Revelations and the Networked Fourth Estate”
International Journal of Communication, vol. 11, 858–878.
Seligson, M. (2002) “The Impact of Corruption on Regime Legitimacy: A Comparative Study of
Four Latin American Countries” The Journal of Politics, vol. 64(2), 408–433.
Shoemaker, P. and Vos, T. (2009) Gatekeeping Theory. New York: Routledge.
Strathern, M. (2000) Audit Cultures: Anthropological Studies in Accountability, Ethics, and the
Academy. London: Routledge.
Sparrow, B. (1999) Uncertain Guardians: The News Media as a Political Institution. Baltimore, MD:
Johns Hopkins University Press.
Tate, N. and Vallinder, T. (1995) The Global Expansion of Judicial Power. New York: New York
University Press.
Thompson, J.P. (2000) Political Scandal: Power and Visibility in the Media Age. Cambridge: Polity
Press.
Thompson, J.P. (2005) “The New Visibility” Theory, Culture & Society, vol. 22(6), 31–51.
Thornton, P. and Ocasio, W. (2008) “Institutional Logics.” In R. Greenwood, C. Oliver, K. Sahlin,
and R. Suddaby (eds.), The SAGE Handbook of Organizational Institutionalism (99–129). Los
Angeles, CA: Sage.
Thornton, P., Ocasio, W., and Lounsbury, M. (2012) The Institutional Logics Perspective. A New
Approach to Culture, Structure and Process. Oxford: Oxford University Press.
Tuchman, G. (1978) Making News. A Study in the Construction of Reality. New York: The Free
Press.
Tumber, H. and Waisbord, S. (2004) “Introduction” American Behavioral Scientist, vol. 47, 1143–
1152.
Wahl-Jørgensen, K. (2014) “Is WikiLeaks Challenging the Paradigm of Journalism? Boundary Work
and Beyond” International Journal of Communication, vol. 8, 2581–2592.
Waisbord, S. (2000) Watchdog Journalism in South America. News, Accountability, and Democracy.
New York: Columbia University Press.
White, D.M. (1950) “The ‘Gate Keeper’: A Case Study in the Selection of News” Journalism
Quarterly, vol. 27, 383–391.
Wortham, S. and Rhodes, C.R. (2012) “Narratives Across Speech Events.” In A. De Fina and A.
Georgakopoulou (eds.), The Handbook of Narrative Analysis (160–177). Malden, MA: Wiley-
Blackwell.
4 Lawfare and the Judiciary-Political Relations
during the Lava-Jato Corruption Scandal

 
In Brazil, holding corrupt politicians accountable has proven to be
notoriously difficult (Power and Taylor 2011). The Lava-Jato investigations
have been heralded as an exception to that rule, although in reality, very few
of the many investigated politicians have been sentenced or charged thus
far. In this chapter, I will explore the various ways that the legal actions
triggered by the Lava-Jato probe have produced or failed to produce
accountability, as well as the perverse effects of judicial intervention in the
political system of Brazil. The chapter provides sections on each of the
relevant accountability institutions and brief case studies of the most
important legal actions and trials – distinct or entangled, depending on the
case – that together constituted the core of Lava-Jato scandal. Together with
the analysis of leaks in the previous chapter, these case studies explain the
trajectory of the scandal. Several of these legal actions generated sustained
media coverage, which contributed to a second stage of the informational
cascade, wherein the same signal of political corruption was recycled. The
question I will answer in this chapter is how judicial processes work in
contemporary Brazilian politics, and how weak accountability mechanisms,
despite their shortcomings, may nevertheless become potent political
weapons when combined with leaks and cascades of disclosure.
I will start the chapter by introducing the main actors in the
accountability and audit institutions of the Brazilian system, describing
their various institutional traits, and discuss their general role in
accountability processes in the political arena. Readers eager to learn about
the specific cases can skip ahead to the subsequent sections, where I will
consider the interplay of political actors, prosecuting authorities, and
judicial institutions. First up is the curious process in the Supreme Electoral
Court, which, based in the Lava-Jato investigations, detected fraud in the
2014 presidential campaign in the amount of R$112 million. The trial,
however, ended up acquitting the defendants, Dilma Rousseff and Michel
Temer. Second, I will show how accountability mechanisms targeting
corruption in the Câmara dos Deputados are incapacitated in the current
configuration of the Congress, with the exception of the former president of
the Câmara, Eduardo Cunha. Third, Cunha’s case had consequences for the
two cases described next, in which individual judges of the STF tried (but
failed) to remove two of the most important Brazilian Senators, Renan
Calheiros and Aécio Neves. Fourth, I turn to the processes against the ex-
president Lula and his family, observing how the timing of the
investigations, denunciations, and sentencing was ultimately crucial to his
chances of running for president in 2018.
Throughout the analysis, I will show that the outcomes of accountability
processes are rarely the punishment of political transgression, and that the
Lava-Jato case was not simply a case of lost political capital for those in
public office. This finding has resonance with extant research, but the
chapter contributes to research on Brazilian accountability institutions by
showing that several macro-level changes and various side-effects surfaced
in the political arena due to legal processes spawned during the Lava-Jato
investigations. In discussing the outcomes of the legal processes, I will
argue that political actors use legal action and corruption denunciations
strategically to wage political war through the institutions of law and the
repressive power of the state, which has been termed “lawfare” (Zanin
Martins et al. 2017, drawing on Comaroff and Comaroff 2006). The term
has already been the object of much debate in Brazil, since the lawyers of
the ex-president Lula began to use the term when criticizing the Lava-Jato
investigations. Seen from the opposite perspective, the lawfare waged in
Brazilian courts is one tactic within the quest to combat corruption, advance
justice and remedy the (perceived) lack of electoral accountability.
Theoretically, I contend that despite the imprecision in relation to the
original concept, it can be fruitful to import the concept of lawfare into the
context of contemporary media spheres and, especially, mediatized politics
(Lima 2004, 2009). In the Latin American historical experience of many
presidential and democratic breakdowns (Llanos and Marsteintredet 2010),
such a concept gives researchers purchase on the mediatized relationship
between legal actions, political agency, and contested constitutions.
Accountability Institutions in the Political Arena

Judicial interventions in politics are common in Brazil at many levels, and

the inclination toward settling political disputes through law has been
termed judicialization (Avritzer and Marona 2014, Werneck Vianna et al.
1999). International scholars have conceived of “judicialization of politics”
as either the kind of macro-change observed when legal concepts and
practices spread through political discourse and political organizations, or
“the reliance on courts and judicial means for addressing core moral
predicaments, public policy questions, and political controversies” (Hirschl
2008: 94). Such increasing reliance on judicial means results in increasing
judicial power in the political arena, which also leads to gradual structural
and communicative changes. Thus, the term judicialization may also
simultaneously encompass the contemporary transformations of judicial
institutions and the structural shifts seen in judiciary communication (Couso
et al. 2010, Peruzzotti and Smulovitz 2006: 19, Tate and Vallinder 1995).
The concept of judicialization may only be relevant here at first glance,
however. In response to scandals, increasing attention to legal frameworks
and judicialization is one political strategy used to signal accountability and
maintain organizational legitimacy. It is therefore to be expected that courts
and accountability institutions enter political arenas in a political system
rife with allegations of corruption. It is also well established that such
allegations may trigger horizontal accountability mechanisms and
subsequently become elements within intra-elite power struggles (e.g.,
Peruzzotti 2006). As most literature on judicialization deals with judicial
interventions and policy review of areas that are not commonly the object
of legal processes, I will not draw on the literature on this subject in the
following. Rather, the questions at stake in this chapter will aim to focus
more on the dimensions and mechanisms of this accountability.
The importance of horizontal accountability in the web of Brazilian
accountability institutions is linked with weak electoral accountability.
Because of this weakness, the horizontal accountability institutions and the
media institutions are typically the main accountability actors when
corruption is disclosed in Brazil. Despite the fact that the activation of
either accountability or media institutions in a scandal should entail
reputational costs for the politicians involved, those suspected of corruption
(as well as those who have been indicted or charged) frequently manage to
get re-elected. Several arguments have been advanced to explain this:

• There is a lack of information regarding corruption in the media at a

local level, often because of illegal overlap between owners of the
media and the political elite, or due to other protections negotiated
through clientelistic networks (Abramo 2007, Lima 2004). Despite the
fact that voters prefer not to elect politicians they know to be corrupt,
corrupt politicians often manage re-election because their
constituencies are not sufficiently aware of the corruption (Winters and
Weitz-Shapiro 2013).
• Another argument, contrary to the above, is that there is plenty of
information available regarding corruption among the political elite,
and precisely because of this, ideological preferences become more
important determinants of votes. In the 2006 presidential elections,
Rennó showed that “[p]erception of corruption matters only when it
comes to choices between candidates on the same side of the
ideological spectrum” (Rennó 2011: 71). The rationale thus being: “If
they are all corrupt, I might as well vote for the politician who, albeit
corrupt, is closest to my ideological viewpoint.” Comparing the re-
election frequency of congressmen in the 2006 elections, Pereira et al.
(2011: 94) found that parliamentarians involved in the Mensalão
scandal that sought reelection were not at a statistical disadvantage
relative to fellow congressmen. Furthermore, at the local level, voters’
clientelistic rationales and expectations of pork-barrel politics may
also play into the choice of voting for corrupt politicians: In 184
municipal campaigns for the position of mayor in the state of
Pernambuco, Pereira and Melo (2015) found that incumbent mayors
could mitigate the negative impact of corruption denunciations and
special audits upon their re-election chances by increasing public
spending.
• It has been argued that Brazilian voters are tolerant of corruption, as
long as the politician in office is also perceived to be efficient. The
adage “rouba mas faz” (“he steals, but gets things done”) famously
described the São Paulo governor Adhemar de Barros (Cotta 2008).
This argument also explains why politicians such as Paulo Maluf
remained in office for decades (until December 2017), despite being
 internationally convicted, wanted by Interpol, and targeted by 572 suits
in Brazil since the 1980s (Arantes 2011: 199). The hypothesis of
“rouba mas faz” has been tested and explored in various studies (e.g.,
Ferraz and Finan 2008, Pereira and Melo 2015, Pereira, Melo, and
Figueiredo 2009; Rennó 2011, and Winters and Weitz-Shapiro 2013).

Several initiatives have been implemented over the past two decades in
order to remedy this weak electoral accountability. Most importantly, the
law of the Clean Slate (Lei da Ficha Limpa, see Koerner 2013, Michener
and Pereira 2016: 500, Presidência da República 2010, Taylor 2011: 170)
was passed in 2010. The law prevents individuals from running for public
office if they are sentenced for corruption, administrative improbity, and
common crimes. It has been applied in thousands of candidatures since it
was enacted, and during the 2016 municipal elections, around 1,600
candidates were impeded from either running or taking office (Pontes
2016).
Despite the Ficha Limpa law coming into force in 2010, the current
national Legislative Assembly still consists of numerous parliamentarians
facing trials: 48 out of 81 senators are under investigation or charged before
the Supreme Court, as is roughly a third of the Câmara at the time of
writing (Macedo 2017a, 2017b, Taylor 2011: 164). The accountability
mechanisms are not as effective as one could hope, and some of the
mechanisms can even be rigged for political effect. While the Ficha Limpa
law remedies this somewhat, a potential perverse side-effect has arisen from
the combination of the Lava-Jato trials and the Ficha Limpa law. This side-
effect would play an important role in the general elections of 2018.
The prohibition of candidature warranted by the Ficha Limpa law
requires either an admission of guilt, or that the politician in question is
found guilty of corruption or political maladministration and sentenced
twice – by an ordinary court as well an appeals court, or in the Supreme
Court. This ratification clause of the Ficha Limpa law makes the timing and
sequence of investigations, trials, and eventual appeals of the Lava-Jato
cases very important. Since the trials against parliamentarians were not
concluded (indeed, many had barely moved to the initial stage) before the
candidate deadline in August 2018, the Ficha Limpa law did not apply to
politicians in office, allowing them to run for re-election. In the case of
Lula, who expected to run for the presidency in 2018, the sentence of 9.5
years in prison on corruption charges (on June 12, 2017) would not by itself
stop him from running, but the decision of the appeals court of the Fourth
Federal Region in Porto Alegre, on January 24, 2018, however, disqualified
him from running due to the Ficha Limpa law. At the time of writing,
further appeals to the Supreme Court and the supreme electoral court were
expected.
The Ficha Limpa law thus floated, like a Damocles sword, above each of
the political actors that once seemed most likely to run for president after
Dilma Rousseff’s second term. Like Lula, senators Aécio Neves and José
Serra (of PSDB), both previous presidential runner-ups, are currently the
subject of investigations by the Supreme Court and sentencing here will
render them ineligible for public office. Unlike Lula (and other Brazilians
without a federal mandate), Neves and other politicians in office however
benefit from the so-called law of privileged right (foro especial por
prerrogativa de função, or in normal parlance foro privilegiado). The foro is
a legal reminiscence of colonial times, automatically relocating to the
Supreme Court all cases involving federal representatives. The prerogative
means that congressmen and ministers automatically move to the top of the
Brazilian three-tier system of courts. Similarly, state governors can only be
tried before the Supreme Tribunal of Justice (Arantes 2002). In 2017,
38,431 individuals could claim the foro on the basis of the federal
constitution, while 16,559 individuals had similar rights under the state
constitutions (Barroso 2017, Bulla and Moura 2017). Upon losing or
renouncing a mandate (or failing to obtain re-election), a case involving a
politician is then redistributed to a local court. If obtaining a new mandate,
for example in a State Assembly, the case is then relocated once more. This
provides denounced parliamentarians with opportunities for manipulating
this mechanism of the legal system, stalling processes, etc.
Beyond the Ficha Limpa law, several other legal instruments threatened
the politicians suspected of corruption. Preventative arrests, for extended
periods of time, constituted another means of legal intervention by the
public prosecutors, and this was the subject of much debate during the
Lava-Jato investigations. Based on the evidence turning up in the probe, the
Supreme Electoral Court also came close to annulling the result of the 2014
presidential elections. Each accountability institution intervenes with
different instruments, so in order to analyze the complexities of the
horizontal accountability interventions executed during the Lava-Jato
probe, I need to clarify several important distinctions between the
accountability institutions. Apart from their differing institutional relations,
the courts, the police, and the prosecutors are also divided by their
professional goals and modus operandi. In the following, each of the
relevant judicial and prosecuting bodies and their institutional
configurations are described.

The State Prosecutors and the Ministério Público Federal

The first accountability institution described here is the Ministério Público
Federal (henceforth MPF). The MPF is the federal public prosecutorial
body of Brazil, tasked with defending societal collective rights, protecting
civil society, the environment, and public heritage, as well as curbing
corruption in public administration. Each state has a local Ministério
Público, and in total there are 10,000 state MP and MPF members
distributed across the nation (Arantes 2011: 187).
The prosecutor-general of the Republic (henceforth PGR) holds office
heading the MPF for periods of two years. The PGR in charge during the
Lava-Jato case was Rodrigo Janot, who held two consecutive mandates.
The incumbent Raquel Dodge (appointed in September 2017), is the first
woman to hold this office. The national association of Brazilian prosecutors
nominates three candidates for the position of PGR, with one of these
candidates continuing to be selected for the position by the president of the
Republic. The candidate will be questioned before taking office by the
Senate’s Commission of Justice and Constitutionality, and the candidacy
will be put to a vote in the plenary of the Senate.1
In investigations into corruption, the members of the MPF work closely
together with state or federal police forces to obtain evidence. The decision
to initiate an investigation lies with the state MP or the MPF, however
prosecutors must request authorization from the relevant local court with
jurisdiction in order to obtain legal evidence pertaining to the crime
scene(s). In the case of Lava-Jato, the initial investigations of money
laundering and smuggling happened in the southern state of Paraná, close to
the borders of Argentina and Paraguay. The investigation has remained in
the capital city Curitiba, where a small team of MPF prosecutors originally
obtained authorization to expand the scope of their probe from judge Sergio
Moro. Since then, several state MPs (of São Paulo and Rio de Janeiro) have
been handed material for local investigation by the task-force in Curitiba.
The prosecutors of Brazil work as a corporation in the sense that the
institution has a distinct organizational goal, a hierarchy of status and a
career ladder, a bureaucratic structure, and strong inwards loyalty. Arantes
highlights that the MPF of Brazil is “motivated by a strong ideological
component of … ‘political voluntarism’” (2011: 187), with members of this
institution viewing society as weak, representative political institutions as
corrupt, and the MP as the “preferred representative of a weak society,
especially in contrast to inept bureaucracies that fail to enforce the law”
(ibid.: 188). The strong professional identity of the MP members is also
externally recognized: The MPF is, in Brazilian administrative circles,
referred to as a “corporation,” and during the Lava-Jato case, many have
accused the MPF of “corporatism,” with reference to the historical
experience of the Portuguese corporatist state and the corporatist ideology
espoused by Brazilian President Getúlio Vargas, inspired by Salazar’s
Portugal (Wiarda 2004). At times in the current debate, the concept of
corporatism often simply denotes the class-interest of a particular group
within the state (e.g., Araujo et al. 2016: 112), acting to defend privileges
vis-à-vis other actors. The MPF has frequently been criticized for enacting
such “corporativismo,” but this sort of critique is variously directed at the
judiciary, teachers in the public schools, the police, public servants in
general, and parliamentarians.
Whether or not this critique rings true in the case of the prosecutors, it is
beyond doubt that the MPF has gained much autonomy and independent
discretion since the days of Geraldo Brindeiro, the PGR of the 1990s who
became known as the filing clerk of the Republic. Autonomy and
independence are prerequisites for the anti-corruption activities of the MPF,
however independence may also pose a problem when considered from
alternate perspectives. It has been pointed out that the corporatist thinking
of the MPF has harmed the national economy (by crippling major
construction firms), the legitimacy of democratic institutions (by repeatedly
and loudly scandalizing the Congress), as well as the legitimacy of the
judiciary.2 Several situations that led to these accusations are detailed in this
chapter. The legitimacy of investigations is frequently defended by
prosecutors from the Lava-Jato task force (e.g., Dallagnol and Martelo
2016), many of whom are active speakers at conferences and public events,
on social media, and even religious ceremonies. In contrast, PGR Janot was
less outspoken, as he was already criticized for appearing in the media
following the initial inquiries in March 2015 (Netto 2016: 145).

The Federal Police

The federal police force (henceforth PF) of Brazil is, in its current
configuration, quite new. The PF has its historical roots in the nineteenth-
century court of the King Dom João VI, while the name goes back to a new
organization of the police force instituted at the end of the Estado Novo
period in 1944. The following year, a unit of this organization gained
national responsibilities; however, neither the Estado Novo federal police
nor the nineteenth-century unit ever grew to become particularly effective
(Arantes 2011: 190). During the earlier, less centralized regimes, the states
of the federation displayed little interest in constructing a national police
force, and later, during periods of greater centralization and more
authoritarian regimes, the military displayed no interest in yielding either
power or jurisdiction to civilian forces. Thus, only with the first Lula
administration in the 2000s did the federal police become a significant force
for combatting corruption and increasing accountability in the political
realm (ibid.: 194). The scope and number of operations has grown from 18
operations in 2003, to 187 in 2007, and to 550 in 2016 (Polícia Federal
2017). The PF had 3,471 positions for detectives and 11,381 for agents,
clerks, and analysts in 2015 (Ministério de Planejamento, Orçamento e
Gestão 2015: 5). Due to budget cuts in 2016 and 2017, the number of
positions is now decreasing.
Because of the rising number of operations, the PF has become a more
visible entity. Arantes notes that the press “has been filled with cleverly
named police operations,” and the catchy names engage public and media
attention, giving “a sense of the result and the likely responsibility of those
being implicated” (Arantes 2011: 201). This symbolic naming of operations
provides simultaneously an informational shortcut, a hashtag-like function
for digital and cognitive retrieval, and a discursive node organizing public
debate. The media strategy of the PF is distinct from that of the MPF (see
above), and agents rarely appear in interviews and columns. This is
probably due to the organizational subordination of the PF to the Ministry
of Justice.
 In investigative operations, the MPF works in tandem with the PF, while
the state MPs work with state police agents. The police agents and the MPF
are legally interlinked:

although the Polícia Federal is subordinate to the executive branch (and

its activities can thus be curtailed or directed by the Justice Ministry),
its main tools for investigating and addressing crimes require the
agreement of independent judges and take place under the oversight of
the MPF.
(Ibid.: 192)

However, neither judges nor MPF direct the police investigations, and thus
legal processes shift between these institutions throughout the phases of an
investigation.
In addition to its subordination to the executive branch of government,
federal police work is less prestigious than judgeship or prosecutorial
positions (and also less well-remunerated than private lawyer work –
although better than other federal civil servants), meaning that the PF has a
weaker institutional identity and sense of independence than the corporation
of the MPF (ibid.).
The federal police force consists of various specialized units for specific
tasks. In cases such as Lava-Jato, a local or regional police director may
select police agents from diverse units to form a task-force with resources
dedicated to a specific investigation. Across Brazil, a range of similar
groups work on other political corruption cases and depending on the
context these groups may incorporate members from other accountability
institutions such as the Federal Tax authorities and financial audit bodies.
Regardless of composition, they depend on judges for authorization to use
force in their investigations (for incidents such as detaining individuals,
forcing them to testify, or breaking the secrecy of bank or telephone
accounts).
The Lava-Jato task-force of the Curitiba-based PF was constructed in
April 2014 from police agents and prosecutors. The PGR installed a
national team of prosecutors on the case in January 2015, while the state
prosecutors of Rio de Janeiro and São Paulo picked out task-forces in June
2016 and July 2017, respectively. The PF’s relationship of dependence or
independence vis-à-vis the MPF was contested on numerous occasions
during the Lava-Jato case. In particular, the right to negotiate and agree to
plea bargains has been a cause for strife between the MPF and the PF
during recent years. While both institutions work to obtain evidence when
negotiating a plea bargain, differences in (political) scope and intention may
emerge between the two. Such differences arise principally because the PF
is organized under the Ministry of Justice, while the state MPs and the MPF
are formally independent.
After the impeachment of Dilma Rousseff, several circumstances have
reduced the efficiency of the PF investigations of the Lava-Jato case.
Beyond the general situation of budget cuts in the federal police forces of
Curitiba and Paraná (budget cuts from R$29.1 million in 2016 to R$20.5
million in 2017), the number of agents managing the case was reduced in
May 2017 from nine to six. In November 2017, a new director of the PF
was appointed by Michel Temer, and with a new director and a new
prosecutor-general, the political control of the Lava-Jato probe is rumored
to have increased. Former PGR Rodrigo Janot publicly voiced his
suspicions in the matter via Reuters news agency (Brooks 2017).
Leaving the investigative and executive branches of the Brazilian
accountability institutions, I now turn to the four courts embroiled in the
trials of 2014–2017, starting with an example contrasting the practices and
organizational ideals of the prosecutors and police to that of the judges and
courts.

The Courts of Brazil

This section describes four courts of Brazil: The Supreme Court (STF), the
Tribunal of Accounts (TCU), and the electoral court (TSE), as well as the
13th Vara of Curitiba where judge Sérgio Moro has been conducting the
trials in the Lava-Jato case. Initially, I will contrast the institutional
perspective of the courts to that of the prosecutors through the example of
preventive arrests. The debate on preventive arrests may elucidate how
accountability is perceived and enacted in different ways in two
governmental institutions, and how the belief systems and logics of these
institutions provide governmental actors different ways to interpret specific
interventions.
 As already mentioned, the Brazilian courts contribute to accountability
in different ways than prosecuting and investigating bodies. The production
of accountability is one sub-goal of the courts, among a range of other
organizational and institutional goals and conceptualizations, dealing with
civil and corporate rights, due process, legality, the fairness of the legal
order, and compliance (Edelman 2016). The goal of producing political
accountability, it must be emphasized, has become a very important topic in
the Brazilian discourse about justice – not least because of the Lava-Jato
case and its many ramifications. To clerks and judges of the courts, this goal
may only be one among many. Although the MPF and PF are intertwined
with courts in corruption probes, the primary actors of courts, that is, the
judges, may still consider themselves distinct from these accountability
bodies on account of their professional identity and distinct work ethos. The
judges of the Brazilian courts form a professionalized bureaucracy in the
Weberian sense, valuing professional reputation, network, and individual
expertise, grounded in the professional knowledge base and language. That
public prosecutors and the judges have distinct professional ideals are
apparent in many aspects of the Lava-Jato investigations, but especially so
in the case of preventive arrests.
The rationale of preventive arrests, in the institutional logic of
prosecutors and police agents, is to secure evidence and avoid the
obstruction of further investigations. During the unravelling of the criminal
network behind the Petrobras graft, the MPF has requested and succeeded
in obtaining authorization for preventive arrests on numerous occasions.
Such authorizations are issued by courts. This kind of arrest is ostensibly
used to impede any attempts at destroying evidence or fleeing the country
(Taylor 2011: 171). However, because the Lava-Jato investigations have
continuously been revealing new layers of corrupt networks, many arrested
individuals have been kept incarcerated for extended periods of time on the
assumption that they would obstruct not just their own case, but also the
next tier of involved persons.
Extended preventive arrests did not used to be standard practice in
Brazilian courts, however. From the professional perspective of Brazilian
courts and the judges, preventive arrests interfere with several principles of
justice, especially individual freedom rights and the assumption of
innocence until culpability is proven. The courts are historically fond of
habeas corpus injunctions for the political elite, commonly letting a high
proportion of defendants enjoy their liberty during trials and appeal
processes (ibid.). The logic here is different from that of the prosecuting
body because judges are concerned with the constitutional right of freedom,
to the possible detriment of the quality of investigations. The predilection of
the courts for maintaining (upper-class) defendants and convicted
individuals in freedom has created a sense of impunity for crimes of
political corruption (Arantes 2011: 195). Meanwhile, Brazil’s courts are less
lenient in cases of petty crimes among poorer segments of the population.
The preventive arrests of the Lava-Jato case have been interpreted as an
exception to the assumption of innocence. The topic has drawn headlines,
because leading businessmen and ex-ministers such as José Dirceu (PT),
Antônio Palocci (PT), and Henrique Eduardo Alves (PMDB) have been
incarcerated between 6 and 12 months before being tried, while Lula was,
as described in the previous chapter, essentially denied the possibility of
taking office as minister in the Rousseff administration with the argument
that his nomination was really a step to avoid a preventive arrest.
The use of preventive arrest remains a hot topic in the public sphere and
between the local courts of the Lava-Jato case (Rio and Curitiba) and
individual STF judges. The diverging interpretations subsequently made it
into the commentary, op-ed pieces, and opinion columns of the Brazilian
news media, recycling the news of corruption disclosure with new layers of
legal discourse and debate on top. The debate can be boiled down to this: At
what point should political actors under suspicion of corruption be held
accountable and arrested? Should it be when suspicion first arises, or with
denunciation, or after evidence has been collected, or after one or multiple
sentences have been declared? Regardless of the correct answer here, I will
merely point out that the agitated debate concerning preventive arrests
reflects an important distinction to be made between the institutional logics
of courts and the judiciary on one hand, and the prosecuting bodies of the
State on the other. The two organizational fields, in general, do not view
transgression of laws and individual or political rights in the same way, nor
do they use the same channels for intervention and sanction.
In sum, the different roles of police and prosecutors compared to judges
have led these two groups to interpret preventive arrests quite differently.
This illustrates that although accountability institutions are organizationally
embedded in the State, the roles and functions of these different bodies are
distinct from each other but are not, however, entirely subsumed under the
democratic ideals of modern state constitutions and checks and balances.
The different institutions belong to separate “macro-level belief systems
that shape cognitions and influence decision-making processes in
organizational fields” (McPherson and Sauder 2013: 167). Concordantly,
the timing of the application of sanctions and methods of intervention differ.
The daily routines and working logic deployed by prosecutors or police
agents in corruption investigations are not dependent upon a democratic
mandate. Like the courts, these agents are ideally independent actors
working to ensure the effective enforcement of laws passed in other sectors
of the State – the executive, legislative, and judiciary branches of
government. The goal of prosecutors and police agents is to obtain weighty
and legal evidence and minimize the harm done to society by actors
perceived as perpetrators. Courts on the other hand are oriented toward
another goal: judging, in an equilibrated manner, the evidence produced in
investigations and then applying the “field-level manifestation of the logic
of the state … in the legal realm … [which] extols the virtues of just
incarceration and formal social control, the markers of traditional punitive
legal enforcement strategies” (ibid.: 172). None of these institutional logics
(Thornton and Ocasio 2008, Thornton et al. 2012) are necessarily based in
the participatory ideals of the democratic state, but despite the formal,
practical-discursive, and professional independence of the courts, they still
form a part of the government, and are in some contexts very attuned to the
political atmosphere. However, the degree to which Brazilian courts are
sensitive to political influence can vary. The Supreme Court, to which I first
turn, is the court most frequently interacting with the political branches of
government, and thus the most central piece in the accountability
mechanism of corruption scandals such as Lava-Jato.

The Supreme Court of Brazil

The Supreme Court of Brazil (STF) is in principle a constitutional court,
although it also holds extended powers for policy revision. STF is
frequently the ultimate appellate court, and acts in criminal cases related to
individuals in public office at the federal level, that is, senators, state
representatives, as well as ministers and presidents.
The STF has been and remains extremely overloaded with cases: While
77,304 verdicts have been handed down in the decade 2007 to 2017, 8,904
of these made by the court plenary (Mariani et al. 2017), but in recent years,
the court has been receiving up to 70,000 new cases per year (Falcão et al.
2016: 66). The US Supreme Court, by comparison, only decides around 80
cases each year (Boydstun 2013: 11). Despite only assessing a fraction of
this plethora of processes, Brazil’s STF has occasionally been protagonist in
the political field. During the last decade the STF has exhibited a growing
independency (Avritzer and Marona 2014: 83) and judged the
constitutionality of a range of (sometimes unpopular but still important)
cases and causes.3
The politically sensitive corruption cases of the Lava-Jato investigations,
not to mention the impeachment process, have resulted in extraordinary
public and media attention directed at the work of the STF. As the case
studies of this chapter demonstrate, the court and the individual judges of
the court have been decisive in the political outcomes characterizing
Brazilian politics in 2014 till 2017. The court has increasingly had “an
important rule-making effect on how the accountability game plays out in
the rest of the political system” (Taylor 2011: 177). Judges Zavascki and
Fachin, in the function of their role as judge rapporteurs of the Lava-Jato
cases, have been particularly important, as have the judges Mendes and
Mello, who have publicly adopted critical postures to the investigation and
the work of the MPF.
The STF consists of 11 presidentially appointed judges who may occupy
the bench until the mandatory retirement age of 70. These 11 judges are
divided into two teams (turmas) of 5 while the president of the court
remains outside of these teams, instead managing the agenda and various
other coordination tasks. In a plenary decision, the president of the court
acts as tiebreaker when votes are tied. The presidency of the STF rotates
every two years to the judge with the longest career in the STF who has not
yet presided over the court. Three of the STF judges also serve in the TSE
(see below).
The make-up of the court (at the time of writing) is a mixture of judges
chosen by Lula (three), one for each of the presidents before him (save
Itamar Franco), four chosen by Dilma Rousseff, and one chosen by Michel
Temer. The court is currently presided by Cármen Lúcia (nominated as
judge by Lula in 2006, having assumed the presidency in 2016).
First team (Primeira Turma)
• Rosa Weber (nominated by Dilma Rousseff in 2011)
• Luis Fux (nominated by Dilma Rousseff in 2011)
• Luís Roberto Barroso (nominated by Dilma Rousseff in 2013)
• Marco Aurélio Mello (nominated by Fernando Collor in 1990)
• Alexander de Moraes (nominated by Michel Temer in 2017)

Second team (Segunda Turma)

• Dias Toffoli (nominated by Lula in 2009)

• Edson Fachin (nominated by Dilma Rousseff in 2015)
• Gilmar Mendes (nominated by Fernando Henrique Cardoso in 2002)
• Celso de Mello (nominated by José Sarney in 1989)
• Ricardo Lewandowski (nominated by Lula in 2006)

The judges are, in principle, able to rule on their own (monocratically) or

collectively in the turmas as well as the plenary of the court. The decision
to bring a case to either collective is taken by individual ministers, or the
court president can schedule cases for the plenary. In collective processes in
the court, one minister is designated to lead the case as judge rapporteur
and suggest a ruling (this position is called the ministro relator), while
another acts as overseeing or auditing judge (the ministro revisor).
Monocratic decisions can also be audited and revised by other STF judges,
but this measure is usually, though not always, avoided.
The case distribution system of STF is digitized, sealed off from other
systems as well as from the judges themselves. It randomly distributes cases
according to types and is supposedly designed to equilibrate the workload
of processes between the judges. In the national political corruption cases,
individual judges normally take the case to either the team or to the plenary,
but the vast majority of other cases in the STF docket are decided by the
judge designated relator. The exact distribution algorithm is therefore of
interest in a court consisting of politically appointed judges.
The Lava-Jato case was originally headed by the judge Teori Zavascki,
because the very first Lava-Jato inquiries into politicians with foro
privilegiado were distributed to him by the algorithm. From 2014 to 2017
Zavascki managed the increasing caseload, authorizing and sometimes
denying inquiries, as well as investigative measures such as the preventive
arrests. However, the Dilma Rousseff-appointed judge died in an airplane
crash off the coast of Paraty in the state of Rio de Janeiro. This happened
precisely at the moment in January 2017 where Zavascki was about to
authorize the plea bargains of 77 Odebrecht executives, who had admitted
to large-scale corruption. When Zavascki suddenly died, the topic of
redistribution of the Lava-Jato cases became politically very sensitive. The
STF denies public access to the source code of the algorithm, so the
decision-making inherent in the code is completely opaque to the public, for
now. Without Zavascki, STF president Cármen Lúcia then authorized the
testimonies of the Odebrecht company as evidence, and the Lava-Jato case
was randomly redistributed to judge Fachin. Having obtained the
legalization of the evidence, the PGR requested and was granted approval
for the opening of further inquiries into more than 200 parliamentarians and
governors.
As already mentioned, the politicians elected for federal public office
(federal representatives, senators, ministers, and presidents) will
automatically have criminal cases judged before the STF because of the
foro privilegiado. In general, cases linger in the STF docket for years on
end; the case type with the longest delay in the STF are the penal cases –
and the penal cases against parliamentarians are the most common penal
cases in the STF (Falcão et al. 2014: 56–57). One example of the tardiness
of the system was the corruption case in the state administration of Minas
Gerais in 1998. The trial in the STF against Eduardo Azeredo, former
national president of PSDB and ex-governor of Minas Gerais, finally
commenced 11 years after the events. The verdict was handed down in
2015, 17 years later, but even then, the sentence of a 20-year prison term
was appealed. The appellate court of Minas Gerais upheld the verdict for
Eduardo Azeredo in 2017. Azeredo maintained another appeal option, and
by 2018, his penalty might be reduced or even suspended entirely because
of his age.
What is interesting about corruption cases is not the fact that such cases
are often massively delayed, as exemplified in the case of the 18-year
process of the ex-governor of Minas Gerais. The crucial question of
corruption cases, especially after the Ficha Limpa law, is the timing in the
STF. As exemplified above, cases of politicians with foro privilegiado
usually take much longer than corruption cases at a local or regional level,
with the official explanation for this being that the STF is geared to judge
constitutional law rather than criminal law. The first stage, between
investigation and indictment, can take years. The processes of penal cases,
which normally move between PF and the MPF, are slowed down in the
STF because auxiliary judges are required to check constitutional legality of
many steps in the processes. Since the start of the Lava-Jato case, most of
the STF investigations targeting politicians with the foro have lasted 2.5
years. The first probes were authorized in March 2015 (47 politicians were
included), and most of the indictments were presented during the final days
of Rodrigo Janot’s mandate as PGR, in September 2017. The first
indictments (against Cunha and Collor, among others) were presented in
August 2015, and while Cunha’s case went to trial in March 2016, Collor’s
case went to trial in August 2017. This second stage, between indictment by
the PGR and the charges being accepted or dismissed in the STF, has only
been sped up in one case apart from Cunha’s: The PT president, senator
Gleisi Hoffmann, was denounced in May 2016 and was charged before the
STF after five months.
If the main body of indictments and trials follow this pattern, the cases
against PT, PMDB, and PP leaders (denounced in 2016 and 2017) will most
likely result in charges in 2018, however their cases will most likely not be
tried before the STF in 2019; that is, after the general elections. Thus, the
foro privilegiado and the process delays are partly to blame for refilling the
Congress with politicians accused (albeit not yet convicted) of corruption.
Verdicts are handed down with even less celerity: If the process of
investigation and pre-process (from indictment to trial) under the auspices
of STF seems slow, the trial processes appear comparatively glacial. On
average, the penal processes last five and a half years in the STF (Falcão et
al. 2014: 57), often because one judge asks for more time to review the
cases, and these review delays have indefinite durations. After verdicts, the
appeal processes last, on average, 2.2 years (ibid.: 81).
In February 2016 however, the STF ruled that prison terms should be
served immediately after a sentence has been confirmed by an appellate
court. This change in the principle of assumption of innocence, producing
the so-called princípio de prisão em segunda instância, represented an
innovation of jurisprudence, since in most non-lethal cases (such as
ordinary political corruption), the assumption of innocence was preserved
all the way through multiple appeal trials and eventual changes of
jurisdictional competency. Until then, the established jurisprudence gave
politicians the option of appealing numerous times while maintaining their
freedom. The result, in nearly all cases of grand political corruption, has
been massive delays in sanctions and a sense of impunity (Arantes 2011,
Macaulay 2011: 229, Taylor 2011: 170ff.). Furthermore, by switching to
other public offices (such as state government or state assemblies, or to
municipal politics), cases could be virtually restarted due to their foro
privilegiado rights changing (ibid.: 182).
There have been several notable exceptions to this general sense of
impunity, which have been important in maintaining the legitimacy of the
STF. The watershed case according to Brazilian media was the Mensalão
case (described in the introductory chapter). This case demonstrated that
political corruption would not always be treated with impunity – although
this milestone verdict required that the STF introduced and completely
reinterpreted the German legal principle of Herrschaftsprinzip as basis of
the verdict for the defendants from the political system (Greco and Leite
2014). Despite the latest changes to legal interpretation that might mitigate
the infinite delays between sentence and prison terms, debates on STF’s
interventions into politics continue to rage in the media, as well as in the
court itself.
Amid the climate of political crisis surrounding the Lava-Jato scandal,
the internal accountability mechanisms of Congress have stopped working
(as described below). The STF is therefore the only real obstacle to having
corrupt politicians in public office. The timing of trials relative to the fixed
dates of the parliamentary elections is at the center of this problem, because
a final verdict in the STF will render those politicians sentenced ineligible
for office as a consequence to the Ficha Limpa law. Furthermore, PT has
criticized a perceived inequality in the deliberation time required to reach
these verdicts, depending on the political party of the defendant. As seen in
the examples from the Senate above, the PT leader Gleisi Hoffmann seems
to have a case moving more rapidly in the STF than other indicted senators,
making hers the only Senate candidacy likely to be barred in 2018. This
critique of the priorities and processing speeds of the STF parallels the
discussion on Lula’s eventual 2018 candidacy (described below). In cases
that drag on without a verdict, politicians can run for office and continue to
exercise power from within the legislative body, and the STF has been
extremely reluctant to remove politicians in office, despite the potential
abuse of political power for hampering investigations (with two important
exceptions described later in this chapter).
 In the period 2015–2016, the majority of the STF judges, with notable
and controversial exceptions, have taken some care to present the court as
actively working toward ending the law of impunity and protecting the
Lava-Jato investigations. Whether that representation, in the end, will turn
out to be justified, remains to be seen. In the final weeks of 2017, many
indictments against politicians were dismissed by the STF’s Segunda
Turma, and judge Gilmar Mendes liberated a number of individuals
preventively arrested in the case. So far, no politician in office has been
tried and convicted before the STF in the Lava-Jato case, and only a few of
the scores of the investigated politicians have been charged before the court
at the time of writing.4 The politicians without foro privilegiado, like the
businessmen involved in the Lava-Jato case, have been sentenced in the
common court of Curitiba, to which I now turn.

The Federal Regional Courts and the 13th Vara in Curitiba

Criminal cases are normally brought before common courts (or the first
judicial instance, primeira instância). The common courts are divided into
topical and geographical jurisdictions, or varas. Different judges thus
specialize in different areas of law. Specializing in crimes of corruption,
embezzlement, and money-laundering, judge Sérgio Moro presides one of
these varas in Curitiba. Moro, a federal judge since the mid-1990s,
previously served as auxiliary-judge in the STF during the Mensalão case
and judged the Banestado case. This case investigated tax evasion and
massive financial movements from Brazilian corporations via a bank in
Paraná to overseas accounts, amounting to at least US$20 billion and
possibly as much as US$124 billion. Like later cases (such as the Mensalão
and Lava-Jato cases), the Banestado case involved elite businessmen,
politicians, and even Henrique Meirelles, then-director of the Central Bank,
currently the minister of Finances. The Banestado case was ultimately
unsuccessful in producing accountability for the political elite, as both the
trials and the parliamentary investigative committee sparked by the scandal
ended mostly in impunity (Cardoso and di Fatima 2013). Ninety-seven
petty criminals and money launderers were sentenced in these trials
conducted by Moro (Netto 2016: 30).
Moro continued to specialize in money laundering and corruption cases
and became the judge of all Lava-Jato inquiries by virtue of the principle of
the “natural judge.” This principle fixes jurisdiction of derived criminal
cases to the judge presiding the originating case. Since the first money
launderer caught in the Lava-Jato investigation was the doleiro Alberto
Youssef, working out of Paraná, cases deriving from Youssef’s corrupt
activities would automatically pertain to the jurisdiction of Moro. Likewise,
all appeals in the Lava-Jato case are sent to the regional appeals court of
Porto Alegre. Appeals courts are the second level of the court system (pt.
segunda instância), and these courts are organized at the federal level, but
are divided into five regions, approximating the geographical regions of
Brazil, yielding a center/northern court located in Brasília, one in Rio de
Janeiro, one in São Paulo which also covers the state of Mato Grosso do
Sul, a court in Recife covering the northeastern states, and the southern
regional court (abbreviated TRF-4), located in Porto Alegre, covering the
states of Paraná, Rio Grande do Sul, and Santa Catarina.
Unlike many other Brazilian courts, the appeal court (TRF-4) linked to
the Lava-Jato cases ratify Moro’s sentences relatively quickly, yielding
average trial durations of eight months in the 13th Vara, plus 13,5 months in
the TRF-4 (Campos 2017). This is rapid for Brazilian courts in politicized
cases, and since TRF-4 rarely overturns Moro’s sentences, instead
preferring to augment them (ibid.), the Lava-Jato trials have been hailed as
an example to follow for other courts.
The much-publicized results of the Lava-Jato case have also propelled
Sérgio Moro to the front of the public imaginary surrounding the corruption
probe. Moro has become the national poster boy of the combat against
corruption, featuring on the front page of numerous editions of the popular
weekly magazines IstoÉ, Época, and Veja since 2014. Nationally, Moro has
received various awards and been the topic of several books, and
internationally, he has been named one of the 100 most influential people in
the world by Time Magazine. Without having a political profile or ever
expressing an interest in running for office, the name Sérgio Moro appears
regularly alongside political actors in popularity surveys conducted by the
large Brazilian poll institutes (e.g., Datafolha 2017). Moro became an icon
– literally – in the protests of 2015 and 2016: Inflatable dolls of the judge,
dressed up as a Brazilian Superman, and cardboard masks and banners
proclaiming “we are all Sérgio Moro” were commonplace artefacts during
the street protests. Standing in for the Lava-Jato investigations, the task-
force, or the “combat against corruption,” Moro’s iconicity may also have
glossed over essential differences between the branches of government and
motivated other judges to attempt a more assertive style of political
intervention.
The public support (whether truly “popular” or not) for the judiciary
was, in a way, an ambition that Moro himself formulated explicitly in a
paper published in 2008, titled “Considerations about the operation Clean
Hands.” This paper deals with the successes of the Italian combat against
corruption during the 1990s scandal known as Tangentopolis, although the
paper also warns of a paradoxical outcome in that combat: Silvio
Berlusconi ascended to power in the aftermath of the successful operation
and collapse of the existing political system in Italy. In the paper, Moro
stated that public attention to corruption cases is necessary to get to the root
of the problem:

Perhaps the most important lesson of the whole episode is that legal
action against corruption is only effective with the support of
democracy. This defines the limits and the possibilities of legal action.
With the support of public opinion, the conditions for progress and
good results exist. Without it, trials can hardly succeed. Also, to get a
favorable public opinion, trials must also reach good results … It is the
enlightened public opinion that can, by its own institutional means,
attack the structural causes of corruption. Furthermore, the judicial
punishment of corruption is always difficult … In this perspective,
public opinion can be a healthy substitute, being able to impose some
punishment of corrupt public officials, by condemning them to
ostracism.
(Moro 2004: 61, translation mine)

The quote, and Moro himself, became the target of much critique,
essentially because this Western-centric ideal of accountability through
public opinion is not necessarily true in Brazil. Critics emphasized that the
spectacle of public opinion, especially knowing the vested or explicit
interests of Brazilian media, was not conducive to the Rule of Law (e.g.,
Guimarães 2016). During the Lava-Jato trials, Moro has largely avoided the
media, only sporadically allowing journalists direct quotes. This contrasts to
the practice of the Lava-Jato task-force, from which several prosecutors
frequently appear in the media (e.g., Dallagnol and Martelo 2016) and as
speakers at public events.

The Supreme Electoral Court

As with the legal areas of labor rights and military law, the area of elections
is ultimately governed and overseen by a specific court – the Supreme
Electoral Court (Tribunal Superior Eleitoral, henceforth TSE). The tasks of
the TSE include overseeing elections and candidatures, regulating and
auditing campaign expenditure during elections, processing and publicizing
election results, and other tasks of an administrative and regulative nature
related to elections. In the 1990s, the court was responsible for
implementing electronic voting machines across the country in an effort to
put an end to accusations of vote fraud. The TSE also, in cooperation and
sometimes conflict with the STF (Marchetti 2008, Taylor 2011: 177),
determines certain questions of political nature, including rules for party
representation thresholds (and the minimum number of votes parties must
obtain to obtain seats), the vertical integration of coalitions across state and
federal levels (the rules for coligações, i.e., electoral alliances), and party
loyalty in Congress.
Historically, the TSE has been known as an institution with a high
degree of legitimacy, which gave it special status during the transition to
democracy in the 1980s (ibid.: 166). Since then, the TSE has been relatively
efficient (compared to the STF) in producing accountability in high-level
cases, removing several governors from their mandates (in Tocantins,
Maranhão, Roraima, Paraíba, Piauí, and, in 2017, Amazonas). In February
2017, the governor of Rio de Janeiro, Luiz Fernando Pezão, lost his
mandate in a verdict by the Regional Electoral Court of Rio de Janeiro, but
the process has been appealed to TSE and has since stalled. This is one of
the most prominent contemporary cases of the lower-level federal electoral
courts producing accountability. Each state has a permanent lower-level
court (abbreviated TRE) overseeing state elections, and at the municipal
level, the temporary accountability institution, which is set up for each
election, is called electoral juntas (Marchetti 2008: 40).
Overall, the electoral court system has limited resources and too few
personnel to ensure fair elections for all citizens across such an extensive
territory. Furthermore, a two-year renewal of the mandate of TSE judge was
once problematic for producing consistent jurisprudence in the court
(Taylor 2011: 165ff.). This timeframe of two-year terms has resulted in a
constant reshuffling of the court’s composition. The court’s bench is
designed to contain a mixture of judges and respected lawyers: Two slots
are filled by lawyers of public renown, each chosen by the president on the
basis of a list of three names nominated by the STF, which has selected
these names from a list of six names nominated by the national bar
association (OAB). The remaining five seats are allocated to three STF
judges and two STJ judges5 (and each of these judges must be nominated
for a slot by consensus in their respective courts of origin).
Members are permitted to sit for up to two terms of two years. The STJ
judges normally sit on the TSE bench for a single term, while two terms
have become the norm for the other members (Marchetti 2008: 885). This
principle of rotation proved to be of the utmost importance during the trial
aiming at the cancellation of the presidential elections of 2014, described
below.
Since the TSE is presided over by one of the STF judges, the two courts
normally act in a concerted way. Some of the high-profiled rulings of TSE
have been challenged on constitutional grounds and thus become the topic
of deliberation in the STF, but the TSE rulings have generally been ratified
there (Taylor 2011: 170). In effect, the STF has delegated competency of
legal interpretation in the electoral area to TSE and avoids direct
interference, although the overlap of STF and TSE judges aligns the courts
somewhat. The STF ratifications grant a shroud of legitimacy to the audits
of party and candidate campaign spending, although it is well known that
the auditors of TSE are unable to scratch the surface of the vast budgets
commanded by the Brazilian parties (ibid.). Special audit procedures can be
initiated by the TSE when whistle-blowers emerge, or, as we shall see,
when political actors eye opportunities to denounce their opponents.

The Tribunal for the Accounts of the Union

A number of governmental audit bodies exist in parallel in Brazil. The
executive branch appoints a comptroller-general (Controladoria Geral da
União), the Ministério Público has offices called Ministério Público de
Contas at both federal and state levels, while the audit organization of the
legislative branch of government is the Tribunal das Contas da União
(henceforth TCU). In this auditing body, routine accounting, state budget
auditing, and special case-by-case investigations are conducted. The TCU
reports on budget irregularities in large-scale government projects and
produces a yearly independent audit of the state finances. TCU normally
only acts as advisor to Congress, but can also (since 2001, when the state
budget regulation was revised) request that the parliamentary budget
committee block fund transfers when severe irregularities are encountered
in projects.
The TCU court is headed by nine so-called ministers, titled like the
Supreme Court judges, despite the fact that TCU judges are not part of the
judiciary, nor necessarily trained in law. Six of the TCU judges are
nominated by the Congress, and three are nominated by the president. Two
of the latter must be career public servants with a background in the TCU.
The ministers are appointed for life but must retire at the age of 70. This
means that “politicians are frequently nominated to join the TCU at the
height or the end of their political career, in what is sometimes seen as a
kind of early retirement” (Speck 2011: 155). The leadership of this
accounting body thus, to a certain extent, reflects the composition and
leading coalition of the Congress. The nomination for a seat in the TCU
“comes with strings attached, since the government with its majority in
Congress does not expect any trouble from its former allies once they are
transferred to the TCU” (ibid.). TCU auditors are, for that reason, rarely
called to testify about audit findings in the committees and hearings of
Congress (ibid.: 147).
Despite not being part of the judiciary branch, the TCU has some powers
to sanction and issue fines to government actors found guilty of
misconduct. In practice, the implementation of sanctions and fines is often
delayed, because the TCU verdicts can and normally will be questioned in
the state and federal courts. This fact, according to Speck, is the weakest
link of the accountability mechanisms of the TCU (ibid.: 144).
The political network and dependency of TCU ministers is probably an
even weaker link, however. During 2015–2016, three TCU ministers
underwent investigation during the Lava-Jato scandal. According to various
testimonies of the case, the ministers Aroldo Cedraz, Vital do Rego, and
Raimundo Carreira have received bribes, possibly to entice them to actively
obstruct audits in the TCU, or as kickbacks related to when they were still
in political office. Moreover, Cedraz’s son, the lawyer Tiago Cedraz, was
subpoenaed by the police in 2017 on the suspicion that he had negotiated
contracts between Petrobras and an American company and received
US$20 million in kickbacks. In one of the plea bargain testimonies of the
Lava-Jato case, the expelled PP ex-parliamentarian Pedro Corrêa claimed
that Cedraz was actually selected for the TCU as a part of a political bargain
struck in the wake of the Mensalão scandal. Cedraz was appointed as
payment for votes to save the mandate of the PP leader José Janene,
according to the testimony. Moreover, the minister Augusto Nardes is
investigated in the operation named Zelotes.
The TCU, in conclusion, is perhaps the most politically aligned
accountability institution, at least insofar as concerns the leaders of the
organization. It is also the body least likely to react to large-scale budget
problems at the federal level: Only in 1937 did TCU refuse to accept the
state accounts (in a situation where then-President Vargas had threatened
the independence of the governmental branches with a new constitution).
The string of 77 consecutive approvals of the state accounts came to an end
in 2015, however, when a TCU report on fiscal delays in the State Treasury
was interpreted as a grave infraction of the Law of Fiscal Responsibility by
Augusto Nardes, conducting the yearly audit of state accounts. His
recommendation of rejection of the state finances was unanimously
supported in TCU, and the rejection precipitated the impeachment
proceedings, described in more detail in the final chapter. This outcome was
the most obvious example of the ways in which accountability mechanisms
and legal actions were deployed in the wake of the Lava-Jato scandal –
although the TCU report on fiscal delays had nothing to do with the
Petrobras corruption.

Judicial Consequences of Corruption Cases

The following sections consist in five case studies drawn from the vast
range of judicial processes that have interfered in the struggles of the
Brazilian political arena. Two of these cases emerged outside of the
criminal actions of the Lava-Jato probe but had their roots in this
investigation of corruption. The case study of the impeachment process is
saved for the final chapter, and so the empirical cases analyzed here are:

• The TSE process for cancellation of the 2014 presidential elections

• Eduardo Cunha’s disciplinary process in the parliamentary Conselho
de Ética
• Two cases of suspension of senatorial mandates (Senators Calheiros
and Neves)
• The nine legal actions against ex-President Lula

From these cases, I have abstracted a range of direct and indirect

outcomes specific to the Brazilian system of accountability institutions and
checks and balances in their current configuration. These outcomes include
overturning and rigging elections, removing politicians from public office,
changing the valorization of political capital and the informal rules for
coalition formation. The outcomes are discussed as a conclusion to the
following brief case studies. Each of the cases had a great impact on the
media agenda. The cases of Cunha and Lula are centerpieces of the
informational cascade described in the next chapter, and for the sake of
clarity the quantitative media content analysis concerning these cases is also
located there. Like these cases, the TSE process and the cases pertaining to
Senators’ mandates were dominant elements of national news for many
weeks, but these cases lie outside of the sample period of the quantitative
study.

The TSE Process for Cancellation of the 2014 Presidential Elections

The case with the most explosive potential of all started with a petition for
cancelling the result of the 2014 presidential elections. The case was
initially comprised of four different petitions, due to the losing opposition
party, PSDB, having filed a range of different actions in TSE. The
following overview of these cases is based on the comprehensive log of
legal documents available at the TSE website (Tribunal Superior Eleitoral
2017).

• On the 2nd of October 2014, PSDB and the coalition behind the
candidacy of Aécio Neves filed a petition for investigation
(investigação judicial eleitoral) into the incumbent president’s abuse
of public mechanisms, such as illegal distribution of pamphlets via the
state postal service and abuse of ministerial positions for campaigning
(AIJE 154781).
• On the 30th of October 2014, PSDB announced it would file a petition
for impugnation (ação de impugnação de mandato eletivo), calling for
an audit of the votes cast during the elections (AIME 761). One
argument for this court action was the use of funds obtained from
corruption in Petrobras during the campaign, based on testimonies
made by Paulo Roberto da Costa, Ricardo Pessoa, Alberto Youssef,
and Pedro Barusco. It was submitted to the TSE on January 2.
• On the 18th of December 2014, PSDB filed another petition for
investigation (investigação judicial eleitoral) into the president’s use
of public ceremonies and events to perform political, and thus non-
public, campaigning activities (AIJE 194358).
• In October 2014, PSDB also announced it would file a representação
(RP 846), technically another kind of petition, also claiming that the
state postal service (Correios) had obstructed Neves’ election mail
campaign. It was submitted to the TSE on January 2. Like the AIJE
154781, this action also targeted the postal service director Walter
Pinheiro.

The allegation that kickbacks from the corruption in Petrobras helped

Rousseff fund her campaign was included in both the second and third
petitions, however it was in the second petition that the denunciation of
corruption was highlighted. Both petitions referenced the media leaks from
the Lava-Jato case. In the following, I will describe the interactions between
the investigative bodies, judicial actors and the political system. Based on
this, I will venture an answer to the question of how a large amount of
evidence of corruption could end up being completely disregarded, focusing
on two types of interaction and the key role played by the TSE president
(and STF judge) Gilmar Mendes.
The eventual consequence of the actions in the electoral court could have
been the removal of the winning president and vice-president from office,
leaving Brazil with a third president in the space of one year. Jurisprudence
as to whether this would result in new elections or the inauguration of the
losing presidential candidate was not clear, however, as different Brazilian
courts had solved similar situations differently in the past. At the time of the
initial petitions, commentators assumed that the cancellation of the 2014
election result would have positioned PSDB favorably. This assumption
was due to the increasing negative focus and coverage of the PT and the
Petrobras scandal, but the timing of the electoral court actions eventually
put PSDB in an unexpectedly vulnerable position. The process was
characterized by many delays, fits and starts.
Initially, the petition AIME 761 was rejected by TSE Justice Maria
Thereza de Assis Moura in February 2015. The rejection was appealed in
March by PSDB and then sent to the TSE plenary, where judge Gilmar
Mendes delayed the process through an extension of the deadline for
review. In August 2015, the ministro relator João Otávio Noronha of the
petition AIJE 154781 asked the STF to share material from the Petrobras
case, and synchronously the judge Luis Fux suggested combining all PSDB
petitions while Gilmar Mendes and two other judges voted to continue the
investigations. Fux’s idea of combining the petitions was debated in the
plenary, but would only be realized the following year. The AIME 761 was
reopened by then-president of TSE, Dias Toffoli, on October 6, 2015. On
December 17, the representation RP 846 was rejected, but PSDB appealed
with the argument that new evidence had come to light. That argument
came to be the Achilles’ heel of the case 18 months later.
On top of the four petitions for cancellation of the election of Rousseff,
PSDB also petitioned the prosecutor-general of elections (PGE) to discard
PT from further elections, as a consequence of the denunciation made by
the ex-director of Petrobras, Nestor Cerveró. In a testimony, Cerveró
claimed that PT received laundered money from Angolan funds through the
actions of the ex-President Lula, in violation of the law prohibiting foreign
influence upon Brazilian elections. This separate petition has lingered in the
TSE docket since.
In February 2016, the four initial PSDB petitions were integrated into
one aggregate suit, maintaining the file name AIJE 194358 under the
auspices of the auditor-general of TSE (a position rotating biennially
between the TSE judges). At that point, the cases were to be judged by
Maria Thereza de Assis Moura, and she restarted the process of collecting
evidence of fraud in the campaign, despite the fact that one year earlier, she
had recommended the case be rejected by the TSE. Various appeals from
PT were then denied in the TSE plenary, ruling that the excessive use of
formal appeals (agravos) was slowing the process down. Events overtook
the TSE process in any case, as Rousseff was impeached, which led to a
new situation for the defense lawyers of Rousseff and Temer. Formerly
united as defendants, the newly inaugurated President Michel Temer now
tried to advance the argument in TSE that the campaign accounts of
Rousseff and Temer were distinct in principle and practice, and that the
eventual annulment of Rousseff’s bid wouldn’t apply to his mandate, as he
was without blame. Several cross-payments from the PT campaign to
PMDB weakened this argument, however.
In August 2016, TSE audits of printing companies contracted by PT
demonstrated that three of these companies – Focal, Red Seg, and Grafica
VPTB – could not prove the capability to produce large amounts of printed
material, nor that they in fact had produced all of the material paid for
during the 2014 campaign. This was taken to be an indication that the
companies were in fact used to produce fraudulent contracts. It was not
clear then who benefitted from this money laundering, since the money
was, in fact, spent by the PT. The mismatch of contracts and products led
the electoral prosecutor, Nicolao Dino, to believe that campaign funds had
financed other areas than the printed material, since the occultation of
transfers could indicate the presence of illegal financing. The printing
company audits became the cornerstone of the inquiry led by the new TSE
Auditor-General Herman Benjamin. Taking over from judge Moura who
rotated out of TSE on September 2, Herman called a number of individuals
imprisoned or sentenced by the Lava-Jato investigations to testify to the
TSE, including the doleiro Alberto Youssef, the ex-senator Delcídio do
Amaral and a number of construction company CEOs.
On October 13, 2016, judge Benjamin called for an investigation of the
printing companies’ bank accounts, and in December, police agents entered
the premises of the printing companies on Benjamin’s request. Further
testimonies were solicited, and Benjamin set up a task-force dealing with
the investigation. Furthermore, Benjamin’s case now included evidence
from the Lava-Jato investigations phase dubbed Acarajé; evidence that
linked the Odebrecht group to certain offshore companies and the marketing
expert of PT, João Santana. Santana had admitted to being paid by
Odebrecht for services delivered to PT as kickback for the advantages the
construction company received in federal procurement and public tenders.
Seventy-seven Odebrecht executives had agreed to a plea bargains in the
Lava-Jato case, and key testimonies of this process were included in the
TSE process. The top executives of the company’s so-called “department of
bribes” appeared before the court, along with the incarcerated heir to the
Odebrecht dynasty, Marcelo Odebrecht, in February and early March 2017.
The inclusion of these testimonies was both a blessing and a curse for
Michel Temer because they posed a greater risk of a cancelled election. On
the other hand, each court session with new testimonies6 moved the court
closer to the rotation deadline of two judges. This was, in a nutshell, the
most important interaction between the political and the judicial system:
Extending the testimony process gave Temer an opportunity to pack the
bench of the TSE.
This type of interaction has been studied in other Latin American
contexts of presidential and democratic crises, and based on comparative
studies of such cases, Helmke concludes that “judges are clearly capable of
altering the various parameters that affect presidential instability” (Helmke
2017: 128). Court-packing is one strategy a president may pursue to secure
stability, and this was exactly what swayed the vote, in the end, in Temer’s
favor. As TSE judges Henrique Neves and Luciana Lóssio left their
positions (in April and May, respectively), Temer could appoint two new
judges to the bench, and he called in Admar Gonzaga and Tarcísio Vieira,
both of them shortlisted by judges of the STF.
The two new judges on the bench meant that a new majority formed
among the judges. The judge overseeing the process, Herman Benjamin,
had declared in the press that the court would provide a historical ruling,
and it was believed that two out of the three STF judges serving on the TSE
bench (Luis Fux and Rosa Weber) would vote with Benjamin for cancelling
the 2014 presidential elections. This would, indeed, have been a historical
moment, since no presidential election had ever been overturned by the
TSE. However, as the process of hearing the testimonies extended over
March and April, Judge Luciana Lóssio’s mandate ended. Lóssio was
expected to vote for cancelling the election (e.g., Mattoso et al. 2017). With
Lóssio now gone, what had previously been a majority of TSE judges
inclining toward cancellation, became a minority on the bench.
The final days of the trial in the TSE plenary, on June 7, 8, and 9, were
dominated by discussion on the original scope of the petitions, and whether
the trial could expand this scope to any and all corruption relating to parties
of the president and the vice-president. The inclusion of testimonies of the
expelled senator Delcídio do Amaral, the Odebrecht executives, and the
marketing expert Santana, did not strongly support the initial accusations
(that had referenced leaked plea bargains from Paulo Roberto Costa and
Alberto Youssef), as these testimonies dealt with other situations of graft
and corruption. The report of the electoral prosecutor affirmed the narrative
of Petrobras contracts leading to kickbacks to the government parties, but
not directly to the election campaign.
The judge in charge of the process, Benjamin, initiated the procedure of
the verdict, unsurprisingly recommending that the election of Rousseff and
Temer should be annulled (Benjamin 2017). The first judge to vote after
Benjamin was Napoleão Nunes, who insisted that the trial could not take
into consideration elements that were not present in the original petition –
such as the Odebrecht testimonies. Nunes therefore recommended denying
the petitions. Judge Luis Fux in turn contradicted Nunes, as Fux affirmed
that rejecting the use of Odebrecht testimonies would mean ignoring
evidence. TSE president Gilmar Mendes scorned this stance during the
debate, calling it “moralist.”
The two judges appointed by Temer in 2017 supported Judge Nunes,
while the two judges from STF (Luis Fux and Rosa Weber) followed the
recommendation of Benjamin. The dissent in the court meant that Judge
Mendes had the decisive vote, and he ultimately denied the petition, stating
that the instability of the country was in question, and that presidents should
not be removed every other day. Both Rousseff and Temer thus ended up
acquitted in the final verdict. PSDB’s lawyers had, in their final allegation
prior to the verdict, highlighted that only Dilma Rousseff and the PT was to
blame for illicit campaign funds, but with the tie-breaking vote of Mendes,
the Rousseff-Temer ticket as a whole was acquitted as the majority of TSE
voted not to separate the candidatures.
Another type of political-judicial interaction of interest here is the
relation between the plaintiff’s petition and the scope of investigation
possible under the current legislation concerning the TSE. While PSDB
had, from the very beginning, accused PT of abuse of public office and
illegal funds for campaigning, throughout the two-year process TSE
maintained that an eventual cancellation would remove both president and
vice-president. However, the ultimate decision to restrict the scope of the
verdict to the original allegation ran contrary to the process of collecting
evidence for the court. In October 2016, the court had gathered a task-force
from PF, tax authorities, and the Central Bank to determine the character of
transfers to the companies printing campaign pamphlets for Rousseff and
Temer in 2014. Soon after, the PGR had recommended including
testimonies from the Odebrecht investigations, and the TSE obliged. In
spite of this, the final verdict, the evidence produced had “expanded the
object of the case” and had to be discarded.
The judicial decision to include and exclude material, in the end, made
all the difference to whether Brazilian voters had the opportunity to
democratically elect a new president. Instead, the country was stuck with a
president tarnished by the very corruption scandal that had undermined the
previous president. Meanwhile, both PSDB and the coalition that had
originally backed Aécio Neves obtained several cabinet positions and no
longer had any interest in the legal action they had started, nor any
consensual candidate beyond Temer to put forth for the presidency.
In his final statement, Gilmar Mendes made ironic comments concerning
the inclusion of any and all allegations floating in the press, and how this
would delay the trial indefinitely. That observation in itself shows that
Mendes and other judges were acutely aware that the timing of the trial in
relation to Lava-Jato was of the utmost importance. Expecting (or perhaps
knowing) more evidence of corruption and illicit campaign funding could
turn up, Gilmar Mendes intervened in order to shut down the TSE trial and
the threat of instability to Temer’s presidency.
In sum, court-packing, the deliberate timing of the judicial action, and
the reduction of scope in the investigations prevented TSE from producing
accountability, despite the evidence collected by the court’s task-force.
Benjamin, commenting upon his colleagues’ stances, stated that TSE
seemed to invert the historical goal of the court by seeking only to
adjudicate matters concerning official donations, denying the validity of
evidence of unofficial transfers. The historical sentence that Benjamin
foresaw early in 2017 ended up being historical only in the sense that it
reflected the sense of impunity toward political corruption. Likewise, two
parallel investigations that were opened by Judge Moura before leaving the
TSE also stalled. These cases investigated the suspicions of PMDB and PP
for campaign fraud in 2014. The electoral court case illustrates how
accountability institutions, despite the best interests of many in the
organization, may still be subjugated to the goals of other institutions,
defeating the whole idea of checks and balances.

Eduardo Cunha’s Disciplinary Process in the Parliamentary Conselho de

Ética
One accountability institution not mentioned in the overview above is the
internal disciplinary committees of the legislative branch of government:
The Conselhos da Ética in the Senate and the Câmara. In this section, I will
discuss the intricacies of this accountability mechanism, exemplified
through the case in which Eduardo Cunha’s mandate in the Congress was
revoked. With this example, I will show how accountability institutions and
processes can also turn out to be bargaining chips in complex political
power struggles. I will highlight the entanglement of the political and the
judicial fields in this legislative body, which is quite different from the
investigations and court trials described previously, and is also perhaps
much more crucial to the functioning of the political landscape in Brazil.
The reason why the two Conselhos da Ética of Congress were left out in
the above discussion of accountability institutions is both conceptual and
practical: Conceptually, these disciplinary committees do not pertain to the
Judiciary, but are governed by the legislative bodies. Moreover, in practice
they have only been used recently in minor internal disciplinary cases and
not for producing accountability to the electorate. Though these disciplinary
committees in principle treat all kinds of lapses of ethical comportment and
hold the power to recommend the cancellation of mandates, they have only
very rarely been activated in the corruption cases engulfing the Brazilian
Legislative in the last three years.
One reason for this surprising state of affairs is the number of
investigations, denunciations, and trials hovering over the legislative
assemblies in the STF. With roughly half of parliament accused, and to a
large extent investigated, there is little incentive for the parliamentarians to
set up internal institutional investigations. Not only would accompanying
disciplinary processes in every case of denunciation take up much time in
parliament, it would certainly also sour relations between the many
fragmented parties in Congress. In a certain sense, both Conselhos have
effectively limited their reach to events occurring inside Congress and
between congressmen, rather than taking a more inclusive view of the
moral code of public office holders.
The Conselhos of Congress are complemented by another internal
accountability mechanism: For more than a decade, the legislative modus
operandi in corruption cases has been to set up parliamentary investigating
committees (called CPIs) looking into political corruption – but seldom
with great consequences for the accused (Power and Taylor 2011: 19). The
prerogative to install and appoint a CPI lies with the president of the Senate
or the Câmara, respectively (and joint commissions between the two
Houses of Congress can also be constructed from time to time). For obvious
reasons, the presidents of the Houses can easily avoid being investigated in
the context of a CPI. So, despite the fact that the PMDB leader in the
Câmara, Eduardo Cunha, was connected to the Petrobras graft in leaks in
December 2014, he could still construct a CPI looking into the case when
he was elected president of the Câmara in early 2015.
As mentioned in the previous chapter, Cunha went on the stand on
March 12, 2015, calmly denying the accusations in the investigative
commission he himself had opened to examine the Petrobras corruption.
Furthermore, Cunha claimed that he did not have any overseas bank
accounts. This claim became the basis of a process for annulling Cunha’s
mandate. It was a complex and excruciating process, and the only
disciplinary process in the Conselho da Ética in the 2015–2017 period
targeting a politician for being involved in the Lava-Jato case. The senators
have, meanwhile, simply avoided to activate the Senate’s Conselho,
preferring to keep up the appearances of a harmonious Upper House.
When evidence of Swiss bank accounts, in Cunha’s name, was leaked
soon after the first indictment of the PGR targeting Cunha (in August
2015), the minor left-wing party PSOL filed for a disciplinary action in the
Câmara. PSOL argued that since Cunha had specifically denied, in the CPI
regarding Petrobras, having overseas accounts, he had lied to his peers.
Lying in a CPI to his peers constituted a clear breach of decorum, enough to
revoke the mandate of a congressman (in theory, but not in practice, a
corruption scandal should also suffice). On October 1, 2015, PSOL and the
party Rede Sustentabilidade requested documentation from the PGR in
September, and on October 13, the two parties and half of the PT
congressmen filed a petition urging the Conselho da Ética to open a case
against Cunha. As president of the Câmara, Cunha maintained the rights to
manage and to command the Board of the Câmara and the institutional
resources before the process began, and even as the process got underway –
despite being obviously partial in the resulting disciplinary process.
Cunha exploited various procedural loopholes and thus postponed the
disciplinary process during October. First of all, the Board of the Câmara,
which is formally responsible for all document handling and administrative
tasks of the Lower House, simply waited the maximum number of days to
hand over the process to the Conselho. This meant that the president of the
Conselho, José Carlos Araújo, could not formally open the case, given the
rules of the Câmara, and only on November 4 did he get to appoint a
speaker (relator) to conduct the disciplinary process. During November, the
speaker Fausto Pinato (then of the party PRB) tried to present his
preliminary recommendation for opening a process. However, half of the 20
congressmen constituting the Conselho, known in the daily press as
Cunha’s “shock troop,” repeatedly delayed the preliminary recommendation
by extending discussions of formalities. The tactic of delaying (in the vein
of the US tradition known as a filibuster) worked successfully throughout
November. Furthermore, as the sub-committees of the Congress are
prohibited from assembling in the time slots when the plenary assembles,
Cunha could schedule the Câmara’s activities, and this allowed him to
simply call for a plenary vote in the Câmara whenever the Conselho was
about to vote, automatically ending meetings of the sub-commission. With
these tactics, Pinato was hindered in even reading his recommendation until
December.
Toward the end of November, as Cunha’s repertoire of delay tactics was
slowly exhausted, it became clear that the speaker, the Council president
and a small majority in the Council would not dismiss the case. The day of
voting for opening the case was set for December 1. In the week leading up
to the session, deliberations within the PT were reported by the press. The
three PT members of the Ethics Council could, in principle, add their votes
to the Cunha’s so-called shock troops, letting him off the hook. Implicitly or
explicitly, depending on the point of view, that would start a truce between
Rousseff and Cunha, protecting the president of the Câmara from
disciplinary action and the president of the Republic from impeachment. In
effect, once the Conselho reached the point of voting for the preliminary
assessment of Pinato, the three votes of PT in the Conselho became a
bargaining chip for Cunha, since he held the prerogative to initiate
impeachment proceedings against the PT president Dilma Rousseff.
The truce, or back-stage deal, between Cunha and PT leaders, never
materialized, however. The Conselho, in the session on December 1, did not
get around to vote. The members of PT looked hesitant and deliberated in
hushed voices, while following the debate of the commission. As the
meeting dragged on, the national president of PT tweeted (and media
promptly reported) that PT had demanded the deputies to vote for opening a
case against Cunha. The meeting ended, and the vote was postponed to the
following day. Finally, under cross-pressure from nearly all of Congress and
the divided allegiance to a party line, on one hand, and the best interests of
their president on the other, the PT members of the Conselho finally
decided how to vote. One of them, Zé Geraldo, outright declared the
supposed truce negotiation to be blackmail, “a knife to the throat,” and that
he and his two colleagues in the Ethics Council were less than happy
holding the trigger for the impeachment (Almeida 2016: 106).
The following day, with the votes of the PT representatives, the
commission formed a majority in favor of Pinato’s recommendation for
opening the disciplinary process against Cunha. The meeting in the
Conselho at the moment of the vote was highly agitated, featuring
everything from verbal abuse to physical infighting. The feared retaliation
arrived the same afternoon, as Cunha declared that impeachment
proceedings would be initiated then, confirming PT’s earlier suspicions of
blackmail. These events became one of the cornerstones of the counter-
narrative of a coup d’état deployed by PT.
Shortly before Christmas, the PGR filed for the removal of Cunha from
office in the STF, with the argument that his position in Congress allowed
him to systematically abuse his prerogatives and protect him from
investigation. The same week, the soundness of that argument was amply
demonstrated: Pinato’s recommendation became overturned on formal
grounds by the vice-president of the Câmara, Waldir Maranhão (PP). He
ruled, on behalf of the Board of the Câmara, that since Pinato’s party PRB
was formally allied to the governing block, and thus to PMDB, Pinato could
not be speaker of the case. In the ruling, Maranhão stated that a politician
could not be the speaker of a disciplinary case against a colleague
participating in the same parliamentary coalition. The ruling ran counter to
logic, since Cunha had long ago given every possible signal that he wanted
a new government coalition, declared his break with government, and since
Pinato had actually recommended revoking Cunha’s mandate, thus
eliminating any suspicions of exchanging favors.
Nonetheless, the president of the Conselho was forced to appoint a new
speaker on the case, this time from the opposition which Cunha was
courting. That move gave Cunha another shot at delaying the case and
possibly getting it filed away. Cunha’s oppositional stance notwithstanding,
the new speaker, Marcos Rogério of the then-opposition party Democratas,
recommended revoking Cunha’s mandate on December 15, but his report
on the case was annulled, again by Maranhão, the vice-president of the
Câmara. The argument brought up by Cunha-allied members of the
Conselho consisted in a formal complaint that they had not been given time
to process Rogério’s recommendation. The process then returned to the
starting point.
As usual in the Brazilian legislature, all processes ground to a halt in
January and half of February. Synchronously, a window of opportunity for
switching parties was opened in the Câmara, and as many as 14 percent of
the Congressmen (71 in total) in the Câmara changed parties. This meant
that the composition of the parliamentary blocks had changed, and in
consequence, that the distribution of representation and substitute
representatives in the Conselho was re-examined and changed accordingly.
In the party-switching frenzy, Cunha had not managed to install a majority
in his favor in the Conselho, however. Rogério’s recommendation for
opening the disciplinary process was to be voted on March 3, five months
after the petition was filed.
One member of the Conselho, Vinícius Gurgel, who was planning to
vote in Cunha’s favor, could not participate in the vote, however, and his
substitute was expected to vote against Cunha. The Board of the Câmara
received a document where Gurgel resigned from his position in the
Conselho (albeit with a falsified signature), and this allowed Gurgel’s party
leader to appoint another Congressman to the position and support Cunha’s
attempt to close down the disciplinary process. The maneuver ended to no
avail with the late-night vote on March 3, when Rogerio’s recommendation
was passed, 11 votes to 10. Cunha then avoided being served by the
legislative clerks, further delaying the process for one week.
At this moment, Cunha also faced corruption charges before the STF –
as the first politician involved in the Lava-Jato scandal. That did not prevent
him from conducting the impeachment proceedings of the Câmara in April,
however. The STF only responded to the PGR’s request for removal of
Cunha from office in May 2016, based on a ruling that charged politicians
cannot be in the line of succession to the presidency (which, with the
eventual impeachment of Rousseff, would locate Cunha as the de facto
vice-president). Using this clause, the STF abstained from creating a
jurisprudence for removing politicians from office who were suspected of
using their position to meddle in their investigations. This legal
interpretation was what the PGR requested when filing for Cunha’s
removal, but the STF denied it.
With the turmoil of the advancing impeachment and the removal of
Cunha on May 5, the presidency of the Câmara was left to Waldir
Maranhão – also under investigation at that point – who had aligned with
Cunha before. Maranhão, who had twice annulled votes in Cunha’s case,
surprisingly also tried to annul the impeachment vote of the Câmara, but
after less than 24 hours of internal pressure proceeded to annul his own
annulment. The former opposition negotiated a new governing coalition,
and once the increasingly isolated Cunha renounced his presidency on July
7, party leaders decided to elect a new leader for the Câmara – one less
associated with Cunha and less easily subdued to the cross-pressures of
Congress.
Rodrigo Maia of the party Democratas was elected new president of the
Câmara, and he scheduled a plenary vote for the recommendation of the
Conselho da Ética for revoking Cunha’s mandate. This vote occurred in
September 2016, 11 months after the initial petition. The vote resulted in a
massive majority backing the revocation of Cunha’s mandate, 450 to 10,
demonstrating that he had become a political pariah at that point. With his
mandate revoked, Cunha’s seven pending corruption cases were relocated
to the regional court of Curitiba running the Lava-Jato investigations. True
to the pattern, judge Sérgio Moro quickly authorized the preventive arrest of
Cunha, apprehending him on October 19.
Despite being the first Lava-Jato investigated politician to be charged
before the STF, Cunha thus never got to have a complete trial before that
court. During the first months of the Lava-Jato probe, in 2014, the PT
member André Vargas was the only other congressman to have his mandate
revoked by the Câmara. In other words, aside from Vargas and Cunha, not
even the politicians who were charged before the STF were made
accountable to their peers through the mechanism of the Conselho, nor did
the parties move to expel them based on corruption charges.
The case of Cunha’s removal was emblematic of the problems inherent
to Congress’ internal accountability processes when Congress is faced with
multiple corruption allegations. At each turn of the process in Conselho, the
votes for the various reports were almost a tie, which indicates that this
body of representatives will bow to the political pressures of their peers to
the detriment of the voters. Furthermore, the quasi-legal framework used in
disciplinary processes turned out to be so fragile that the process could very
well have ended up acquitting Cunha. The tactics used to delay Cunha’s
removal from office were legally valid, although (literally) violently
contested, which shows the democratic deficit of disciplinary procedures in
the Câmara – procedures that have not been challenged although they are
clearly malleable. His removal by the STF, however, provided some legal
jurisprudence to the following cases, where the STF also intervened in the
legislative branch of government.

The Case of Renan Calheiros and the Constitutional Mini-Crisis

A parallel to Eduardo Cunha’s case removed Senate President Renan
Calheiros, albeit briefly, from office. The verdict, in the case of Cunha, was
that no parliamentarian charged before the STF could preside over any
public office that would place him or her in the presidential line of
succession. Traditionally, presidents are replaced by vice-presidents, and
when both are absent, the vice-president of the Câmara acts as head of state.
The next in line is the president of the Senate and the president of the STF.
The STF, on November 3, 2016, initiated a plenary discussion of whether or
not the Cunha case should turn into legal precedence, but postponed the
decision indefinitely. At the point of postponement, the majority of judges
in the STF plenary had already declared their votes, which meant that the
court had a virtual, but not enacted, ruling on the matter.
Like Cunha, Senate President Renan Calheiros was targeted by the Lava-
Jato investigations in March 2015, being perhaps the most prominent name
on the PGR’s list. In addition to that investigation, Calheiros became the
target of 11 successive investigations for fraud, corruption and
embezzlement. Of these cases, ten derived from the Lava-Jato probe, and
on top of these Calheiros was investigated in a separate inquiry of the
Zelotes case, mentioned below, as well as in a much older case from 2007
that, back then, almost cost him his career.
The oldest case, filed away in the STF for years, in which Calheiros
became indicted and charged, dealt with the monthly payments to his
former lover and mother to one of his children. The money allegedly came
from the construction company Mendes Junior, in return for political
support for legislation that benefitted the company. Together with a scandal
of indirectly controlling media outlets in his home state Alagoas (which is
prohibited for persons in public office), these cases brought Calheiros’ first
mandate as president of the Senate to an end in December 2007, when he
resigned after months of pressure. However, he remained in the Senate and
returned to the presidency of the Senate in February 2013.
The 2007 case of exchanging political support for clandestine payments
to a mistress was put on the agenda of the STF on December 1, 2016, with
the court accepting some of the elements of the PGR’s indictment, meaning
Calheiros now faced charges. The fact that Calheiros had been charged by
the STF placed his position as president of the Senate at odds with the STF
verdict in Cunha’s case, which ruled that no charged individuals could
remain in the line of succession to the presidency. Four days later, STF
judge Marco Aurélio de Mello received a petition, from the party Rede, for
removing Calheiros from the position of president of the Senate. The very
same day, Mello gave a preliminary ruling and suspended Calheiros from
office. At first, Calheiros evaded the civil servant that was dispatched to
notify him, and upon receiving the suspension, he simply ignored it – with
the support of the Senate. The vice-president of the Senate, Jorge Viana (of
the PT), and the Board of the Senate chose to wait out the impasse,
expecting an eventual confirmation of the preliminary ruling by the STF
plenary. The failure to comply with a STF ruling, in theory, could mean that
Renan Calheiros was now in danger of being apprehended by the police.
The conflict arose as the Senate was about to go on holiday break in
December. This moment coincided with the very last days of Calheiros’
mandate as president of the Senate. For two days, the sense of constitutional
conflict filled the headlines, but on December 7, the STF plenary
overturned Mello’s ruling. Six of the nine judges present in the court argued
that since Calheiros was not immediately next in the line of succession,
there was no urgency to remove him at the very end of his mandate, and
that the court should avoid ruptures of the “harmony between the branches
of Government” (Mello 2016). So, in the end, Calheiros managed to
weather the storm and left office as planned, handing the seat of the
presidency in the Senate over to fellow PMDB senator Eunício Oliveira in
February 2017. Oliveira also came under investigation in several inquiries
of the Lava-Jato case. Following the Calheiros impasse, the principal
parties of Congress deliberated over a constitutional amendment shielding
presidents of both Houses from this kind of judicial intervention. Calheiros,
moreover, called for reforms to reduce the salaries of judges and spoke in
favor of harsher punishments for judges who abused their authority.

The Case of Aécio Neves and the Second Constitutional Crisis

The conflict between the Senate and the STF resurfaced in May 2017. The
same basic question was at stake: The First Turma of the STF had, after
repeated requests from the PGR, ruled that Aécio Neves be removed from
his mandate, although the judges preferred not to rule that Neves be
arrested. While in the cases of removing Cunha and Calheiros, the legal
argument concerned the presidential line of succession, the STF had less
jurisprudence on which to build the Neves case. Neves, like Michel Temer,
had been caught on tape in conversation with the meat magnate Joesley
Batista. In a meeting which was secretly recorded, Neves had asked Batista
for bribes of R$2 million to pay for defense lawyers in the Lava-Jato case.
Neves designated his cousin to receive the money, who, upon receiving it,
hid the bags of cash with an aide of Neves’ fellow senator from Minas
Gerais, Zeze Perrella.7 Since the deliveries of cash made by Neves’ cousin
were monitored by the PF, there was little doubt that Neves had been caught
red-handed, but at the time the STF did not act in a consistent manner: As
the information turned public, judge Fachin first decreed Neves be
suspended from mandate and denied him the rights to leave the country or
talk to others under investigation in the Lava-Jato case. Neves appealed and
continued to appear in the Senate, talking to other politicians including
other senators under investigation. The Senate clerks did not even remove
Neves from the voting roll-call list.
As the limelight during this week in May was mostly directed at
President Temer and his attempts to salvage his own credibility, Neves
remained in Congress, ignoring the ruling. The corruption probe before the
STF into Joesley Batista was separated from other Lava-Jato investigations,
and because of this a new judge rapporteur was required. The case was
redistributed to Judge Marco Aurélio de Mello, who two months later
reversed the ruling. Even though the PGR appealed Mello’s ruling, Aécio
Neves had now had time to ignore the scandal and consolidate the coalition
of Michel Temer at a critical moment, just after the Batista leaks had shaken
the government to its core. Neves’ party, the PSDB, wound up divided in
half over the question of supporting Temer and whether or not to suspend
Neves permanently from presiding over the party.
Only in September 2017, four months after the tapes became public, did
the STF deliberate on the matter of Aécio Neves again. This time, the case
was taken up by the five ministers of the Primeira Turma, who decided,
three votes to two, that they would reinstate the initial measures of judge
Fachin by restricting both Neves’ personal liberty (through a nightly curfew
and prohibition of travels abroad) and his political rights by suspension of
his mandate. Had they decided upon the arrest of Neves (as the PGR had
requested), this decision would have had to have been ratified by the Senate
– and nobody expected this to happen since Neves, runner-up in the 2014
presidential election, was central to the political games unfolding in the
Senate and the government coalition.
Therefore, to avoid this situation, three of the five judges in the Primeira
Turma decided not to order the arrest, instead restricting Neves’ freedom
and political rights – measures not hitherto seen. The earlier cases in the
Lava-Jato scandal had exhibited some similarities, and also some
differences: The arrest of the senator Delcídio do Amaral was ordered in a
clear-cut in flagrante situation, while removing Cunha from his mandate
was an exceptional measure without constitutional grounds. In both cases,
the two parliamentarians had already lost their political support as the
evidence mounted. In the case of Cunha, the presidential succession was
included in the legal argument. This became the single point of
jurisprudence for the (temporary and unsuccessful) removal of Calheiros.
According to the Constitution, however, removing parliamentarians from
mandate is the prerogative of the respective legislative bodies and arrest of
parliamentarians in office requires ratification from the peers in Congress.
Arguing for the restriction of Neves’ liberty, Judges Fux and Barroso
referred to the Brazilian criminal code which lists the restrictions analogous
to prison – and these analogous restrictions are not mentioned in the
Constitution’s clauses on checks and balances which made the judges argue
that these measures can be imposed on senators without ratification from
the Senate plenary.
The Neves case was then put on the plenary agenda in the STF, resulting
in another reversal on October 11. The small majority in the Primeira
Turma in favor of restricting Neves’ rights during the investigation was
matched by a small majority of three judges in the Segunda Turma, voting
against the intervention: Judges Fachin and Fux, Barroso, Weber and Celso
de Mello preferred to avoid an outcome where Neves obstructed
investigations and affronted the public sense of justice. The other five
judges did not want to infringe upon the checks and balances laid down in
the Constitution, preferring to leave it to the Senate to decide. STF
president Cármen Lúcia ended up voting with the latter group, although she
stated that she did agree with Judge Fachin’s view (Supremo Tribunal
Federal 2017). The following week, the Senate overruled the restrictions
imposed by the Primeira Turma (with the votes 44 for and 26 against), and
two weeks later, the Conselho da Ética of the Senate also absolved Neves
by filing away the disciplinary case against him.
Besides the recorded conversation with Batista, Aécio Neves had been
the target of various other lines of inquiry in the Lava-Jato probe. Neves
was simultaneously implicated by Odebrecht executives, by imprisoned
politicians (Delcídio do Amaral and Pedro Corrêa), and by the doleiros
Alberto Youssef and Carlos “Ceará” de Souza Rocha, leading to nine
different inquiries.8 None of these caught the attention of the media to the
same degree as the case described in this section.
Neves’ corruption cases, which led to discontent in the party and to his
temporary stepping down from the party’s presidency, became emblematic
for the general image of the leading PSDB politicians. The party has been
the ideological leader of the Brazilian right for two decades, and the only
right-wing party to have fielded presidential candidates in the second round
of voting since 1994. The established leaders of the party (apart from
Neves, the São Paulo governor Geraldo Alckmin, presidential runner-up
against Lula in 2006, and ex-minister José Serra, runner-up against Lula in
2002 and Dilma Rousseff in 2010) were also investigated in the Lava-Jato
probe, implicated by the Odebrecht testimonies. With the constitutional
crisis and the near-rupture of the party provoked by Neves case, PSDB was
pressed for time to find a consensual presidential candidate.

The Nine Legal Actions Against Ex-President Lula

While PSDB suffered in the polls in 2016 and 2017, Lula’s support among
electors was not hampered much in the polls by corruption accusations.
Comparing the cases of Neves and Lula, it becomes apparent that voters
may indeed prefer not to switch political allegiances in the face of
corruption denunciations, because the opponents, or the alternatives, look
just as corrupt. The situation preceding the 2018 elections underscored
Rennó’s observation of Brazilian electoral accountability very clearly:
“[p]erception of corruption matters only when it comes to choices between
candidates on the same side of the ideological spectrum” (Rennó 2011: 71).
Unlike the right wing, which had a number of virtual contenders for the
2018 presidential campaign, the left side of Brazilian politics could provide
no obvious alternatives to Lula. With little incentive to switch sides, Lula’s
public support remained high throughout 2016 and 2017. However, when
asked about which candidates they would never vote for in 2018, the group
of voters rejecting Lula also grew, while his trials in various Brazilian
courts progressed (Datafolha 2017: 5). In the following, I will describe
these investigations and trials in greater detail.

The Triplex Case

In April 2015, almost one year before the MPF indicted Lula for corruption,
the weekly magazine Veja denounced Lula and Léo Pinheiro, the president
of the large Brazilian construction company OAS (Coura and Marques
2015). Pinheiro had been arrested and imprisoned preventively in
November 2014, in the seventh phase of the Lava-Jato probe. The front
page and a feature article of Veja (edition #2423, April 2015) revealed how
OAS had built a few select beachside apartments in a larger construction
project originally planned by a bank employee association named Bancoop.
The real estate project, located in Guarujá, São Paulo, was co-owned by the
shareholders of Bancoop, with the imprisoned PT treasurer João Vaccari
Neto also having a share – as well as a commanding position in Bancoop.
According to the article, the beachside apartment complex was allegedly
constructed at Lula’s request. In fact, according to Veja’s interpretation of
the OAS director’s testimony, Bancoop had been used as a façade to funnel
money into PT campaigns, and when Bancoop went bankrupt in 2009, the
company never completed construction of the apartments it was supposed
to provide for the associates. This information, however, was merely a
repeat of an O Globo article from December 2014 (Oliveira 2014), detailing
how Lula’s wife had acquired a share in the association and the rights to an
apartment. After the bankruptcy of Bancoop, OAS executed the
construction of some of the buildings, in the end only providing apartments
to 5,000 of the 8,000 shareholders, Lula’s wife among them. The apartment
(known as a triplex) allegedly intended for Lula’s wife, moreover, had an
elevator fitted, separate from the rest of the building.
The case of the Bancoop bankruptcy and its alleged role as façade for PT
and Lula was brought before the prosecutors of São Paulo state in August
2015. Veja continued to publish on the case and reported that Lula had been
seen at the apartment several times. Footage of this only surfaced in March
the following year, however. The sale of the apartment was never finalized,
as Lula and his wife declined to buy it in November 2015. That they never
did get to enjoy the alleged fruits of this corruption did not hinder the
prosecutor of the São Paulo state Cassio Conserino from affirming, during
an interview in Veja (edition #2462) in January, that the ex-presidential
couple were to be indicted. According to Conserino, Lula had concealed his
ownership of the apartment in order to hide the spurious relationship to the
construction company OAS.
The media storm following this interview was sustained into the
following week, when federal police agents of the Lava-Jato investigation
paid a visit to the beachside building in Guarujá on January 27. The agents
investigated the imprisoned PT treasurer Vaccari Neto and a lawyer named
Nelci Warken, who eventually would be found to have connections to the
Mossack Fonseca company, of Panama Papers fame. The investigation did
not target Lula specifically then, and the São Paulo prosecutors were still
yet to denounce Lula. In fact, following the advance announcement of the
denunciation in Veja, Lula’s defense lawyers succeeded in challenging the
competency of the São Paulo prosecutors. In a political climate becoming
more polarized by the day, the São Paulo prosecutors originally looking into
the Bancoop case called for a press conference on March 10. Here, the
prosecutors Conserino, Blat, and Araújo presented a denunciation of Lula,
his wife, and their oldest son Fábio Luis, and requested authorization of the
judge to arrest Lula along with Vaccari Neto and two OAS employees. The
request for his arrest had a favorable timing in relation to the protest
marches scheduled in the major capitals of Brazil three days hence.
The request for Lula’s arrest was first filed with the São Paulo judge
initially allocated to the case, Maria Priscilla Oliveira. However, in a series
of legal skirmishes, the right to not only indict but also to adjudicate the
case of the beachside apartment was pushed back and forth between
different courts during March and April 2016. Staying out of the media
limelight, Oliveira abstained from the case by affirming that the triplex case
derived from the cartel formation and graft of Petrobras contracts. This
meant relocating the jurisdiction to the Curitiba court presided over by
Sérgio Moro. Oliveira also pointed out several missing links in the
accusation of the São Paulo prosecutors, and added that any request of
preventive arrest to apprehend evidence was now rendered meaningless,
since the prosecutor Conserino had already jumped the gun by conceding
the interview to Veja.
On March 14, the case files were sent on to Curitiba – the same day that
Lula’s decision to accept nomination as minister in the Rousseff
administration was made public. That day, before the leaked telephone
recordings described in the previous chapter, Lula’s nomination was
interpreted in the national media as a way of ensuring that the Curitiba court
would not authorize the preventive arrest of Lula. One example of this
connection could be found in Folha on March 14:

The ex-president was more willing to accept the nomination after the
judge Maria Priscilla Veiga Oliveira, of the 4ª Vara Criminal da
Capital, declined to authorize his imprisonment and decided to send the
petition to be analyzed in Curitiba, where the PT leader is investigated
in the Operation Lava-Jato.
(Bergamo 2016, translation mine)

The following day, this interpretation became the dominant prism for
understanding not just the nomination, but in particular the telephone
recording between Rousseff and Lula that was leaked from the Curitiba
court, described in Chapter 3.
The intervention of the STF judge Mendes ensured that the triplex case
remained in the hands of both the Lava-Jato task force and Judge Moro in
Curitiba. The prosecutors of the task force denounced Lula in September
2016, and the case went to trial a few days after the denunciation. The
phase of testimonies lasted for six months, featuring ex-President Cardoso
among others, concluding with a five-hour interrogation of Lula before
Sérgio Moro. Lula continued to deny the allegations, claiming that he had
never owned the beachside apartment and that the trial constituted a
political persecution. Moro sentenced Lula to nine years and six months in
prison on July 12, 2017, while the OAS president Pinheiro was sentenced to
ten years and eight months for paying kickbacks to Lula in the form of the
apartment, the interior design, and the extra elevator, amounting to R$2.25
million. The kickbacks to Lula were allegedly drawn from a larger pool of
R$16 million, originating in a Petrobras contract with OAS. Pinheiro’s
sentence in the triplex case was added to his previous sentences in the Lava-
Jato case.

The Atibaia Ranch Case

Léo Pinheiro’s close relationship to Lula during and after his presidential
mandates was the target of another inquiry into kickbacks occulted in real
estate. The Veja article of April 2015 detailed how a ranch in Santa Bárbara,
Atibaia (in the interior region of greater São Paulo) was also under scrutiny
as the ranch and the rapid renovation of the site possibly constituted
kickbacks for Lula. A pool was enlarged, a porch added, and so on,
allegedly at Lula’s request, despite the fact the estate was not actually
owned by Lula, but by friends of his. According to the investigation, the
total value of the renovation amounted to around R$1 million.
On November 17, Folha revealed that the police were looking into the
OAS accounts in order to determine what the company had been doing at
the Atibaia ranch and exploring the possible relationship to the contracts
between OAS and Petrobras (Ferreira and Rocha 2015). The case was
largely forgotten in the final weeks of 2015, because Lula’s acquaintance
José Carlos Bumlai was arrested in the Lava-Jato probe, and Rousseff soon
after faced impeachment proceedings. However, the Atibaia site became the
focus of intense scrutiny in 2016, in parallel to the investigations into the
triplex case. The two cases were often bundled together in media coverage,
but actually had different trials in the Curitiba court (once the triplex case
had returned from the digression of the process to São Paulo). According to
the prosecutors, Odebrecht and OAS had split the construction costs for the
ranch site and in return received political influence in government tenders
via Lula. On August 1, 2017, a fortnight after being sentenced in the triplex
case, Lula was charged in Curitiba in the case pertaining to the ranch in
Atibaia.

The Palocci Case and the Instituto Lula

A third line of inquiry, also investigated and conducted in Curitiba,
explored the evidence arising from the testimony of Marcelo Odebrecht.
After leaving the presidency, Lula had given lectures and talks at various
business events across South America. This activity had been the focus of
interest in Brazilian news outlets, speculating that the payments were not
merely the market price for the talks by an internationally renowned ex-
president, but a way of making illicit transfers seem legitimate. Marcelo,
the former president of Odebrecht, testified that his business empire had not
only paid for some of Lula’s talks in the period following his presidential
mandate, but had also acquired a site in the Vila Clementina neighborhood
of São Paulo for constructing a new building that would eventually have
housed the Instituto Lula.
The institute, which was functionally a political communications office,
was the base of Lula’s post-presidential political oeuvre. To be more
precise, the institute was the seat of Lula’s political activities; at least until
the police operations of March 2016 in the triplex case, when the offices
were ransacked by police agents. After this, Lula increasingly took to
traveling about the country, rallying political support. The institute, since it
opened in 2011, had been located in the Ipiranga neighborhood of São
Paulo, but according to the testimony and evidence presented by Marcelo
Odebrecht, the company had purchased a new lot for the institute’s offices.
Hindered by red tape in the municipal authority, no building was ever
constructed. As with the inquiry into the triplex, the case thus seemed rather
hypothetical, since Lula could not really have enjoyed this bribe, and
because the denunciation, like the earlier cases, was not linked to any
specific political projects promoted by Lula while in office.
The originally fragile case was greatly helped by a plea bargain,
however. By following the money, the PF succeeded in finding evidence
that linked the site in São Paulo to Lula’s former minister of Finances,
Antônio Palocci. Palocci, also a founding member of PT, was arrested in
September 2016 on these grounds and chose to plea bargain after six
months of preventive arrest in Curitiba. In his plea bargain, he confirmed
the story told by Marcelo Odebrecht, explaining how he had controlled a
large fund of resources granted by the Odebrecht company in exchange for
extended political goodwill. These funds, after Lula left office, had been
earmarked in order to have Lula use his influence and act as a mouthpiece
for the company’s interest during the Rousseff administration. Rousseff,
taking the mantle, was apparently not as aligned with the business elite. The
Odebrecht family believed that Lula could sway her at the crucial moment
if need be. Palocci admitted that he was the designated go-between for Lula
and Marcelo Odebrecht, and furthermore, he claimed that he had
accompanied Dilma Rousseff at a meeting where everyone present
witnessed how well aware Luna was of the corruption in Petrobras.
The relationships between Lula in his post-presidential period,
Odebrecht, and OAS, resulted in these three cases of alleged corruption in
2015, 2016, and 2017. Three more cases, not directly connected to
Petrobras and the subcontractors, became trials in the courts of the Federal
District of Brazil, as they investigated possible criminal acts committed by
Lula while residing there as president.

Operation Zelotes and Tax Exemption by Provisional Decrees

In March 2015, a police operation investigating a systematic scheme of tax
evasion was carried out. The operation, code-named Zelotes, had as its
principal target CARF (Conselho Administrativo de Recursos Fiscais), a tax
appeal agency under the Ministry of Finances. The initial estimates
identified close to R$19 billion being withheld from the state during a
period of five years, as the counselors of CARF had systematically favored
a range of companies, including three major banks (Safra, Bradesco and
Santander), the steel company Gerdau, and several construction companies.
Furthermore, the inquiry targeted the Globo-affiliated media company RBS.
Perhaps due to the operation targeting the media conglomerate Globo, the
Zelotes operation was mostly covered in the left-wing press: In April, the
left-wing magazine CartaCapital pointed out that despite a month of
coverage, other media, especially those from the Globo group, had been
reluctant to report on the case. In contrast to the media-savvy style adopted
by Curitiba’s court, which readily published material from investigations,
the Brasília-based court at which the probe was conducted did not
disseminate material on the case. Like the Lava-Jato case, a task-force
conducted the probe, under the auspices of federal judge Vallisney de Souza
Oliveira and the federal prosecutor Frederico Paiva, but without the
publicity and journalistic access.
As the next stages of the Zelotes investigation proceeded, evidence arose
that linked some of the investigated companies to provisional decrees
issued by the presidency (medidas provisorias). On October 1, 2015, the
Estado front page featured the scoop that two law firms, SGR Consultoria
Empresarial and Marcondes&Mautoni, had not only lobbied for lenient
verdicts in the appeal council CARF, but also for specific tax exemptions
for the automobile industry in Congress and with the executive branch. The
exemption, according to Estado, was specified in the provisional decree MP
471, issued in 2009 by Lula, later ratified by Congress, with effects until
2015.9
Later in October 2015, the police investigations centered on one link in
the supposed chain of bribes from the automobile sector: A paper trail that
linked the law firm Marcondes&Mautoni, used by automobile
manufacturers, to Luís Claudio da Silva, another of ex-President Lula’s
sons. Investigating this trail, the task-force obtained search warrants and
examined the house and office of Luís Claudio da Silva. With this, the
Zelotes case began to be covered by the mainstream press. While this paper
trail was only one of the links between government and the automobile
industry, a range of news items were produced on October 26 concerning
the police action in his consulting service, as well as the law firm paying
Luís Claudio da Silva for consultancy. The nature of da Silva’s consulting
services, the exact content of a report he delivered to the law firm, and
possible plagiarism in the report was the object of the news. The frame, in
other words, was the suspect nature of his consultancy, and that the report
was only delivered to cover for the fact that the R$2.5 million paid for the
service was, in fact, a (somewhat late) kickback for the provisional decree.
On these suspicions, the two lawyers behind Marcondes&Mautoni were
arrested preventively, and ministers of the Lula and Rousseff
administrations were subpoenaed. On January 6, 2016, Lula testified before
the federal judge in Brasília. Both Lula and Rousseff were subpoenaed
again in the case on January 21, as were many ex-ministers. Prominent
testimonies followed, including the PT ex-ministers Gilberto Carvalho,
Guido Mantega, and Miguel Jorge. Three months later, on April 30, 2016,
the prosecutor of the case obtained the authorization of the court to open
another line of inquiry, investigating senators Renan Calheiros, Romero
Jucá, and Gim Argello (an ex-senator of the PTB party who had already
arrested in the Lava-Jato case, 14 days before). Testimonies in the case had
mentioned that some R$45 million had been offered to these senators to
ensure that the medida provisoria was passed, favoring the car
manufacturers. This probe was one of a dozen inquiries at that moment
targeting President of the Senate Renan Calheiros.
 The Zelotes investigations resulted in two trials against Lula, his son, the
ex-minister Gilberto Carvalho, and a number of lobbyists and CARF
councilors. In the first case, Lula was charged before the 10th court of the
federal district for selling political influence to the Brazilian automobile
industry and to the Swedish company SAAB (in order to secure the sale of
36 military aircraft), for money laundering and prolonging the MP 627 with
criminal intent. In the second, Lula stood accused of corruption in the
issuing of the tax exemption bill MP471.

Obstruction of Justice in the Cerveró Case

One of the former directors of Petrobras, Nestor Cerveró, was the object of
intense interest in late 2015. While the director Paulo Roberto Costa had
opted for a plea bargain back in 2014, Cerveró was only apprehended in
January 2015, and did not admit to the charges while imprisoned in
Curitiba. In Congressional hearings, the former director of Petrobras’
International Division remained silent. As the months wore on, PT senator
Delcídio do Amaral tried to find a way for Cerveró to flee the country –
supposedly to avoid another incriminating plea bargain. Amaral, former
Petrobras director during the Cardoso administration and former colleague
of Cerveró, was secretly recorded on tape deliberating how to manage the
escape. Cerveró’s son Bernardo had participated in a meeting with Delcídio
and decided to tape it. As he handed over the tape to the authorities, the
extraordinary situation of apprehending a senator in public office became a
reality. While this can only happen (as discussed in the Neves case) with the
ratification of the Senate, few Senators had the nerve, under public scrutiny,
and in the face of flagrant crimes, to deny the request for Amaral’s
immediate arrest.
The tape provided by Bernardo Cerveró was one of the conditions upon
which the reduction of Nestor Cerveró’s sentence rested. Cerveró also
detailed how he had helped senators like Amaral and Renan Calheiros to
channel money from Petrobras into their political campaigns, and claimed
that the details of a remarkably bad Petrobras deal (the buy-out of the
Pasadena refinery in Texas) was well-known to Dilma Rousseff, then
director of the Administrative Board of Petrobras. Furthermore, he
suspected that both Lula and Rousseff were well aware of the graft of the
state oil company. Senator Amaral, in turn, claimed that his failed attempt at
having Cerveró flee the country was actually a direct command from Lula.
The PGR indicted Lula in May 2016, but the charges based on Amaral’s
testimony did not hold up to scrutiny: In September 2017, state prosecutor
Ivan Claudio Marx requested that the inquiry be suspended and the 10th
court of the Federal District discard the case.

Operation Janus
The operation named “Janus,” on May 20, 2016, followed hot on the heels
of the impeachment of Dilma Rousseff. The operation investigated one line
of credit within the BNDES (the Brazilian state investment bank) extended
to the construction company Odebrecht in a set of international sub-
contracts in Africa, Cuba, and the Dominican Republic. On October 7,
2016, the prosecutors of the Federal District indicted Lula, Marcelo
Odebrecht, and nine other men for fraud and corruption, securing credit
lines in the BNDES, eventually leading to contracts between Odebrecht and
the Angolan state. The alleged operator of the kickbacks, according to the
denunciation, was a nephew of Lula’s ex-wife, Taiguara Rodrigues, who
channeled money through subcontracts between Odebrecht and his
company Exergia Brasil. Some of these funds allegedly paid for monthly
allowances to Lula’s brother Frei Chico.
The court of the federal district was the scene of three more cases against
Lula, of which one was rejected on the grounds of weak evidence and one
was split into two formal denunciations. Furthermore, in the final days
before the PGR Rodrigo Janot was substituted as head of the MPF by
Raquel Dodge, Janot indicted Lula two more times, with one of these cases
being redistributed to the federal district court. Janot’s final indictments
identified Lula as the criminal mastermind behind the Petrobras corruption.

The Criminal Mastermind behind the Graft in Petrobras

On September 5, 2017, Janot indicted Lula and the leaders of the PT for
pursuing the grand scheme of corruption in Petrobras, resulting in
kickbacks of R$1.5 billion in total. This case, by virtue of its objects and the
history of inquiry, was sent straight to the STF. Together with three
indictments targeting other political groups (the erstwhile coalition partner
PP, as well as two separate indictments of the PMDB group in the Câmara
and the PMDB senators), the indictment against Lula and PT detailed how
the division of directories in the oil company was masterminded by Lula
and the financial strategists of PT. The accusation also stated that Dilma
Rousseff was conscious of the corruption and in cahoots with the PT
Senator Gleisi Hoffmann, Hoffmann’s husband Paulo Bernardo, and ex-
Ministers of Finance Guido Mantega and Palocci.
Less than 24 hours after these indictments, the PGR filed another
indictment, this time accusing Lula, Rousseff, and ex-Minister Aloízio
Mercadante (a minister in both of their administrations), of once again
obstructing justice in their attempt to prevent Delcídio do Amaral from
negotiating a plea bargain. The indictment also denounced the criminal
intent in nominating Lula for minister in the final days of the Rousseff
administration. The PGR affirmed that the two indictments were interlinked
but divided because the criminal organizations acting in the cases were
distinct. Judge Fachin of the STF dispatched this second case to the court of
the federal district three days later, since none of the targets enjoyed the
foro privilegiado at this point, after the impeachment of Rousseff.
With one case discarded in the Federal district, Lula faced trials (or
appeal processes) in eight cases total at the end of 2017. At the time of
writing, they have had the following results so far:

1 (Curitiba) The ranch case: Lula was sentenced to 9.5 years of prison by
Judge Sérgio Moro in July 2017, and the appeal court TRF-4
augmented the sentence to 12 years in January 2018.
2 (Curitiba) The beachside apartment case: Lula is charged.
3 (Curitiba) The Institute construction site case: Lula is charged.
4 (Federal district) The SAAB, automobile industry, and MP627 case:
Lula is charged.
5 (Federal district) The provisional decree MP471: Lula is charged.
6 (Federal district) Influencing BNDES in Odebrecht’s Angola
operations: Lula is indicted.
7 (STF) The Petrobras graft mastermind: Lula is indicted.
8 (Federal district) The obstruction of justice through Mercadante: Lula
is indicted.

The combined prison sentences of these trials may amount to more than
100 years in prison. Moreover, the Instituto Lula is under another
investigation for money laundering, although the director of the Institute,
Paulo Okamotto, has so far been acquitted in one of the cases involving
Lula.

Outcomes of Judicial Actions in the Political Arena

In parallel with the Lava-Jato investigations, and frequently in consequence

of events pertaining to that case, judicial actions have influenced and
changed the political processes of Brazil. The outcomes of these judicial
actions have not however been simply the punishment of political
transgression in its various forms. Furthermore, the Lava-Jato case was not
simply a case in which public office-holders lost their political capital.
Rather, a myriad side-effects and macro-level changes are visible in the
political arena as a result of the legal processes spawned during the Lava-
Jato investigations. I will summarize these effects and categorize them as
follows:

• overturning elections;
• rigging elections;
• removing individuals from mandates of public office;
• protecting individuals from actions against their mandates of public
office;
• extinguishing or severely damaging parties;
• changing political capital and costs of forming coalitions.

Overturning Elections
Without a doubt, the legal actions leading to the impeachment of Dilma
Rousseff and, subsequently, the acquittal of the campaign of Rousseff and
the (then) vice-president, were the most the high-profiled judicial outcomes
in the period 2014–2017.
Both the electoral court case and the impeachment processes against
Rousseff and Temer could have overturned the result of the presidential
election of 2014. This would constitute the ultimate judicial intervention in
a democracy. None of these processes, however, actually convicted the
election result of 2014 of corruption – despite the evidence produced and
the political climate of crisis that characterized the period. As a result of the
timing and the outcomes of the processes, Brazil lost a democratically
elected president, and was left with an acquitted but still investigated
president with record-low approval ratings. While the successful
impeachment of Rousseff was executed on a quasi-legal basis (Wink 2017)
grounded in fiscal delays, the electoral court case was dismissed on formal
grounds pertaining to the scope of the original petition for cancelling the
elections. Thus, both cases had to evade the specific revelations of the
Lava-Jato case in order to create these particular political outcomes, despite
the fact that neither process would have been initiated without the Lava-Jato
probe.

Rigging Elections
Several of the Lava-Jato denunciations and trials are of critical import to the
upcoming 2018 elections. Lula’s sentence in the Triplex case, and the
various delays in cases being judged before the STF will allow a number of
investigated and denounced parliamentarians to run for office in 2018
(including the presidency), but keep the ex-president outside. Under normal
circumstances, keeping a corrupt ex-president from running for the
presidency should be beneficial to democracy. However, the denunciations
against Lula after he left the presidency seem only tangential to the much
larger problem of graft in state companies, in the sense that Lula was
perhaps influential, if not exactly calling the shots. In economic terms, the
real estate cases of Lula pale in comparison to parallel political corruption
cases such as Cunha’s and are dwarfed next to the tax evasion of some
larger Brazilian companies.
Another situation triggered by the Lava-Jato case was a bill, proposed in
parliament in 2017, that would change the method of allocating votes from
proportional voting districts to a most-voted district system. Besides
wasting many cast votes, such electoral reform would further help secure
incumbent parliamentarians their seats thanks to greater access to funds
(licit public funds as well as illicit ones), all the while remaining in office
and maintaining their precious foro privilegiado. Thus, this legislation is
just another means of rigging elections without losing the appearance of
legitimacy, although the intention is anything but legitimate.

Removing Individuals from Mandates of Public Office

In this chapter, several specific cases have been highlighted, in which
political actors have removed opponents from their mandates. In the cases
of Dilma Rousseff and Eduardo Cunha, such attempts were successful.
There are three other individual cases which should be mentioned, however.
The case of Renan Calheiros has already been described: Calheiros was
removed as president of the Senate for a few days, through the monocratic
intervention of STF Judge Marco Aurélio de Mello quickly overturned by
the STF plenary. Senator Aécio Neves avoided a similar fate by one vote in
the STF.
A more important case of this type of outcome pertains to President
Michel Temer. Temer was not only denounced twice by the PGR, which, if
the Congress had so chosen, would have removed him from office. Temer
was also, even when still vice-president, the target of impeachment
petitions. The first of these argued that what was true of the decrees issued
by Dilma Rousseff which were interpreted as illicit loans outside of the
Congressional license, also had to be true for identical decrees. Since
Rousseff had been on several international journeys, Temer had in fact
signed several of these decrees as acting president. This should have thus
made him complicit in the same illegal act as Rousseff. Later impeachment
petitions against Temer, once he assumed the mantle, based on obstruction
of justice and breaches of decorum while in presidential office, revealed in
the leak of the JBS owner Joesley Batista. These petitions were stalled
indefinitely by the new president of the Câmara, Rodrigo Maia (of the party
DEM) – and perhaps were perpetual political weapons held in reserve by
Maia. This aspect of judicial or quasi-judicial action brings me to the next
outcome, the protection from legal actions.

Protecting Individuals from Actions Against Their Mandates of Public

Office
The president of the Câmara has proven to be of extraordinary importance
in situations of political instability. This is due to prerogatives vested in the
office. He or she is situated third in line for the presidency, while at the
same time has the power to decide whether to pursue impeachment
procedures. The scheduling of roll-call votes and sessions of the Câmara
and the combined Houses must also pass through his or her office,
potentially leading to significant delays in other proceedings such as
disciplinary processes. Parliamentary votes may even be annulled on
procedural grounds at the intervention of the president of the Câmara, who
may decide to restart such a process. These are just some of the many
instruments used (and abused) as protection or as a blackmailing device, as
with the case of Eduardo Cunha.
What is perhaps even more extraordinary is the fact that aside from
Eduardo Cunha, no internal cases for removal of mandate have occurred in
three years in either house of Congress, despite – or perhaps rather because
of – the mountain of evidence produced during the Lava-Jato investigations.
An obvious explanation for this extended principle of silence would be the
congressional fear of an all-in melee situation. Keeping up appearances in
corruption scandals seems to be the easy way of tackling the negotiation
problem that appears if central actors in Congress opt to threaten other
actors’ mandates. It is not just the president of the Câmara who holds
protective powers; it is in effect all major parties that to some extent partake
in the protection of mandates. This explains why, in the Lava-Jato case, we
only see minor left-wing, non-establishment parties such as PSOL and Rede
opting to denounce politicians in Congress. Aggression through Congress-
internal disciplinary/judicial action, targeting political mandates, is
normally shunned, especially so once the snowball of denunciations starts
rolling. Such internal obstruction to action against mandates was also
apparent in the case against Aécio Neves, where not only the STF declined
to intervene, but the Senate plenary and the Senate disciplinary committee
also protected his mandate.

Extinguishing or Severely Damaging Parties

Several legal steps have been taken to judicially or virtually extinguish the
political parties PT and PP. PT is currently the target of a lawsuit in the TSE
arguing that the party as a whole has been funded from sources abroad,
which is prohibited, and that this should make the whole party ineligible.
PSDB filed the action on January 20, 2016. STF and TSE judge Gilmar
Mendes opened a similar case on August 5, 2017, prompted by the TSE
action concerning illicit funding in Dilma Rousseff’s presidential campaign.
PP, meanwhile, is the object of a civil suit for administrative improbity to
the amount of R$2.3 billion and the loss of mandates and rights to
candidatures for most of the party leaders. The suit was filed by the PGR on
March 30, 2017. Both of these legal actions – if successful – would have a
huge impact in Brazilian politics, since PT and PP are the second and
fourth-largest parties in Congress, respectively, commanding a full fifth of
the seats in total. Neither of the cases have advanced, however.
Furthermore, on October 6, 2017, the Congress passed a bill to protract the
payment of such fines in civil suits. In effect, if found guilty, the fine of PP
could be divided into 24,000 monthly fines, which would stretch the fine
hypothetically across the next 2,000 years (Faria and Peron 2017).
Curiously, the parties are themselves perhaps the weakest link of all in
the web of accountability institutions, and are not even considered as such
in works on the subject (such as Power and Taylor 2011). Elsewhere, parties
are producing accountability to signal distance, saving the image of the
party, and minimize electoral consequences, often by expelling members
accused or guilty of corruption. This has only occurred thrice in the Lava-
Jato case, with the PT moving to expel members on each occasion. André
Vargas was expelled early on (in 2014) for corruption denounced by
Alberto Youssef, however since then only Antônio Palocci and Delcídio do
Amaral have been expelled; in a strange twist to the case not due to
confessions of corruption, but rather for lying and accusing the party
leaders Lula and Rousseff of corruption.
The other twenty-odd PT politicians investigated or charged in the Lava-
Jato case have not been targets of any party-internal accountability
procedure, and neither has any politician in PSDB, PP, PMDB, PR, PSD,
PTB, PTC, PCdoB, PDT, or DEM, to this date. This negligence can perhaps
justify the move toward removing parties from the ballot, and it certainly
shows that the parties do not consider it beneficial to strengthen their ethical
position and public image by activating internal accountability procedures.
Collective image, or the association between politician and party, may be
very weak in Brazil, and the costs of potential infighting may anyway be
too high for anybody to bother with this kind of initiative in a politico-legal
environment already filled with legal actions.

Changing Political Capital and Costs of Forming Coalition

Ending, or attenuating, the rampant use of appointments to state companies
and agencies in exchange for kickbacks, is perhaps the most positive effect
of the Lava-Jato case. Even this effect has a problematic side-effect, strange
as it may sound. As Brazilian executives lose this instrument of pork-style
politics, the demand for other compensations from veto players in Congress
(Tsebelis 2002) may rise, as may the political costs of forming coalitions.
This could lead to increased use of (legal) pork-barrel policies and
improved ideological cohesion of Brazilian governments. In theory, this
should bode well for the moral condition of governing and forming
coalitions. With the hindsight provided by the Lava-Jato investigations, it
now seems probable that for decades, the use of illicit payments was an
agglutinating factor in the coalition-building of Brazilian governments. It
remains to be seen, when the dust has settled, if top positions in state
company and clientelistic influence over tenders will once again be in
demand as payment for votes in the fragmented Congress. The (hopefully)
lower influx of illicit funding in the wake of the Lava-Jato case might
obstruct the possibility of building and maintaining super-coalitions, thus
bringing more deadlock situations to the Brazilian democracy (Avritzer
2016). The bargaining value of state company appointments may have been
underestimated hitherto in the literature on Brazilian coalition formation
(Pereira et al. 2016, Raile et al. 2011).
On top of these outcomes might be added two sub-goals of legal
processes and interventions which help to reach outcomes such as the
above. First, many legal actions, especially the range of appeals available in
the Brazilian system, are used as a means of stalling for time – motivated by
political intentions or for the sake of shifting the public agenda. This leads
to the second instrumental use of legal processes, which is that of setting
the media agenda, or at least filling that agenda. The latter use, agenda-
setting, is considered in the following chapter, while the former sub-goal led
to several of the outcomes discussed above: Michel Temer’s court-packing
of the TSE bench and the Senate’s protracted and successful stone-walling
of STF verdicts are but two of the outcomes of such delay tactics. In the
final sections of this chapter, I will discuss the macro-level effects of these
outcomes and the malleable character of justice and accountability in
Brazil.

Lawfare in the Brazilian Debate and in International Research

What does the plethora of contested legal actions mean to Brazilian
democracy and accountability? As we shall see in the following chapter,
processual movements of legal actions are one of the main sources for
political journalists in Brazil, generating much of the material in the
newspapers sampled. In the above categorization, I have shown the
functions of the judicial interventions in the wake of political scandal.
However, this question must also be asked on a more abstract level.
I would avoid describing the contemporary context of Brazil, marked
and marred as it is by corruption and the vested political interests and
manipulations of justice, as “politicization of justice” or as “judicialization
of politics,” although superficially, both processes co-exist (Hirschl 2004,
2008, Tate and Vallinder 1995). That accountability mechanisms are
activated by corruption scandals is, of course, to be expected. I will argue,
however, that the term “politicization of justice” does not do the situation in
Brazil justice, if the reader will excuse the pun. Rather, I prefer the term
lawfare, because it opens the field of political science to legal-
anthropological insights into the performative and contested aspects of law
and the State – and possibly provides links to the context of contemporary
media spheres and, especially, mediatized politics. This sensitizes research
in a way better suited to capturing the events emerging from the Lava-Jato
scandal, as opposed to research into changing discursive and organizational
forms in Brazilian politics and laws. Simply put, lawfare brings emphasis to
actors and actions, while the perspective of judicialization is more apt for
theorizing societal change.
Cristian Zanin Martins, defense lawyer for ex-president Lula, raised the
concept of lawfare in Brazilian media in October 2016. Zanin Martins used
the concept as a response to an opinion piece in Folha, October 30, penned
by two of the Lava-Jato prosecutors (Dallagnol and Martelo 2016). In the
piece, Dallagnol and Martelo pointed out that the investigation was not
politically motivated, since it targeted only the parties that had been in
command of the graft in Petrobras. Contesting this claim, Zanin Martins
repeatedly made use of the term “lawfare” during the following months, as
in this critique of the Curitiba task-force of Lava-Jato and the indictment of
Lula for corruption in the triplex case:

After the hearings in Curitiba, it must be affirmed that this legal action
of the triplex is another frivolous process, a part of the “lawfare”
 phenomenon that consists in utilizing judicial processes as a means to
an end – an end which is political persecution.
(Zanin Martins 2016, translation mine)

Critique had emerged with the same tenor on the left wing of the
Brazilian political arena since 2015. The main thrust of the critique was not
just the obvious bias of the media attention and framings (described in the
next chapter), but also the concerns about a bias of judicial and
investigative attention. In particular, Lula’s lawyers claimed that the
corruption investigations against him were intentionally speeding up, in
order to avoid Lula running for president in 2018, while other cases
lingered in court dockets or stalled because of repeated reversals of rulings.
Whether or not Lula was and is the victim of judicial-political persecution
to a greater degree than Cunha, Neves, or other political opponents, the
concept of lawfare is relevant to connect the judicial interventions to the
role of the State.
In academic circles, the word “lawfare” was used once back in 1975 in
the anthology The Way Out – Radical Alternatives in Australia, in an article
concerning mediation in legal processes, and then forgotten. The authors,
Carlson and Yeomans, wrote:

Utilitarian law is the law of the State, of order, of business, of war,

contract and crime – the law of ruthlessness, retribution and
punishment. In the last 200 years, this law has uniquely dominated the
Western world. It has swallowed the humane justice of humanitarian
law, creating State monopolisation of law-making. Thus the
inquisitorial or enquiry technique is gone, the adversary or accusatory
procedure alone applies in our courts. The search for truth is replaced
by the classification of issues and the refinement of combat. Lawfare
replaces warfare and the duel is with words rather than swords.
(Carlson and Yeomans 1975: n.p.)

The article departs from a Parsonian perspective and discusses various

nations’ use of mediation in legal disputes. Seemingly disconnected from
the way Yeomans and Carlson used the term, it reappeared in legal
scholarship dealing with warfare after the 9/11 attacks in the United States.
Legal scholars, working on law, attribute the term to a U.S. Air Force
lawyer, who asked, in November 2001, if law was “becoming more of the
problem in modern war instead of part of the solution” (Dunlap 2001: 1).
Here, the term takes on meanings parallel to judicialization, but limited to
the context of international warfare: “Dunlap … describes lawfare as
attempts by opponents of the U.S. projection of power to use law to
discipline and hamper that power” (Waters 2009: 891). Werner points out
that the term is used as “a label to criticize those who use international law
and legal proceedings to make claims against the state, especially in areas
related to national security” (Werner 2010: 62). The restriction and
encirclement of legislative branches by judicial branches, identified by Tate,
Vallinder, Hirschl and other scholars discussing judicialization, is re-
encountered in the restriction of warfare through international judicial
processes and diplomacy.
In parallel, but unconnected to the above,10 the term lawfare was
included in the writings on the post-colonial condition by John and Jean
Comaroff (Comaroff and Comaroff 2006). Here, the term almost has the
inverse meaning, covering the abuse of socio-political state power dressed
up as law in post-colonial contexts:

… what imperialism is being indicted for, above all, is its commission

of lawfare: its use of its own rules – of its duly enacted penal codes, its
administrative law, its states of emergency, its charters and mandates
and warrants, its norms of engagement – to impose a sense of order
upon its subordinates by means of violence rendered legible, legal, and
legitimate by its own sovereign word. And also to commit its own
ever-so-civilized, patronizing, high-minded forms of kleptocracy.
Lawfare – the resort to legal instruments, to the violence inherent in the
law, to commit acts of political coercion, even erasure … is equally
marked in postcolonies, of course. As a species of political
displacement, it becomes most readily visible when those who act in
the name of the state conjure with legalities to act against some or all
of its citizens.
(Ibid.: 30)

The Comaroffs also mentioned judicialization (here and elsewhere):

“Not only is the non-postcolonial world more litigious than ever, but the
judicialization of politics is proceeding apace everywhere” (Comaroff and
Comaroff 2007: 148). “Lawfare” has since then been recycled by various
anthropologists, while similar ideas have simultaneously emerged and
created a burgeoning sub-field of anthropology, researching the juridico-
political struggles at the margins of societies, legal systems, and states (e.g.,
Anders and Nuijten 2001, Das and Poole 2004, see also Gupta 1995). Here,
the term merges with theoretical frameworks geared to the performative,
transitive, and malleable aspects of “law” in everyday practices across the
globe.
Lawfare, in the definition given by the Comaroffs, and the related
anthropological interpretations of law as a range of instruments for
coercion, seems both apt and somewhat imprecise to describe the situation
of Lula and other politicians, however. First of all, in purely abstract terms,
the anthropologists normally describe this coercion as subjugating national
political subjects – in a word, citizens. Lawfare is not, in this literature,
perceived as a homing missile targeted at individual, political actors
(although in practice, all citizens coerced by the abuse of law would
experience lawfare as such). The use of legal violence and state-legitimized
oppression is frequently observed in marginalized groups, who are
exploited or denied their right as citizens by political elites. In the abstract,
to use the term lawfare to describe a tactic for overthrowing a singularly
influential political leader, suspected of grand corruption, is a very
uncommon use of that term.
Second of all, looking at Lula’s concrete cases (to take the obvious
example), we might doubt that the use of legal trials even constitutes
coercion. Granted, Lula was forced to testify in March 2015, despite having
testified voluntarily several times before and after. Similarly, he was denied
the status of minister on grounds that were not applied to ministers of the
Temer administration just months later (which I will detail below).
However, the notions of violence, discipline, and bodily coercion are
perhaps not as relevant in the case of Lula as with those cases normally
explored by anthropologists. On the other hand, the symbolic violence to
Lula’s image is quite tangible and effective, most visibly seen in the street
protests of 2015 and 2016 where inflatable dolls made to look like Lula in a
prisoner’s clothes were brandished by protesters. The imagery deployed
here invokes the desire to see Lula behind bars, or in other words, to see
him subjugated to the state’s exceptional power.
 This brings me to another theoretical question of the applicability of the
term lawfare – namely, the role of the State. In the anthropological research,
lawfare defined as intentional maladministration and abuse of legal power
is the operational mode of kleptocracies and political elites with access to
the ostensibly legitimate powers of states. In the Brazilian cases appearing
in the Lava-Jato case, however, we need to reassess what we mean by the
term “state.” To begin with, we cannot fathom the MPF as “the state” in the
same way as the legislative and executive branches of government are “the
state.” It seems, in the anthropological literature, that powerful elites in the
post-colony cases are commanding the legal and coercive instruments of the
state, but in the Lava-Jato case, sections of the political elite – the center-
left parties along with the governing party PMDB – lost their grip on the
investigative and judicial branches of government. Apart from Eduardo
Cunha, two of Temer’s former close associates and ex-ministers for PMDB
are preventively arrested in the Lava-Jato case, while the current ministers
Moreira Franco and Eliseu Padilha are under investigation in various cases,
and also several aides have been arrested in 2017. Conversely, the
continued survival of Michel Temer as president demonstrates that his
presidency is legitimized by other actors, such as half the TSE bench, the
STF judge Gilmar Mendes, and a majority in Congress.
In theoretical terms, the Brazilian case in general and Lula’s case in
particular are perhaps too complex to be subsumed under the concept of
lawfare: When prosecutors and judges operate sometimes independently,
and sometimes in tandem with the executive and legislative branches, terms
from post-colonial studies are not adequate to describe the structures of
political and intra-governmental struggle.
What is fruitful in the concept of lawfare, though, is the attention to the
legitimizing effects of the instruments of law. I will quote here a recent
study of contested jurisdictions that perfectly reflects one aspect of the
partial and biased nature of the judicial decisions taken during the Lava-Jato
case – a quote that, incidentally, also captures how legitimacy was bestowed
selectively by the Supreme Court in Brazil. This study of law and the
pragmatics of jurisdictions highlights how lawfare is present when
discretion and legal power are distributed and spatially located within
overlapping judicial systems:
 These often hidden architectures (cf. Ford 1999), or cartographies, of
discretion are both the condition of possibility for the distribution of
official power and the targets of legal battles over where such power
resides and how it can be expressed. Politicized legal warfare, or
“lawfare” … unfolds within the technically rendered routes, zones, and
fractal hierarchies of jurisdictional space, relying on, contesting, and
reinventing them in the process. By attending to the intricacies of these
juridico-spatial formations and the poetic … structures[,] they are
expressed as performances both mundane and spectacular.
(Kahn 2017: 6)

This quote draws attention to the cases where Brazilian courts

manipulate jurisdictions and the foro privilegiado, but superficially
maintain the legitimacy: While Lula was barred from acting as minister in
the last days of the Rousseff administration, Michel Temer successfully
nominated Moreira Franco for minister in February 2017. The PMDB
strategist Franco had been the government’s executive secretary of the
Program for Investment Partnerships since the first cabinet was put together
by Temer, but as executive secretary he did not have the privilege of being
judge before the STF. In February 2017, as the Odebrecht testimonies
became authorized as evidence in the Lava-Jato case, Temer decreed (by
way of a provisional decree) that the secretariat was now a ministry –
changing nothing else in the cabinet composition, but endowing Franco
with the foro privilegiado. This ensured that Franco would not be under the
jurisdiction of the Lava-Jato task-force and Judge Sérgio Moro, who
commonly authorizes preventive arrests.
As with the Lula case, various judges across the country intervened, but
their verdicts were eventually overturned by the STF judge Celso de Mello.
He ruled that the nomination could not constitute an obstruction of justice,
since ministers are not made immune to investigation by the foro but must
still respond to justice in the STF. The monocratic decision of Celso de
Mello was thus exactly the opposite of the ruling by STF Judge Gilmar
Mendes eight months earlier. Such conflicts of rulings in the STF can of
course be questioned – but again, the timing of such appeals is determined
by the judges, and the appeals can linger in the STF docket forever. The
provisional decree making Franco’s secretariat a ministry reached its
deadline in September 2017, but was passed into law with a very tight
margin of votes in the Congress.
Lawfare is also a performance of legitimizing interventions, not least
through the creation or absence of scandalous coverage in the media. Thus,
the tarnishing of Lula’s image was not solely enacted through courts, but
perhaps more so through the media coverage of the trials, as his project of
performing as a credible virtual candidate was continuously undermined.
The combination of investigations, legal actions, and the possibility of
being barred by the Ficha Limpa law was repeatedly presented as his major
political and legal problem (while the corruption charges were presented as
the major image problem and moral problem for Lula and his party). Again,
the timing of the legal actions is of interest: the possible PSDB candidates
José Serra, Geraldo Alckmin, and Aécio Neves were all implicated in the
Lava-Jato investigations, but since each of these politicians maintain the
prerogative of being tried before the higher courts, neither of their cases had
a celerity that could match the cases moving against Lula. This meant that
either of the PSDB pre-candidates were in a much better judicial position
than Lula – although they themselves were increasingly hampered in polls
by allegations of corruption.
Even though the concept of lawfare, in its anthropological formulation,
is perhaps more oriented toward the exclusion and coercion of social groups
in post-colonial societies, it is also well suited as a sensitizing concept in
the study of corruption scandals. As with the flows of information and leaks
studied in the previous chapter, the selective application of accountability
and legal measures in response to corruption denunciations is always of
potential utility to some political actor. The cases of judicial intervention
and non-intervention in the various case studies of this chapter demonstrate
that law, even in a context with a seemingly abundant web of accountability
institutions, is a malleable thing. Despite claiming to serve the universally
legitimate combat against corruption, trials against corrupt politicians can
have perverse effects – and may even be manipulated for political
advantage.

Conclusion
The many dramatic judicial interventions into Brazilian politics and the
equally spectacular lapses of justice have created a democratic, if not legal,
impasse: When the law is executed only partially, and when this furthers
political controversies while neglecting equality before the law, then that
use of legal actions distorts the political field and undermines democracy.
During the Lava-Jato scandal, neither courts, congressional committees,
nor judicial mechanisms could be relied upon to regularly produce
accountability in an equilibrated manner, to resolve policy questions, or to
promote a stable democratic State.
The existing literature on the flaws and merits of Brazilian accountability
mechanisms has tended to assess such mechanisms, triggered by corruption
scandals, based on whether justice was served or impunity prevailed. With
the insights of this chapter, I draw attention to the systemic effects of
accountability, sanctions, and interventions, because the mechanisms of
checks and balances and accountability should not and cannot simply be
measured by the results at the individual level.
The systemic ramifications of judicial interventions and legal struggles
do not only influence jurisprudence and the day-to-day media agenda. As
we have seen in the TSE case and Lula’s cases, state-given laws, and the
coercive power that resides as the ultimate boundary of such laws, can be
deployed not only to coerce and suppress physically, but also to make or
break political narratives. In this sense, “lawfare” is perhaps even more
interesting as an analytical concept than the notions of judicialization of
politics or politicization of the judiciary, because the concept of lawfare
points to the discursive power and narrative capacity wielded by specific
actors, through legal instruments and representations of trials and courts.
We might do well to keep in mind that the war, combat, or crusade
against corruption are powerful metaphors, and the use of such metaphors
legitimizes precisely the (symbolically or physically) coercive and violent
manifestations of state power. In Brazil, parading politicians and
businessmen in handcuffs from police transport vehicles into prison cells
has enormous symbolic power, and the coverage of Lula being escorted
from his home to testify in March 2016 was perhaps the moment most
galvanizing to the impeachment of Dilma Rousseff. The incident clearly
showed Lula as the passive object of law, rather than a politically vital
subject. Since then, public opinion polls have been asking for the
population’s views on the imprisonment of Lula. Thus, the coercive threat
of the State is discursively present and maintained on the media agenda –
even if it never manifests into a concrete incarceration of the ex-president.

Notes
1 The vote in the Senate after the hearing of the nominated PGR is normally considered merely
symbolic. Upon choosing Raquel Dodge, some discussion arose as to whether the president could
disregard the order of the shortlist provided by the prosecutor’s corporation. Dodge was only the
second most-voted candidate on the list, behind the electoral prosecutor-general, Nicolao Dino.
Nicolao Dino had, in the TSE trial a few months earlier, requested the cancellation of Temer’s
presidency and, moreover, is the brother of the communist governor of the state of Maranhão,
Flávio Dino.
2 Supreme Justice Gilmar Mendes has repeatedly echoed the critique raised by Cunha and Temer,
stating that prosecuting processes led by Janot lacked legitimacy, that Janot made a “complete
mess” of the investigation, made a “sorry figure,” and that the plea bargain negotiated in 2017
with the Batista brothers was the “greatest tragedy in the history of the PGR.”
3 A number of crucial instances of judicial review may be mentioned here: Affirmative action in
the public universities (the so-called racial Quota deemed constitutional in 2010), the 2011
affirmation of the constitutionality of gay marriage, and the ruling of 2016 where the STF
acquitted doctors practicing illegal abortion, potentially paving the way for legalization of
abortion in Brazil are three important examples. Stem cells, anti-Semitism, amnesty laws, and
strike rights have also been the object of judicial review in the STF; see Koerner 2013: 699 for
more examples.
4 In January 2018, only eight Congressmen were charged before STF: Gleisi Hoffmann and Vander
Loubet of PT, Aníbal Gomes and Valdir Raupp of PMDB, Fernando Collor de Mello of PTC, and
Nelson Meurer, Luiz Fernando Faria, and José Otávio Germano of PP.
5 The Supreme Tribunal of Justice (STJ) is the supreme appellate court dealing with non-
constitutional issues. The court, composed of 33 judges, works to ensure homogeneous
jurisprudence across the Brazilian courts, and the court also judges governors (in accordance with
a special case of foro privilegiado).
6 The final step in the testimony phase saw both the imprisoned João Santana and his wife Monica
Moura testifying to the TSE in April 2017. Nicolau Dino, the electoral prosecutor, concluded that
the Rousseff-Temer ticket had used R$112 million illicitly, stemming from the corruption of state
companies and channeled into the campaign via Odebrecht payments to João Santana (R$20
million), payments to three coalition partner parties (PROS, PCdoB, and PRB, R$21 million in
total) and 4 million paid as compensation for the broadcasting slot of PDT during the race.
Furthermore, R$17 million had been laundered by Odebrecht’s subsidiary Petrópolis, and 50
million had been paid as delayed kickbacks for a tax exemption provisional decree (MP470 of
2009) which favored Odebrecht. The latter amount, according to Marcelo Odebrecht, had not
been paid in full during the 2010 campaign, and had only been used in the 2014 campaign. With
these testimonies and the final report of the prosecutor, the process of sentencing began in May
2016, but was interrupted as the court accepted the defendants’ plea for more time and restarted
on June 6, 2017. By then, Temer’s presidency had become embroiled in yet another mega-leak,
threatening the stability of his government (see Chapter 3).
7 Zeze Perrella is notorious in Brazilian politics since the incident in which a helicopter owned by
his company, packed with 455 kilograms of cocaine, was seized by the federal police. Perrella
 was not charged in the case.
8 The eight investigations targeting Neves in the STF can be reviewed at the Court’s site,
www.stf.jus.br, and found via search for the inquiries numbered INQ4246, INQ4244, INQ4444,
INQ4414, INQ4423, INQ4436, INQ4392 and INQ4506.
9 MP 471 meant tax reduction incentives for car construction companies in the regions North,
North-East and Center-West. Originally valid only in 2009, the decree was extended till 2015.
The reduction was specifically to be determined from the expected value of the IPI (Imposto
sobre Produtos Industrializados) and then return as reimbursement of two other product taxes, the
PIS and the Cofins. In practice, the bill resulted in tax benefits for two car construction firms
(MMC, linked to Mitsubishi, and CAOA, linked to Ford, Hyundai, and Subaru). Two more
provisional bills, from Rousseff’s government, also looked suspect to the investigation: The MP
512 (decree of 2010) was, like MP471, a tax reduction incentive, targeted at manufacturers of
automobile motors, wheels, and other parts, again determined from the IPI and only for
manufacturers in the regions North, North-East, and Center-West. The MP 627 (decree of 2013)
annulled a temporary tax, the Regime Tributário de Transição (RTT), and changed the tax scheme
for Brazilian multinational companies, as well as establishing an eight-year deadline for the
Brazilian companies based outside of Brazil for paying two other taxes, the IRPJ and CSLL.
10 The Comaroffs were later made aware of the military-legal scholarship use (Comaroff and
Comaroff 2007: 144) but had originally used the term to connote “the systematic effort to exert
control over and/or to coerce political subjects by recourse to the violence inherent in legal
instruments” (ibid.: 145).

References
Abramo, C.W. (2007) “Brazil: A Portrait of Disparities” Brazilian Journalism Research, vol. 3(1),
93–107.
Almeida, R. de (2016) A Sombra do Poder. Os Bastidores da Crise Que Derrubou Dilma Rousseff.
Rio de Janeiro: Editora LeYa.
Anders, G. and Nuijten, M. (eds.) (2009) Corruption and the Secret of Law: A Legal Anthropological
Perspective. Aldershot: Ashgate.
Arantes, R. (2002) Ministério Público e Política no Brasil. São Paulo: Sumaré/Fabesp.
Arantes, R. (2011) “The Federal Police and the Ministério Público.” In T. Power and M. Taylor
(eds.), Corruption and Democracy in Brazil (184–217). South Bend, IN: University of Notre
Dame Press.
Araújo, C.M., Costa, S.F., and Fittipaldi, I. (2016) “Boa noite, e boa sorte: determinantes da demissão
de ministros envolvidos em escândalos de corrupção no primeiro governo Dilma Rousseff”
Opinião Pública, vol. 22(1), 93–117.
Avritzer, L. (2016) Impasses da democracia no Brasil. Rio de Janeiro: Civilização Brasileira.
Avritzer, L. and Marona, M.C. (2014) “Judicialização da política no Brasil: ver além do
constitucionalismo liberal para ver melhor” Revista Brasileira de Ciência Política, n. 15, 69–94.
Barroso, R. (2017) Questão de Ordem na Ação Penal 937. Brasília: Supremo Tribunal Federal.
Retrieved from www.stf.jus.br/arquivo/cms/noticiaNoticiaStf/anexo/AP937QOrelator.pdf.
Benjamin, H. (2017) Relatorio. Brasília: Tribunal Superior Eleitoral. Retrieved from
www.tse.jus.br/eleitor-e-eleicoes/eleicoes/eleicoes-anteriores/eleicoes-2014/prestacao-de-contas-
eleicoes-2014/acao-de-investigacao-judicial-eleitoral-no-1943-58.
Bergamo, M. (2016) “Lula deve aceitar convite de Dilma para ser ministro” Folha de S. Paulo, p.
A5, March 14. Retrieved from
 www1.folha.uol.com.br/colunas/monicabergamo/2016/03/1749879-lula-deve-aceitar-convite-de-
dilma-para-ser-ministro.shtml.
Boydstun, A. (2013) Making the News. Politics, the Media, and Agenda Setting. Chicago, IL:
University of Chicago Press.
Brooks, B. (2017) “Janot diz que Dodge e Segóvia estão desacelerando investigações de corrupção”
Reuters Brasil, published on December 10. Retrieved from
https://br.reuters.com/article/topNews/idBRKBN1DV52Q-OBRTP.
Bulla, B. and Moura, R.M. (2017) “Supremo indica maioria a favor de restrição do foro” Estado de
S. Paulo, p. A4, November 21.
Campos, P.P. de (2017) “Juízes são mais duros que Moro ao revisarem penas em 2ª instância” Veja,
edition #2538, published on July 12, 2017. Retrieved from http://veja.abril.com.br/politica/juizes-
sao-mais-duros-que-moro-ao-revisarem-penas-em-2a-instancia/.
Cardoso, G. and Fatima, B. di (2013) “Movimento em rede e protestos no Brasil. Qual gigante
acordou?” Dossiê Mídia, Intelectuais e Política, vol. 16(2), 143–176.
Carlson, J. and Yeomans, N. (1975) “Whither Goeth the Law.” In M. Smith and D. Crossley (eds.),
The Way Out – Radical Alternatives in Australia. Melbourne: Lansdowne Press.
Comaroff, J. and Comaroff, J. (eds.) (2006) Law and Disorder in the Postcolony. Chicago, IL:
University of Chicago Press.
Comaroff, J. and Comaroff, J. (2007) “Law and Disorder in the Postcolony” (abridged version)
Social Anthropology vol. 15(2), 133–152.
Cotta, L.C. (2008) Adhemar de Barros (1901–1969): A origem do “rouba, mas faz.” Master’s
Thesis. São Paulo: Universidade de São Paulo.
Coura, K. and Marques, H. (2015) “Os favores da empreteira” Veja, edition #2423, April 29.
Couso, J., Huneeus, A., and Sieder, R. (eds.) (2010) Cultures of Legality. Judicialization and
Political Activism in Latin America. Cambridge: Cambridge University Press.
Dallagnol, D. and Martelo, O. (2016) “Lava Jato, de onde veio e para onde vamos” Folha de S.
Paulo, October 30. Retrieved from www1.folha.uol.com.br/opiniao/2016/10/1827555-lava-jato-
de-onde-veio-e-para-onde-vamos.shtml.
Das, V. and Poole, D. (eds.) (2004) Anthropology in the Margins of the State. Santa Fe, NM: School
of American Research Press.
Datafolha (2017) Intenção de voto presidente – setembro de 2017, PO813939. São Paulo: Datafolha.
Dunlap Jr., C.J. (2001) “Law and Military Interventions: Preserving Humanitarian Values in 21st
Century Conflicts.” Paper prepared for the Carr Centre for Human Rights Policy, November 29,
2001. Retrieved from http://people.duke.edu/~pfeaver/dunlap.pdf.
Edelman, L. (2016) Working Law: Courts, Corporation, and Symbolic Civil Rights. Chicago, IL:
University of Chicago Press.
Falcão, J., Hartmann, I.A., and Chaves, V. (2014) III Relatório Supremo em Números: o Supremo e o
tempo. Rio de Janeiro: Escola de Direito do Rio de Janeiro da Fundação Getulio Vargas.
Falcão, J., Arguelhes, D.W., and Recondo, F. (eds.) (2016) O Supremo em 2015. Rio de Janeiro:
Escola de Direito do Rio de Janeiro da Fundação Getulio Vargas.
Faria, T. and Peron, I. (2017) “Reforma alivia dívidas dos partidos até com Lava Jato” Estado de S.
Paulo, October 6, 2017. Retrieved from http://politica.estadao.com.br/noticias/geral,reforma-
alivia-dividas-dos-partidos-ate-com-lava-jato,70002029781.
Ferraz, C. and Finan, F. (2008) “Exposing Corrupt Politicians: The Effects of Brazil’s Publicly
Released Audits on Electoral Outcomes: Evidence from Audit Reports” Quarterly Journal of
Economics, vol. 132(2), 703–745.
Ferreira, F. and Rocha, G. (2015) “Polícia investiga sítio usado por Lula” Folha de S. Paulo,
November 17, 2015, p. A4. Retrieved from www1.folha.uol.com.br/poder/2015/11/1707354-
policia-investiga-sitio-usado-por-lula.shtml.
Greco, L. and Leite, A. (2014) “Die ‘Rezeption’ der Tat- und Organisationsherrschaft im
brasilianischen Wirt-Schaftsstrafrecht” Zeitschrift für Internationale Strafrechtsdogmatik, vol. 6,
285–291.
Guimarães, J. (2016) “Midiatização Instrumental versus Publicidade Democrática na Operação Lava
Jato.” In J. Guimarães, M. Andrade Cattoni de Oliveira, M. Mont’Alverne Barreto Lima, and N.
de Menezes Albuquerque (eds.), Risco e Futuro da Democracia Brasileira: Direito e Política no
Brasil Contemporâneo (19–32). São Paulo: Fundação Perseu Abramo.
Gupta, A. (1995) “Blurred Boundaries: The Discourse of Corruption” American Ethnologist, vol.
22(2), 375–402.
Helmke, G. (2017) Institutions on the Edge: The Origins and Consequences of Institutional
Instability in Latin America. Cambridge: Cambridge University Press.
Hirschl, R. (2004) Towards Juristocracy: The Origins and Consequences of the New
Constitutionalism. Cambridge, MA: Harvard University Press.
Hirschl, R. (2008) “The Judicialization of Mega-Politics and the Rise of Political Courts” Annual
Review of Political Science, vol. 11(1), 93–118.
Kahn, J. (2017) “Geographies of Discretion and the Jurisdictional Imagination” Political and Legal
Anthropology Review, vol. 40(1), 5–27.
Koerner, A. (2013) “Judiciário e moralização da política: três reflexões sobre as tendências recentes
no Brasil” Pensar, vol. 18(3), 681–711.
Lima, V.A. da (2004) “Sete teses sobre mídia e política no Brasil” Revista USP, São Paulo, no. 61,
48–57.
Lima, V.A. da (2009) “Revisitando as sete teses sobre mídia e política no Brasil” in Comunicação &
Sociedade, vol. 30(51), 13–37.
Llanos, M. and Marsteintredet, L. (eds.) (2010) Presidential Breakdowns in Latin America: Causes
and Outcomes of Executive Instability in Developing Democracies. London: Palgrave Macmillan.
McPherson, C.M. and Sauder, M. (2013) “Logics in Action: Managing Institutional Complexity in a
Drug Court” Administrative Science Quarterly, vol. 58(29), 165–196.
Macaulay, F. (2011) “Federalism and State Criminal Justice Systems.” In T. Power and M. Taylor
(eds.), Corruption and Democracy in Brazil. The Struggle for Accountability (218–249). South
Bend, IN: University of Notre Dame Press.
Macedo, I. (2017a) “Os senadores sob suspeita e o que eles dizem sobre as acusações em análise no
STF” Revista Congresso Em Foco, vol. 26, July 18, 2017. Retrieved from
http://congressoemfoco.uol.com.br/noticias/os-senadores-sob-suspeita-e-o-que-eles-dizem-sobre-
as-acusacoes-em-analise-no-stf/.
Macedo, I. (2017b) “Quem são e o que dizem os 238 deputados e senadores investigados no STF”
Revista Congresso em Foco, vol. 26, July 25. Retrieved from
http://congressoemfoco.uol.com.br/noticias/quem-sao-e-o-que-dizem-os-238-deputados-e-
senadores-investigados-no-stf/.
Marchetti, V. (2008) “Governança Eleitoral: O Modelo Brasileiro de Justiça Eleitoral” Dados –
Revista de Ciências Sociais, vol. 51(4), 865–893.
Mariani, D., Lupion, B., and Almeida, R. (2017) “Qual é o grau de discordância e concordância entre
os ministros do Supremo” Nexo Jornal, published on March 21, 2017. Retrieved from
www.nexojornal.com.br/especial/2017/03/21/Como-os-ministros-do-Supremo-se-aproximam-ou-
se-distanciam-entre-si-de-acordo-com-suas-decisões.
Mattoso, C., Casado, L., and Uribe, G. (2017) “Julgamento de Dilma-Temer no TSE já começa com
disputa por um voto” Folha de São Paulo, March 29, 2017. Retrieved from
www1.folha.uol.com.br/poder/2017/03/1870655-julgamento-de-dilma-temer-no-tse-ja-comeca-
com-disputa-por-um-voto.shtml.
Mello, C. de (2016) Voto. Brasília: Supremo Tribunal Federal. Published on December 7, 2016.
Retrieved from www.stf.jus.br/arquivo/cms/noticiaNoticiaStf/anexo/referendoADPF402CM.pdf.
Michener, G. and Pereira, C. (2016) “A Great Leap Forward for Democracy and the Rule of Law?
Brazil’s Mensalão Trial” Journal of Latin American Studies, vol. 48, 477–507.
Ministério de Planejamento, Orçamento e Gestão (2015) “Tabela 1 – poderes executivo, legislativo e
judiciário.” Retrieved from
www.planejamento.gov.br/secretarias/upload/Arquivos/servidor/publicacoes/dados_ldo/anexo_i-
tab1-quant_fisico_pessoal-pexec.pdf.
Moro, S. (2004) “Considerações sobre a Operação Mani Pulite” Revista CEJ, vol. 8(26), 56–62.
Netto, V. (2016) Lava-Jato: O juiz Sergio Moro e os bastidores da operação que abalou o Brasil. Rio
de Janeiro: Editora Sextante.
Oliveira, G. (2014) “Cooperativa entrega triplex de Lula, mas três mil ainda esperam imóvel” O
Globo, December 7, 2014. Retrieved from https://oglobo.globo.com/brasil/cooperativa-entrega-
triplex-de-lula-mas-tres-mil-ainda-esperam-imovel-14761809.
Pereira, C., Melo, M.A., and Figueiredo, C. (2009) “The Corruption-Enhancing Role of Re-Election
Incentives? Counterintuitive Evidence from Brazil’s Audit Reports” Political Research Quarterly,
vol. 62(4), 731–744.
Pereira, C., Rennó, L., and Samuels, D. (2011) “Corruption, Campaign Finance, and Reelection.” In
T. Power and M. Taylor (eds.), Corruption and Democracy in Brazil (80–102). South Bend, IN:
University of Notre Dame Press.
Pereira, C. and Melo, M.A. (2015) “Reelecting Corrupt Incumbents in Exchange for Public Goods:
Rouba mas faz in Brazil” Latin American Research Review, vol. 50(4), 88–115.
Pereira, C., Bertholini, F., and Raile, E. (2016) “All the President’s Men and Women: Coalition
Management Strategies and Governing Costs in a Multiparty Presidency” Presidential Studies
Quarterly, vol. 46(3), 550–568.
Peruzzotti, E. (2006) “Media Scandals and Social Accountability. Assessing the Role of Scandals in
Argentina.” In E. Peruzzotti and C. Smulovitz (eds.), Enforcing the Rule of Law. Social
Accountability in the New Latin American Democracies (249–271). Pittsburgh, PA: University of
Pittsburgh Press.
Peruzzotti, E. and Smulovitz, C. (eds.) (2006) Enforcing the Rule of Law. Social Accountability in the
New Latin American Democracies. Pittsburgh, PA: University of Pittsburgh Press.
Pontes, G. (2016) “Exclusivo: conheça os candidatos que caíram na Lei da Ficha Limpa” Congresso
em Foco, September 30. Retrieved from http://congressoemfoco.uol.c.
Power, T. and Taylor, M. (eds.) (2011) Corruption and Democracy in Brazil. South Bend, IN:
University of Notre Dame Press.
Polícia Federal (2017) Estatística de Operações. Retrieved from
www.pf.gov.br/imprensa/estatistica/operacoes.
Presidência da República (2010) Lei Complementar 135 De 4 De Junho De 2010. Retrieved from
www.planalto.gov.br/ccivil_03/Leis/LCP/Lcp135.htm.
Raile, E., Pereira, C., and Power, T. (2011) “The Executive Toolbox: Building Legislative Support in
a Multiparty Presidential Regime” Political Research Quarterly vol. 64(2), 323–364.
Rennó, L. (2011) “Corruption and Voting.” In T. Power and M. Taylor (eds.), Corruption and
Democracy in Brazil (56–69). South Bend, IN: University of Notre Dame Press.
Speck, B. (2011) “Auditing Institutions.” In T. Power and M. Taylor (eds.), Corruption and
Democracy in Brazil (127–162). South Bend, IN: University of Notre Dame Press.
Supremo Tribunal Federal (2017) “Decisao de julgamento, ADI 5526.” Retrieved from
www.stf.jus.br/portal/processo/verProcessoAndamento.asp?
numero=5526&classe=ADI&origem=AP&recurso=0&tipoJulgamento=M.
Tate, N. and Vallinder, T. (1995) The Global Expansion of Judicial Power. New York: New York
University Press.
Taylor, M. (2011) “The Federal Judiciary and the Electoral Courts.” In T. Power and M. Taylor
(eds.), Corruption and Democracy in Brazil (162–183). South Bend, IN: University of Notre
 Dame Press.
Thornton, P. and Ocasio, W. (2008) “Institutional Logics.” In R. Greenwood, C. Oliver, K. Sahlin,
and R. Suddaby (eds.), The SAGE Handbook of Organizational Institutionalism (99–129). Los
Angeles, CA: Sage.
Thornton, P., Ocasio, W., and Lounsbury, M. (2012) The Institutional Logics Perspective. A New
Approach to Culture, Structure and Process. Oxford: Oxford University Press.
Tribunal Superior Eleitoral (2017) “Informações gerais sobre a Ação de Investigação Judicial
Eleitoral nº 194358.” Retrieved from www.tse.jus.br/imprensa/noticias-
tse/2017/Marco/informacoes-gerais-sobre-a-acao-de-investigacao-judicial-eleitoral-no-1943-58.
Tsebelis, G. (2002) Veto Players How Political Institutions Work, Princeton, NJ: Princeton University
Press.
Waters, C. (2009) “Beyond Lawfare: Juridical Oversight of Western Militaries” Alberta Law Review,
vol. 46(4), 885–911.
Werneck Vianna, L., Carvalho, M.R., Palácios, M. de, and Burgos, M. (1999) A judicialização da
política e das relações sociais no Brasil. Rio de Janeiro: Revan.
Werner, W. (2010) “The Curious Career of Lawfare” 43 Case Western Reserve Journal of
International Law, vol. 43, 61–72.
Wiarda, H.J. (ed.) (2004) Authoritarianism and Corporatism in Latin America Revisited. Gainesville:
University of Florida Press.
Wink, G. (2017) “Judicialização da política ou politização da justiça? Uma Análise Crítica do
Discurso do Libelo Acusatório contra a presidenta Dilma Rousseff no processo de impeachment”
Moara, vol. 47, 152–177.
Winters, M. and Weitz-Shapiro, R. (2013) “Lacking Information or Condoning Corruption: When Do
Voters Support Corrupt Politicians?” Journal of Comparative Politics, vol. 45(4), 418–436.
Zanin Martins, C., Zanin Martins, V.T., and Rafael, V. (eds.) (2017) O caso Lula – a luta pela
afirmação dos direitos fundamentais no Brasil São Paulo: Contracorrente.
5 The Informational Cascades of Impeachment
and the Lava-Jato Scandal

 
The previous chapters have demonstrated how disclosure of corruption
emerges intermittently as leaks – when information is provided to
journalists of the larger media outlets by the public prosecutors, by lawyers
or even by the defendants – and how, subsequently, that disclosure is
processed and instrumentalized throughout the judicial arenas of Brazil.
Between the sudden leaks and the continuous grinding of the legal actions,
the attention of the main Brazilian news outlets has been synced to the anti-
corruption agenda, dancing to the tune of the leaky, legal processes. Even
so, the various cases have not reached their end in an equal pace; many
political actors under scrutiny have largely escaped accountability, and only
a few cases have continuously been in the limelight. In this chapter, I will
explore the distribution of information flows and attention. Whereas the
previous chapters have given qualitative, historical and political
explanations for the outcomes of the disclosure of corruption, both in the
form of leaks and continuous coverage of legal action, this chapter works
quantitatively, using the concept of cascades to understand the impact of
disclosed information in the Lava-Jato scandal. The chapter contributes to
the growing body of research on infostorms by developing an approach to
tracing such information phenomena in the sphere of media.
Snowballing, avalanche-like informational media phenomena such as
moral panics, media storms, and media hypes have been theorized before,
but mostly as separate concepts within media studies (Thompson 2000,
Tumber and Waisbord 2004, Welch et al. 2002, Wien and Elmelund-
Præstekær 2009). In a certain sense, a mechanism similar to media hypes
was also implied in more general media theories about agenda-setting.
Agenda-setting theory predicts the relation between media coverage and the
media audiences’ perception of salience of issues (McCombs and Shaw
1972, McCombs 2005). Agenda-setting theory could be used to analyze the
skew of salience transfer effects in the rapid surges of coverage that
characterize scandals. This has rarely been explored (but see Boydstun 2013
and Walgrave et al. 2017), however, and such surges of media attention have
only recently been linked to a general theory of information flows through
the notion of informational cascades in social contexts (Hendricks and
Hansen 2016).
Earlier concepts explaining the power of media scandals have sometimes
been linked theoretically to various fields of sociology, but in distinct ways.
Where Thompson (2000) discussed the power of scandals in political fields
using a Bourdieusian capital approach, the moral panics described by Welch
and colleagues were conceived as a parallel to Durkheim’s functional
understanding of ritual, because surges of media panic acted to expel folk
devils and confirm societal order (Welch et al. 2002). In a third strand of
theory, the concept of media hype coined by Vasterman was connected
loosely to framing theory. Several of these authors have pointed out that
media-instilled panics and news waves entail social amplification of signals
disproportionate to real frequency of the problem in question (e.g.,
Vasterman 2005: 516, Welch et al. 2002: 15). This conceptualization of
news as signals in public spheres is a crucial component of the
informational cascade framework as well.
In the following, I will explain the concept of informational cascades and
operationalize this concept through a content analysis research design,
drawing on agenda-setting theory. By using the concept of informational
cascades, rather than ad-hoc theories defined by characteristics specific to
certain media situations (such as hypes, panics, or storms), media theory
may integrate these theories and link them to the general mechanisms of
information flows and decision making. This integration requires an
explication of the relationship between media systems and the collective
processing of information in public spheres. Such an account may
contribute to our understanding of chain reactions emerging between media,
political actors, and public opinion; chain reactions that can be particularly
nefarious in media environments rife with so-called “viral” news, “fake
news,” or “post-factual news.” However, what is central here is neither
“fake” news, although disinformation will be touched upon, nor the usual
question of agenda-setting studies – for example, to ascertain how much of
an impact media coverage had upon Brazilian audiences in the course of the
Lava-Jato investigations. Such a study of the level of reception would
require representative samples of both the population and the media
coverage. Instead, this chapter will focus only on the information flows that
emerged in the leading dailies during the Lava-Jato scandal and the build-up
of salience in relation to particular political actors.
The analysis of this chapter helps explain the conundrum of the Lava-
Jato scandal: How could the majority of the politicians implicated in the
scandal escape social accountability and form a new government while
embroiled in the very same corruption investigation that undermined the
former government? Most theories of scandal affirm that the media holds
great power during scandals, and the agenda-setting effects of media
scrutiny are theoretically linked to this potential inherent in scandals: All
else being equal, we would expect that more coverage of a political action
(which is being characterized as transgressive and scandalous) would
increase the likelihood of some kind of political response – resignations,
public apologies, etc. While this expectation undoubtedly rings true in some
political systems, it does not capture what happened in the Lava-Jato case.
In fact, this corruption probe might, at a glance, seem paradoxically potent
and impotent: The media reported ceaselessly on the case for three years,
but while scores of businessmen were brought to trial and sentenced, with
ample evidence of political corruption, very few politicians have been
charged before the STF at the time of writing (although scores of
congressmen are being investigated). Only two mandates have been
annulled, and the political parties are completely unwilling to expel
members who are denounced or sentenced for corruption. This flies in the
face of the theoretical and common-sense expectations of political scandals,
since the ongoing public shock of disclosed corruption has hardly affected
the vast majority of the congressmen. The paradox of enormous media
attention and limited accountability, I argue, can be explained partly as a
problem of the informational cascade in the public spheres of Brazil.
In the following, I will consider theoretically how information about
corruption can be shaped and passed on in cascades, expanding the
theoretical apparatus presented in Chapter 1. Second, I operationalize the
concept of cascades through a sampling strategy of print and web media,
exploring the frequency, salience, and sequence of news on the various
corruption cases and the impeachment proceedings. The data collected with
this strategy provides a proxy for the general distribution of media attention
in the public spheres of Brazil; and the distribution reflects editorial and
journalistic selection processes, which are crucial in setting up an
informational cascade in the media.
The findings of the content analysis demonstrate that the dominant news
outlets’ coverage of this wide-ranging scandal exhibited a centripetal
pattern, leading to massive overexposure of three political actors, while
information about other actors was marginalized and largely disregarded in
the news. Two mechanisms of the media system were crucial: The
overvaluation and the selective recycling of information about corruption.
These mechanisms meant that news linking Lula to the Lava-Jato probe
eclipsed news that dealt with other politicians’ cases of corruption.
Delving into the processes where information was passed from political
and judiciary elites, through mainstream media, and on to the stock markets
of Brazil, I also find evidence that the informational cascade contributed to
creating not just a moral crisis, but subsequently an economic and political
crisis. The Lava-Jato scandal can be interpreted as an exemplar case,
magnifying certain features such as skew and explosiveness inherent in
media agenda processes (Boydstun 2013: 208). In the interplay of politics
and media, these features yield information phenomena in the public spheres
akin to financial bubbles (Hendricks and Hansen 2016).

The Concept of Informational Cascades

This chapter builds on the definition of informational cascade given in the
first chapter. To unpack this definition further, I will expand three aspects of
the theoretical framework before operationalizing the concept. First, I will
briefly discuss the definition of information in the framework I am
proposing, including the definitions of misinformation, disinformation, and
information quality. Second, since this book works with media texts, it is
necessary to link the abstract definitions of information to theories that deal
with media objects; in particular, the theory of framing. Framing theory
provides insights concerning the packaging of information in concrete
media texts, which is an under-developed part of the informational cascade
literature (but see Hendricks and Hansen 2016: 85ff.). Closing these
preliminary remarks, I discuss how intertextuality is used in media texts on
scandals to reproduce and extend certain journalistic frames. Intertextuality
thus becomes a key aspect of journalistic narration in scandals, present in
various clues (visual, textual, or metatextual, i.e., hyperlinks). Through such
clues, audiences are invited to connect the dots between sequences of media
texts, and cued intertextuality may thus be an important factor for the
emergence of cascades and pluralistic ignorance in public spheres. The three
theoretical considerations of corruption scandals and informational
cascades, framing, and intertextuality provide the background for the
sampling strategy and operationalization of the concept of informational
cascade deployed.
As described in the first chapter, conceptually, informational cascades
consist in rationally acting actors, who are sequentially passing along
information through their public actions. The actors cannot derive private
knowledge possessed by other actors, but nonetheless, all observing actors
take the actions to be significant down the sequence stream. When making
their respective choices, these actors might then base their actions on false
assumptions about the private knowledge of others. A “cascade starts and
new information stops accumulating” when the “preponderance of evidence
supports one action” and actors repeatedly reproduce one public signal
(Bikhchandani et al. 1998: 158).
Hendricks and Hansen warn that such sequences might bias the choices
of an electorate or a group of decision-makers a great deal. They offer
several illustrations of everyday situations where cascades harm our
decision making. In one such illustration, they identify informational
cascades in the environment of airports: Cascades occur when we are
running frantically around an airport terminal, trying to find the right gate
after only hearing half the announcement about a gate change. Observing
other passengers with a calm, directed gait and tickets that are seemingly
identical to ours, we decide to follow these apparently well-informed
travelers. “They seem to know where to go, so I should follow them” goes
the internal argument, which is potentially very wrong. This kind of
thinking, they argue, occurs every so often in public spaces, or, indeed, in
the formation of public opinion (Hendricks and Hansen 2016: 61, 245).
Hendricks and Hansen point out how informational cascades might be
harmful to democratic societies because cascades tempt citizens – voters as
well as decision-makers – to make their choices based on the public
informational sequence rather than the quality of the information.
Similar concerns were foundational for modern media studies. The
simple model of media introducing ideas into the minds of the audience
(like a highly effective hypodermic needle) was however debunked by
Lazarsfeld and colleagues, who demonstrated more complex information
flows: People adopt attitudes to political issues by first looking to local
opinion leaders (which could be other people affiliated with a party, or
simply people regarded as knowledgeable in political matters), rather than
solely taking cues from media (Lazarsfeld et al. 1944). Individual actions,
then, are based on the value invested in the signal sent by these opinion
leaders, even though individuals will not know every aspect of the opinion
leaders’ internal deliberations and possible vested interests. In the limited-
effects model proposed by Lazarsfeld and colleagues, the public signal of
media (in campaign coverage) was attenuated or overridden by the signals
of people’s social environment. In both hypodermic-needle and limited-
effect scenarios, however, individual decision-making was considered to be
influenced by other social agents, sometimes in spite of private knowledge
or convictions. Herein lie the roots of the problem of information, public
knowledge, and signaling.
In Chapter 3, we saw how information is leaked or fed to journalists
(usually those working for prestigious media outlets). These media outlets,
in turn, disseminate the information, once journalists and editors have
selected and packaged the information in specific framings (Entman 1993)
and through specific formats and scripts (Altheide and Snow 1979: 10, van
Dijk 1988: 34). The exact selection of information that will eventually reach
audiences is limited by the gatekeepers of media and the sub-politics of the
sources (Waisbord 2000: 93). Researchers should not naively assume that
members of the audience cannot reinterpret the disseminated information in
heterogeneous ways (Biroli and Miguel 2013: 81), framing problems in
other ways than the media framed them (see below): Signals of corruption
could be disregarded or mistrusted by audience members, especially if one
has private knowledge that contradicts the public signal of the media.
However, in national corruption scandals, private knowledge about what
goes on in the backstage of the political arena is naturally extremely rare.
Even Brazilian congressmen, in fact, must sometimes rely on the rumors
reported by the small news columns of O Globo, Folha, and Estado.
Congressmen, unlike most of us, may however have a much better chance
of understanding the implications of such rumors. Like ordinary citizens,
even such insiders in political circles still stand at a fourth or fifth remove in
the chain of information. Denunciations of corruption, for instance,
conceptually look like this in a simple transmission model, ignoring the
potential reinterpretations occurring downstream (see Figure 5.1 above).

Figure 5.1 The chain of information in a denunciation of corruption.

The figure illustrates, in a simplified manner, how information is passed

on to audiences via media actors. As described in Chapter 3, investigative
authorities (MPF and PF) are not the only ones who can pass on information
to journalists; defendants and their lawyers are just as capable of delivering
information to journalists if they judge it to be beneficial to their case.
Judicial actors in courts could also pass on information to journalists, but in
the cases proliferating from the Lava-Jato investigations, information about
probes normally comes to public attention months before the denunciations
and documents are actually handed over to clerks of the courts. The notable
exception to this part of the information flow is the judges. Judges with
jurisdiction of a case, such as Sérgio Moro in the Lava-Jato case, see some
elements of cases before they are investigated, because investigative
measures such as arrests, telephone taps, or inquiries about telephone and
bank registries need a judge’s authorization, based on some manner of
description of the circumstances of the case. Whistle-blowing can also
happen outside of criminal investigations (which was the case in the
Mensalão scandal, but rarely occurred in the course of the Lava-Jato case).
In any case, the information reaches the pages of political news, or the
front pages, or even the daily news broadcasts. At this point, a bandwagon
effect between media outlets might come into play, and when various media
outlets repeat the same signal over and over again, an informational cascade
might form. The bandwagon effect starts with the practical and epistemic
job of the individual journalist, who, in order “to flesh out any one supposed
fact … amasses a host of supposed facts that, when taken together, present
themselves as both individually and collectively self-validating” (Tuchman
1978: 86). This means that many assumptions are continuously reproduced
in news, and this recurrently validates the original assumption. The
hegemonic interpretation of national politics, for example, could be
reproduced over and over by journalists because this will present the
shortest route to validating and interpreting new information culled from
sources. The self-validation also occurs at the institutional level, where

[m]edia institutional actors rely extensively on each other not only as

sources of reference but also as sources of institutionalized legitimacy,
even though they lack first-hand access to the event. This reflects the
general tendency in contemporary journalism of including the
perspectives of media institutions, commentators, correspondents and
other media professionals in the news coverage as opposed to involving
sources outside the media institutional realm.
(Kristensen and Mortensen 2015: 360)

Thus, when media outlets rely on other media outlets for information, and
build on the same assumptions about issues, a feedback-loop emerges that
reproduces the same signal many times over. This might then build up into a
cascade, as visualized in Figure 5.2.
An informational cascade needs, as mentioned in the quote of
Bikhchandani et al., enough weight to override the private knowledge and
presuppositions. When people have little private knowledge, they confide in
news texts (at least if the media outlet is perceived as trustworthy, which of
course is not always the case). If all (trusted as well as distrusted) media
outlets cover the same topic in the same way, private doubts might be
dispelled. The cascade, or chain-effect, only triggers when there is “a
preponderance of evidence [that] supports one action or the other by just
enough to outweigh one individual’s private signal” (Bikhchandani et al.,
idem). Once one actor decides to act upon that information, this constitutes
another piece of evidence for others to see, because his or her action is also
a signal. So, in a corruption scandal, when more and more people start to
call for the resignation of a given politician, that call seems to be justified
for yet more people, and joining the bandwagon, the call for action
reinforces itself until the proverbial mob with forks and torches
materializes: The bandwagon effect results in a self-reinforcing feedback
loop (Hendricks and Hansen 2016: 258).

Figure 5.2 The informational cascade between media outlets.

 In this chapter, I will attempt to trace exactly this dynamic in the media,
working from a universe of 8,200 news texts on corruption from the two
large media outlets Folha and Estado, as well as 1,335 headlines from the
same outlets plus their competitor O Globo. With this dataset, I can track
how some signals of corruption were repeated over and over in news texts,
while others died out, and furthermore trace how the selected and
disseminated information was carried into public spheres (through opinion
polls and stock market data). Like the general model of a cascade, the
particular informational cascade of the Lava-Jato scandal picked up steam
gradually, until it reached a point where new information was promptly
disregarded, no matter how relevant. Before describing the research design,
the notions of information and information quality needs to be accounted
for. Working from a definition of information, I will also prepare the ground
for a discussion of how corruption is framed in news texts.

Information Definitions
An exceedingly broad range of information definitions has previously been
proposed, deriving from such disparate disciplines as communication
sciences, semantics, anthropology, cybernetics, and physics. In the
following, however, only a few definitions will be discussed. This is done
not just for brevity’s sake, but also in order to grasp the kind of
informational phenomenon most relevant to a contemporary media system.
Between the contributions of Floridi (2002, 2005, 2011) and Fallis (2009,
2011), the field of philosophy of information has inherited a central but
debated distinction from thinkers such as Grice (1957) and Dretske (1981),
concerning the element of veracity in information. For Floridi, information
can only be defined as “well-formed, meaningful and truthful data” (Floridi
2011: 80), while Fallis tries to take contextual factors such as imparting
processes, intentionality and semantic layers into account. Fallis (2011)
objects to this, claiming that untrue information might still convey meaning
that can be verified. Fallis tries to sort different kinds of lies into the
categories of misinformation, disinformation, and information, depending
on the intention behind defective semantic content. Obelitz Søe (2014,
2016) has usefully clarified the conceptual intersection of different kinds of
information discussed by Floridi and Fallis, proposing a framework of four
categories: information (which is verifiably true), disinformation (which is
intentionally false), misinformation (unintentionally false), and information-
as-natural-meaning (in the sense of Grice 1957, in which natural meaning
refers to states of reality that are true independent of any observations and
discursive statements). So, while this latter category cannot include meaning
determined by social conventions, the former three frequently do include
information determined according to social norms, practices, and linguistic
usage.1
Actual political and social agents may very well choose to deteriorate
information quality for specific communicative purposes and political goals,
and so for pragmatic purposes, more than conceptual precision, defective,
derived, and even alethically distorted information is still regarded as
information here. This thrusts the concept of information toward the notion
of framing found in media studies (Entman 1993, Sniderman and Theriault
2004). I return to this move shortly.
A media text (including material broadcast through radio, internet, or
television) can be understood as a signal containing many bits of
information. For example, the introductory sentence or paragraph called a
“lead,” as defined by most journalists, normally establish the who, the what,
the when, the where, and the why of a news item (Carey 1986: 148,
Tuchman 1978: 100). Leads establish several facts, purported to be true. A
report on a statement by a politician, for instance, is a speech act from
which various bits of information are extracted by the reporter. The speech
act itself and the extracted information can be interpreted as one or more
discrete signals, e.g., as evidence, as a denial, or as parts of a testimony.
While all of the informational content of the speech may be intentionally
false (which means it is disinformation, strictly speaking), the fact that the
politician spoke is still verifiable. News in scandals can include disclosed
information of the Gricean, natural-meaning variety (a person caught on
tape, for instance, leaving a physical mark, or an audio-visual mark, or a
signature). More commonly, information about corruption in news texts is
of the convention kind (e.g., allegations about a transgressive and prohibited
action, performed by a public agent or agents, which could not have been
performed unknowingly nor innocuously by that agent). In either case, the
information is framed as corruption intentionally by the journalist through
textual or visual markers. That is, both information of the conventions kind
and information-as-natural-meaning need to be packaged along with other
bits of information to make the transgressive nature of a political action
clear.

Framing of Information in Corruption Scandals

Related to the question of information quality, the concept of framing covers
the way in which information can be formed and then imparted exactly and
truthfully, but partially, in a way to create states of potential misinformation
with the audiences. Partial transmission of information is possible in larger
packages of information, such as the bundling of information we find in
news. Deliberate deterioration of information in order to create
misinformation – what is commonly called spin – is an element of specific
political and journalistic practices of shaping information textually.
Although framing has already been mentioned in the second chapter, I will
briefly revisit the concept of framing and argue why, ultimately, a study of
the various framings of corruption is not a fruitful research design, although
theoretically linked to the topic of information flows.
The academic perspectives on framing range from cognitive to
constructivist approaches across discursive and symbolic environments. In
one formulation, framing

involves selecting a few aspects of a perceived reality and connecting

them together in a narrative that promotes a particular interpretation.
Frames can perform up to four functions: define problems, specify
causes, convey moral assessments, and endorse remedies … Framing
works to shape and alter audience members’ interpretations and
preferences.
(Entman 2010: 391)

Logically, framing also entails de-selecting some information, for

example leaving out other potential causes or remedies; the aspect of de-
selection is also relevant to the aggregate level of news framing, to which I
return below. Some scholars have broadened Entman’s definition, viewing
“frames as structures that draw boundaries, set up categories, define some
ideas as out and others in, and generally operate to snag related ideas in their
net in an active process” (Reese 2007: 150). This means that some media
researchers tend to view frames as discursive formations, while other
researchers (e.g., in the field of journalism studies) would name the
journalistic and editorial activity of selecting and packaging information
“framing.” In this more limited sense, framing is a professional practice
linking events to a repertoire of storyable forms – reporting, say, on politics
as a horse-race or a conflict – using specific metaphors and locating actors
textually within certain scripts (van Dijk 1988).
As argued in the second chapter, the frames utilized by Brazilian
journalists to cover political corruption are highly predictable: Suspicion is
cast upon a political actor or a political action, and the suspicion is the
object of conflict, being contested or affirmed to varying degrees by the
involved actors (such as whistle-blowers, prosecutors, police agents,
political supporters or opponents, and of course the denounced actors
themselves). In the media texts reporting on the Lava-Jato investigations,
various reasons for condemning corruption are brought up: Respect for the
Constitution, in the name of democracy, or the electorate, to save the
Petrobras company, or with reference to morality, economy, or national
industrial development. But no matter who is voicing such concerns in the
news texts, and regardless of the explicit presence of any condemnation of
corrupt acts, the frames of the texts imply political transgression from some
perspective (Damgaard 2015). Whether or not responsibility is clearly
attributed to an identifiable political actor (Iyengar 1990), the transgression
lies at the heart of the frame, as the device that defines the problem, moves
the plot, and makes the news item newsworthy.
Alternative frames may spin off from the simplistic denouncement
narrative as the story develops, but in terms of public signals of corruption
and information quality, the framing of spin-off news hardly matters. At the
intertextual level, spin-off news items still refer to the original narrative of
transgression and denunciation. For this reason, framing is not useful in this
study: When news mentions a corruption case, it can be understood as a
public signal that the accused politician is corrupt, even when competing
frames (Sniderman and Theriault 2004) are present in the same text. This
argument brings me to a final consideration, regarding the ways that the
Lava-Jato scandal grew into a vast discursive behemoth, looming over both
the impeachment proceedings and the various strategies of lawfare
discussed in the previous chapter: The mere reference to the Lava-Jato
probe linked a multitude of disparate texts together intertextually.

Intertextuality
Earlier research has discussed the role of intertextuality in scandals, which
may shed light on the narrative linkages in the coverage of the Lava-Jato
scandal. Elizabeth Bird notes that new information in corruption scandals
connects to earlier disclosure intertextually, like strings on a bead, which
gives an episodic quality to scandals. Scandals take on forms dissimilar to
“the terse, inverted-pyramid, ‘news’ style” and develop through follow-up
stories (Bird 1997: 101). A particular news item in the coverage of a scandal
can be considered as “one element in a mesh of stories,” emerging from a
broader repertoire of cultural narratives, stereotypes, and scripts (ibid.: 104).
That is, the news items are discursively linked. Wortham and Rhodes (2012)
argue that in general, discourse connects across speech events, and linked
speech events are essential to social life. Because of this structuring, events
can be recognized as such, belonging to a set or being distinct from the
background noise of history. The events connected thus constitute
narratives, which “are joined together in chains or trajectories, through
discursive processes that link speech events to each other, such that signs
and individuals move along chains of narrating events that occur in different
spatio-temporal locations” (ibid.: 161).
Brazilian news texts commonly represent the Lava-Jato corruption case
as an ever-expanding episodic narrative – almost like a feuilleton or the
Brazilian soap opera genre telenovela – that engulfs many actors in the
political spectrum. Usually, this was marked by the simple inclusion of a
reference to the Lava-Jato case itself. By referencing the Lava-Jato case, or
one of the parallel investigations of grand political corruption, neither
headlines nor leads of news texts or broadcast news need to include words
like “bribe” or “corruption.” The intertextually established newsworthiness
of the Lava-Jato investigations and high-profiled politicians is what signals
corruption. Even media outlets which are normally highly trusted by
Brazilians to produce quality coverage can utilize certain strategies to imply
corruption without explicitly labelling an action “corrupt.” It is enough that
some bit of information can signify a transgression from some societal
perspective (Damgaard 2015).
Intertextual implicature is a framing strategy that can help journalists
evade libel lawsuits and enhance the news value in certain news items. For
example, information about the rebuilding of the Presidential Palace’s
swimming pool was included as news by Folha in November 2016, because
no contract was encountered, while at that moment, the construction
company Odebrecht was under much scrutiny, as was the ex-president Lula.
The renovation of a public building – indeed, buildings which are part of the
national heritage of Brasília – could thus be cast as possibly corrupt,
although the renovation in all probability will never become the centerpiece
of a trial. However, the signal was one in a range of signals apparently
testifying to the corrupt relationship between the ex-president and the
private company.
In this vein, associating an agent’s action, indicated by disclosed
information, with transgression, is a politically potent signal, even in cases
where the action does not qualify as a legal transgression. The information
about the swimming pool was not misinformation, because it was not
incorrect; rather, we may say that the publicizing of this information is
characteristic of disinformation because it is not newsworthy unless it
testifies to corruption. In other words, the inclusion as a news item in itself
signals that something fishy was going on – otherwise, one could not make
sense of the information as “news.” Perfectly ordinary actions may thus be
bundled in the media along with other information and come to be
associated with transgression or even substituting for evidence of
transgression. If perceived “correctly” (that is, from the view of the
manipulative imparter), members of audiences might then hold false beliefs
about the information – believing some politician to be corrupt – although
no disinformation nor misinformation was disseminated, only circumstantial
information that implied corrupt dealings.
Intertextual implicature, furthermore, has ramifications for the sampling
strategy, because the construction of a corpus of news texts about corruption
need to go beyond criteria of simple key words. It is also necessary to
include texts that report on the companies, businessmen or political actors
involved, where the spin-off relation to a case of political transgression is
merely a relation of allusion or implication. I will turn to this and related
questions of sampling in the following.

Operationalization and Sampling Strategy

In the following, I will discuss a methodology for exploring an
informational cascade in the news. While Hendricks and Hansen (2016)
analyzed the decision-making problems arising from cascades, the problem
of this book is rather the emergence of a cascade and the institutional
configurations leading to such a phenomenon. Other researchers have
addressed similar phenomena recently, and their methods could be adapted
for this purpose: Amber Boydstun, although constructing a study around the
term “media storm” and patterns of media attention, sampled news items of
front pages and coded the news items for issues (Boydstun 2013: 11, 82).
Russell and Waisbord, analyzing so-called “news flashpoints,” or bursts of
coverage, also analyzed frequencies of news items (but only items related to
the Snowden leaks). Both examples dealt with the New York Times, and both
worked with a content analysis methodology (Russell and Waisbord 2017:
860).
By opting, like these two studies, for a research design based on content
analysis, the present study could connect to agenda-setting theory, which in
its original formulation explored the relation of media content and voter
perceptions of societal problems (McCombs and Shaw 1972). McCombs
and Shaw utilized a macro-level content analysis method, coding the overall
thematic content of various news items. Later, agenda-setting scholars tried
to integrate framing as a second level of agenda setting in order to get to a
more detailed and context-sensitive level of analysis (Ghanem 1997,
McCombs 2005). Researchers of information phenomena in public spheres
view agenda setting as “a concept which may be studied using tools from
formal epistemology and network theory to evaluate the differences between
the mobilization of both robust knowledge and ignorant belief
configurations in networks” (Wiewiura and Hendricks 2017: 2). With the
studies by Boydstun, Russell, and Waisbord in mind, I will argue that the
inverse should be equally feasible: Detecting cascade-like informational
phenomena through content analysis and understanding cascades through
agenda-setting theory.
In the following, I therefore undertake a content analysis of the news
items that deal with corruption. The content analysis does not compare the
amount of news on corruption to news dealing with other issues but aims at
the distribution of news items devoted to the plethora of corruption cases in
the political system of Brazil. Comparing the Lava-Jato scandal and
impeachment to other issues in Brazilian public spheres would simply
underscore something obvious, namely that the political situation was the
object of intense media attention.
 The goal here is also not to find agenda-setting effects of the media
coverage at the level of reception, as looking for agenda-setting effects of
the several years of intense media coverage of corruption would not go far
enough: There is ample evidence in various surveys that can be correlated to
the reception of the corruption scandals and the impeachment: During the
sample period, the topic of corruption, for the first time, overtook
unemployment, security, and health as the problem considered most
important for society (Datafolha 2015: 3) – despite being the South
American country with the lowest bribery rates (Pring 2017: 31). Both the
Rousseff and the Temer administrations experienced record-low approval
ratings (Datafolha 2017), and trust in political institutions and parties was
the lowest on the continent (Latinobarómetro 2017: 25–27). None of this, of
course, can be tracked to specific news items or events, since the barrage of
information on corruption dominated the political news throughout 2015
and 2016.
Rather, the first goal here is to map out the evolution of media attention
(operationalized as news items) to the various national corruption cases
(defined below). Second, this content analysis might detect feedback
patterns in the coverage of particular cases, where coverage heaps up,
drawing away resources and attention to other cases. To construct the
content analysis, I will first define a corpus of texts, and argue for the
validity and utility of this particular sampling of the Brazilian media system.

Sampled Media Outlets

Corruption is a topic being dealt with across a totality of interlocking media
discourses, which are fluid and engage in never-ending intertextual
dialogue, and so the analysis of the national media agenda is in principle an
almost infinite endeavor. However, given the extant knowledge of
information distribution and vertical integration in Brazil, summarized in
the second chapter, we know that the media system is quite concentrated
(Cabral 2017, Lima 2004, Matos 2008), and that a few newsrooms drive the
media agenda. In the above-mentioned content analysis research designs,
the authors similarly constructed representative samples of the national
media agenda (e.g., Boydstun 2013: 82). In Boydstun’s study, the outlet
sampled was picked out to be exemplary, representative, and a driver of the
national media agenda. Along the same lines, I will argue that it is
methodologically sound to narrow the research object by sampling from
only three media outlets.
First off, due to the vertical integration of media outlets, there are only a
few newspaper and broadcasting network newsrooms covering the national
political scene. These newsrooms distribute information through their
associated networks of news agencies and regional subsidiaries in radio,
television, and print. Abramo showed that in the mid-2000s, 90 percent of
agency-credited news originated with Estado and Folha (Abramo 2007: 99).
Furthermore, as demonstrated in Chapter 3, the leaks of the Lava-Jato case
largely appear in these two media outlets and in Globo, Veja, and IstoÉ. In
other words, to detect a cascade of information about political corruption,
we must begin with these news outlets.
The prestigious daily papers, in contrast to broadcast news and weekly
magazines, are particular useful news platforms on which to gauge regular
information distribution, because daily newspapers have dedicated (and,
usually, fixed) space and resources to devote to political coverage. The
newspapers report on politics every day, which might not be the case for,
say, the evening news. Daily coverage of political issues gives a sense of the
waxing and waning attention of the media. In this sense, broadcast news
journals and weekly magazines are less reliable, because their selection
criteria for inclusion are constrained in different ways than web and print
media of the prestigious newspapers. TV news are limited to specific time
slots, and will balance political news with other topics, thus giving
incomplete pictures of politics on days where other issues crowd on the
national agenda – sports, international terrorism, natural disasters, etc.
Weekly magazines of Brazil are similarly constrained to a format in which a
certain amount of items deal with political scandals. Across the duration of a
week, daily newspapers may report on a variety of cases, whereas the
political reporting (or scandal coverage, as it were) of the Brazilian weeklies
are normally feature articles of some length with a single topic, often
expanding the front page’s framing of the most notable (or most scandalous)
news item of that week.
In other words, for this part of the study, I leave out broadcast and
weeklies, not because they do not matter, but because their particular
formats filter the coverage of political matters much more tightly than do
the newspaper format. Whereas the agenda of weeklies and televised news,
on any given day, is negotiated between politics and other topics, this is not
the case in the print and web news of the quality newspapers. In aggregating
the landscape of news in the daily papers, we get a better sense of the
relative weight of different corruption cases, including smaller cases that
may eventually grow in importance, and avoid the methodological
complication of media formats with stronger selection pressures, where
political news competes with international news, sports, and other topical
features.
Briefly stated, as a representative sample of the various news media of
Brazil, I have picked out a few central media outlets and organizations
(those that get the scoops and distribute news along their vertical networks
of agencies and subsidiaries). Two sets of sampled data were collected: A
sampling of front pages and a sampling of articles. The daily journals
selected for front page sampling were O Globo, Folha, and Estado de S.
Paulo – the three largest and most prestigious papers of Brazil. The headline
samples provide an index of the salience of each corruption case on the
national media agenda on a daily basis. If a corruption case is mentioned on
one or more front page, there is a chance that it may also reach the news
broadcasts. The article sampling, meanwhile, took both printed and web-
publicized news items and opinion pieces from Estado and Folha, leaving
out O Globo in order to avoid excess data redundancy.
This second sample provides an index of how attention to all cases of
corruption was distributed by including the cases that did not make it to the
front pages. The front-page count is inherently weighted, since only so
many headlines can fit in a front page, while a sampling of all published
articles of print and web face fewer restrictions. The print versions of
quality newspapers, as pointed out above, normally have many pages
dedicated to the coverage of politics, while their online versions face no
restrictions inherent in the format. Thus, these kinds of media give a
measure of the resources and attention given to scandals and investigations
independent of other news of national or international importance.
The sample of articles includes articles published on the newspapers’
online editions. With the increasing internet access rates across Brazil, and
the increasing reliance of web media as sources for news, it would be
unwise not to take online news outlets into account. The print news media
and the regularly scheduled newscasts are no longer the first to arrive at the
scene and report breaking news. Often, newsrooms will present a scoop or a
breaking story online immediately, gradually building up some commentary
and perspective and printing some of it on the following day, along with a
report of the event. When examining disclosure of corruption, it is not
infrequently the online news outlets that come into focus. Beyond that, it
might also be relevant to consider how online news are mediated doubly –
not only as a particular media format in itself, but also as content that may
be more or less shareable on social media. For these reasons, I have also
stored the available social media statistics published on the media outlets
content pages, although such statistics comes with their own methodological
problems.
Let me briefly summarize this section: In addition to the front pages of
the three national newspapers, I opted to collect both web and print versions
of news items on corruption from Estado and Folha in a single SQL
database structure with various metadata and a coding of the issues being
reported. The database is one of two results of an archiving strategy, with
the other result being a full-text collection of the article content. As the web
constantly changes, it was necessary to fixate the texts of the corpus.
Therefore, on a daily basis, I collected all textual material pertaining to
Brazilian corruption cases publicized in the two media outlets. I did so by
checking each of the index sites listing the whole range of political news
items, and cross-checked with the print versions of the same outlets. From
these lists of publicized online material, I selected each and every one
relating to political corruption and downloaded the whole text while also
noting down a range of metadata. In a few instances, the news item had a
steadily developing form (such as live updated material from votes in
Congress or STF), and in those cases, I copied the material after the final
update, usually between 3 and 24 hours after the last update.

Sampled News Items

The problems of defining corruption conclusively have consequences for the
ways in which one can operationalize “information about corruption” in a
research design dealing with media texts. Understanding corruption as a
framed and constructed object, it becomes problematic to use legal
definitions as the yard-stick for determining what is and what is not
corruption. Linguistic criteria could then be imagined; i.e., including all
news items that contain synonyms for corruption, fraud, etc. Both of these
criteria fall short, however, because news items are not necessarily dealing
with corruption explicitly (as in the swimming pool example above) but
may merely imply this assessment of political actions by connotation or
association through framings that emerge intertextually.
Therefore, I opted to study the research object “media texts with
information about corruption” from a minimalist and heuristic definition:
Information about corruption can be found in political news items that have,
as primary cause of their newsworthiness, a possible political transgression.
It is, admittedly, almost tautological to define corruption news as news items
treating corruption as the newsworthy object. However, this definition
eliminates news items about other issues that merely mention corruption in
passing. Meanwhile, the “aboutness” criterion is not a tautology, but a
pragmatic identifier, and this criterion is implied in the definition when
pointing to the cause of newsworthiness. This is the crucial part: If a news
item is newsworthy because the events described are said to be caused by
(or intertextually linked to) a political transgression (alleged or actual), then
it should be included in the sample. In other words, the inclusion criterion
requires that the primary issue of the news item is “corruption.”
It is certainly not a perfect definition, because some kinds of news items
can slip through the cracks. When, say, Congressional debates are paralyzed
because of corruption allegations, a news item reporting on this fact would
count as a news item about corruption, while coverage of Congress debates
merely mentioning corruption allegations would not be included – although
both could be construed by audiences as containing signals about
corruption.
In a few cases, news items have been left out which could arguably have
been included. In general, I included all articles covering political
corruption or corruption in state companies (since these have politically
appointed executive officers), as well as all articles treating or discussing
alleged improper economic or political conduct of the President, ex-
presidents, ministers or ex-ministers, since these articles are partly
overlapping thematically (see the discussion regarding the problem of fluid
corruption definitions above). This means that the articles concerning
impeachment petitions were included, though they in many cases do not
deal with corruption per se, but certainly deal with perceived illicit political
actions. This inclusion was fortunate, because it allows for comparisons
between the most visible political topics of the sampled period –
impeachment and the Lava-Jato case.
 However, I excluded articles about corruption in national and
international football associations, as well as articles about the train cartel of
São Paulo, as that case was running out of steam and went on to disappear
totally from the media agenda. Articles about ex-president Fernando
Henrique Cardoso’s mistress were also excluded, as the framing in most of
these articles (whatever the population in general may think) did not cast the
Cardoso case as corruption, but as extra-marital activity, making it
scandalous for sexual-moral reasons rather than for public-office moral
reasons.
Leaving aside those caveats, I have deployed the definition given and
kept track of the various corruption cases and their spin-off probes,
identifying news items related to these cases and the politicians under
scrutiny. Many of the cases have already been described in the previous
chapters, while this chapter provides insights into the general distribution of
media attention, albeit only with an approximation of the aggregate universe
of media discourse.

Sampling Period
In the front-page sample, the headlines of the three daily papers were
collected for six months leading up to the vote of the Câmara, on April 17,
2016, approving the removal of Dilma Rousseff as president. The database
of all news items on Brazilian political corruption in Estado and Folha
encompasses a slightly longer period, until May 12, 2016, the date when the
Senate approved the removal and Michel Temer assumed the presidency. In
both samples, the sampling period starts in October 2015. The events of the
sampled period were crucial in setting up the political outcome of the
impeachment process, as momentum gathered in favor of ousting the
president. Specifically, this was the period in which the impeachment
petition that eventually was the cornerstone of the impeachment proceedings
was handed in, but also the moment where the Lava-Jato investigations
reached the critical phases of denouncing top politicians, including Lula and
Cunha. The period is therefore interesting because it is the period of
converging scandals – the stories of political corruption and potential
removal of the head of state intertwine in this period throughout the media
sphere. The theme of impeachment first appeared as a serious possibility in
the selected media outlets during this period, and the political consequences
of the Lava-Jato investigations are only manifested from this period onward,
with the initiation of the disciplinary process against Eduardo Cunha.
Impeachment had been mentioned at distinct moments before this period –
most notably immediately after the elections and in February and March
2015 – but the onset of this topic as a common front-page topic can be dated
to October 2015.
The total sample amounts to 1,335 headlines concerning corruption cases
and 8,200 news items in the database of articles. The database and the tally
of headlines and topics are available online, at the time of writing, at the
University of Copenhagen’s Centre for Latin American Studies site
(www.clas.hum.ku.dk/research).

Cascading Corruption News: A Content Analysis

In this section, I will present the findings of the content analysis of sampled
articles and front-page headlines dealing with corruption. This is done first
diachronically, to illustrate how news media built up the informational
cascade, and then synchronically, showing how the total coverage appeared
across the sample period. By aggregating the frequency of news items, I can
demonstrate how news on certain cases was largely disregarded in the total
coverage of political corruption. The coding of issues in the content
analysis, as discussed above, is based upon the attribution of responsibility,
or, if no transgressive actor appears, the investigation. Because all different
cases and the denounced actors were coded in this sample, it allows for
relatively fine-grained analysis of media attention.
The sample period had a few lulls, with days with only one or two
headlines dealing with political corruption, but during most of the sampled
weeks, the national newspapers ran two or three front-page headlines daily
and many articles about the Lava-Jato investigations, impeachment
proceedings, or other cases of political corruption. During the period,
corruption or other forms of political transgression were attributed to dozens
of political actors in the 188 distinct cases identified. At the beginning of the
sample period, in October 2015, the headlines and articles treated a diverse
range of corruption subjects: Lula, Cunha, the CPI of Petrobras, as well as
the case of the pedaladas and the TCU decision to reject the accounts of the
state (see next chapter). However, the diversity was gradually reduced, as
only a few cases, which have been described in the previous chapters,
attracted sustained attention in the sampled outlets.
The news items and front-page headlines presented in Figure 5.3 deal
with Eduardo Cunha, then president of the Câmara, the various inquiries of
the Lava-Jato investigations targeting him, as well as his maneuvers in the
Conselho da Ética to avoid being removed from office. Though Cunha’s
house and offices were searched by police agents in mid-December, and the
PGR requested his removal the following day, the headlines of the week
before Christmas mostly focused on the debate of the STF regarding the
regulations for conducting an eventual impeachment. The holidays and
January 2016 had declining attention to all corruption news, including the
impeachment and Cunha’s cases. A few peaks of attention to Cunha’s cases
are present from January until he conducted the impeachment vote of the
Câmara in March. These peaks appear on the dates where his (first) trial was
on the STF agenda (February 16 and March 3). News dealing with Cunha’s
corruption cases had been increasingly infrequent since the PGR requested
his removal on December 16, 2015. After the impeachment vote, Cunha was
temporarily removed from the position of president of the Câmara in
consequence of his ongoing corruption trial, and this generated 76 articles
(mostly web-based) the following day in Folha and Estado.
Figure 5.3 Number of headlines and news items on Eduardo Cunha’s
corruption cases.

While news attention to Eduardo Cunha remained mostly steady during

the period, news on Lula’s corruption cases eventually eclipsed news
concerning Cunha, starting at the turn of the year. Events related to Lula
commanded the attention of the media at several moments (see the previous
chapter for details). Figure 5.4 demonstrates the frequency of news items,
broken down into six of these cases, as well as the frequency of headlines.
In October 2015, the part of the Zelotes investigation related to Lula was
made public. In November, the arrest of his acquaintance, the cattle king
José Carlos Bumlai, drew headlines. In the last week of January and all of
February, attention increased steadily to the two real estate cases (the ranch
in Atibaia and the beachside triplex in Guarujá). In early March, the media
focus on these cases dovetailed with the breaking news of Lula’s testimony
in the Lava-Jato case, Lula’s indictment by the São Paulo state prosecutors,
and, after two months of scandal coverage, his ultimately unsuccessful
nomination for minister. The increasing media attention to Lula’s cases (and
activity in the investigations) also coincided with the great street protests of
2016, which I return to below.

Figure 5.4 Number of headlines and news items on Lula’s corruption

cases.
 Figure 5.5 shows the 1,176 headlines concerning the main actors of
Brazilian politics that were sampled in the six-month period from October
2015 to April 2016. In the same period, 150 other headlines dealt with
corruption allegations or probes against other actors (governors, the vice-
president, senators, and mayors foremost among these). The company most
frequently denounced for corruption in the headlines is Odebrecht, but this
is not anywhere as frequent as news items on the corruption of politicians.
As discussed in the methodological sections of this chapter, the
information flows leading to cascades must be understood as dynamic
phenomena. This means that when signals are amplified, other signals may
be attenuated and eventually lost in the din, as this graph of the most
covered corruption cases demonstrates The attention to Lula and Cunha
continued into November, with a spike of attention as the PT senator
Delcídio de Amaral was also arrested in the Lava-Jato probe, followed by
peaks of media interest in the first half of December, when Cunha signaled
the start of the impeachment process. October and November’s variety of
signals about corruption was drowned out in early 2016 as the newspapers
dedicated their attention to cases of Lula. The Carnival and the zika virus
attracted attention in February, which meant that the front pages had a brief
lull in attention to corruption. On March 4, 2016, the graph demonstrates
that as Lula was taken to testify in São Paulo airport and the marketing
expert of Rousseff was arrested, the intensity of media attention to
corruption increased and, shortly thereafter, decisively shifted the focus
toward the question of impeachment.
Figure 5.5 Number of front-page headlines in Globo, Folha, and Estado
on corruption cases.

The pivotal point of Figure 5.5 is the days following the leak of recorded
conversations with Lula, on March 16, where Lula’s cases began to be
overtaken by headlines dealing with the impeachment. From that point on,
impeachment news gathered speed and grew tremendously, dwarfing and
eventually (in May) completely suppressing other news about politics and
corruption.

Frequency of Corruption News Items across the Sample

Looking at the total sample of news items from the printed and online
versions of Folha and Estado, it is apparent how the impeachment process
and the corruption cases of Cunha and Lula were massively exposed. In
comparison, news items on other investigated politicians (both opposition
and the erstwhile-government partner PMDB that jumped ship) were few
and far between. The following graph shows the frequency of the most
reported topics (with more than 30 news items in total) besides news on the
impeachment. The impeachment was covered in 1,651 news items in
Estado, while Folha publicized 1,727 news items in the sample period from
October 15, 2015, until the ousting of Rousseff on May 12, 2016. Between
Folha and Estado, there is very little variation in the relative distribution of
attention to the range of corruption cases:
Figure 5.6 Comparison of number of news items on corruption cases
between Folha and Estado.

Apart from the 3,378 news items concerning impeachment proceedings

(left out in the graph), the most covered cases were the various
investigations of Lula (with a total of 261 front-page headlines and 1,507
news items across six cases) and the PT senator Delcídio do Amaral (320
news items in total). The only story that rivaled those about the PT was the
complex story of Eduardo Cunha’s disciplinary process and various
corruption investigations (675 news items and 160 front-page headlines).
News about corruption in Petrobras in general, most frequently mentioning
PT and/or PMDB with no specific politicians associated, reached 600 in the
sample period. Items concerning Dilma Rousseff’s involvement in the
Petrobras graft and laundered money from the Petrobras scheme feeding her
2014 campaign reached more than 290 items in total, while impeachment
news items numbered more than ten times that figure (3,378) in the two
news outlets.
This intense focus on these four persons (Rousseff, Lula, Cunha, and
Delcídio) outshone by a considerable margin the attention to the rest of the
investigated politicians. At the top of the second tier, the president of the
Senate, Renan Calheiros, was the focus of 60-some news items (summing
news items on several cases), just ahead of Aécio Neves. While Neves was
only subjected to formal investigations later in the Lava-Jato probe (outside
of the sample period), Calheiros accumulated investigations in 12 different
instances in the period. The alleged corruption of other central politicians
(such as the senators Romero Jucá of PMDB and Antônio Anastasia of
PSDB) was not even half as frequent a topic as news on Calheiros and
Neves. Like the once-impeached president, now senator, Fernando Collor de
Mello, these third-tier politicians were the objects of fewer than 25 news
items each in total across the two news outlets. Meanwhile, the sons of ex-
president Lula (Luís Claudio and Fabio Luís) were mentioned in more than
50 news items about political corruption, in spite of the fact that neither of
them ever ran for public office.
In relative terms, the attention to the impeachment of Dilma Rousseff and
her possible relation to the Petrobras graft outsized all news topics
concerning corruption. For instance, the ousting of the president was
covered five times as often as the investigation of the president of the
Câmara, and impeachment was covered 11 times as often as the case of the
senator Delcídio do Amaral. The relative weight of corruption news
between news on Rousseff and Neves can be expressed in the ratio 73:1,
while the ratio between news on Lula and Neves was 32:1, and Cunha
relative to Neves 15:1. News on corruption probes into PMDB senators like
Romero Jucá (later to take office as Minister of Planning and Budgets) and
Valdir Raupp were almost lost, about 500:1 in relation to impeachment
news, and about 250:1 relative to news on Lula’s corruption cases.
The former PT minister José Dirceu and the former PT treasurer João
Vaccari Neto were both on trial in the sample period. Both were objects of
several headlines, and their trials were the topic of 70 and 32 articles,
respectively, reporting on several cases.2 This added to the general
directionality of the cascade: That PT was behind the corruption in the oil
company.

Frequency of Corruption News Headlines in the Sampled Front Pages

On the front pages in the sample period, a similar pattern was evident. The
table below presents the frequencies of the most frequent cases in the
headlines, including headlines and front-page items which cover corruption
in Petrobras in general without responsibility attributed to specific political
actors. The issues of impeachment and fiscal delays (pedaladas, described
in the next chapter) are separated from other headlines that cover Dilma
Rousseff’s alleged corruption.
Like in the sample universe of articles in print and web media, the
headlines demonstrate a PT-centric focus of the front-page space of Estado,
Folha, and O Globo. As was the case in the article sample, only Cunha had
a visibility close to Rousseff and Lula, though, as the impeachment
proceeded, not nearly rivaling it. Apart from the above-mentioned actors,
the most-mentioned corruption cases of the headlines included an
embezzlement scheme in São Paulo public schools apparently run out of the
state assembly (12 headlines), the trial of José Dirceu (19 headlines), the
improbable but still newsworthy potential probes into the ex-president
Fernando Henrique Cardoso (5 headlines), as well as several of Lula’s ex-
ministers.
Michel Temer, before taking over the presidency, was mentioned in five
headlines in relation to corruption cases in the sample period. Temer,
furthermore, was mentioned twice in headlines concerning the electoral
court process determining the validity of the 2014 campaign, and twice in
relation to the stalled processes of impeachment petitions requesting his
removal from office. Despite many testimonies mentioning Temer and
PMDB leaders, media attention only shifted toward them as the party
grabbed influential cabinet positions following Temer’s ascendancy. That
attention had effects, however: Within 34 days of Temer seizing power,
three ministers had stepped down (see Chapter 3).
In the chronological analysis, it is clear that the prestigious Brazilian
media outlets, across the board of the political beat and on various
platforms, developed an increasingly narrow focus in the sample period.
Even after the PGR requested the STF to have Cunha removed, the news
media tended to devote less and less attention to the president of the Câmara
while he orchestrated the process of impeachment, eventually commanding
the session of the vote. Other political actors denounced for corruption were
intermittently present on the agenda, but never for long. The only political
actor that received sustained attention until right before the impeachment of
Rousseff was Lula. While the leak of Lula’s conversation and the coverage
of his testimony presented Lula as the key political actor under fire, the
coalition parties of the government signaled that they would jump ship, like
Cunha and PMDB, and vote for the impeachment. This would allow Temer
to form a new cabinet with the center-right parties. The diachronic analysis,
meanwhile, shows that the corruption cases involving exactly these actors
had only been covered fleetingly by the sampled media outlets.

Table 5.1 Number of front-page headlines in Globo, Folha, and Estado

on corruption cases

These findings prompt the following question: Why did the newsrooms
of the large media organizations prioritize the news items as they did,
placing the spotlight on the impeachment process and denying agenda space
to reporting on corruption with the political parties coming into power? The
findings of the content analysis above, naturally, cannot explain the editorial
selection processes nor the information gathering or leaking that preceded
editorial choices. To understand organizational conditions that determine
informational cascades of scandals such as the Lava-Jato and the
impeachment of Dilma Rousseff, I will discuss two information phenomena
found in the media: the selective recycling of information and overvaluation
of information about corruption. Both are information selection pressures
that became crucial for ousting Rousseff, as well as legitimizing her
impeachment. Before doing so, I will consider the organizational
environment of media scandals and market pressures that may have
contributed to the emergence of these information phenomena.

The Attention Economy and Investment of Media Actors

The environment of “big media” – such as the prestigious Brazilian
newspapers and the top broadcasting networks – is an economy of attention,
and a highly important measure of the success of each media outlet is
audience outreach. Maximization of the audience by way of attracting
attention is the central logic or imperative of such media (Landerer 2013).
Once information from corruption investigations has been disclosed to
journalists, the leak enters into a special economy of attention inside the
newsrooms: A scandal, to editors, represent an attention-grabbing pool of
different stories that can be mined continuously with a certain amount of
recycled information. Therefore, dedicating manpower to a scandal is a
relatively cheap investment compared to the number of stories it can yield
(Beale 2006: 401). That editors work with this continuous mining of
corruption cases for information and new spin-off news is clearly visible in
the graphs presented above.
Furthermore, as scandals usually beg the question, “did somebody else
know this?” it is easy to escalate the scandal to other governmental actors or
to a national level, if locally contained. Such an escalation holds the promise
of a greater potential audience. Thus, scandals tend to inflate enormously in
terms of space and attention, but in some cases only ephemerally – what
Vasterman has termed hype cycles (Vasterman 2005, Wien and Elmelund-
Præstekær 2009).
Corruption scandals involving national politics are prime news material
that checks all news value boxes, meaning a potentially nation-wide
audience. So, in terms of audience, the newspaper with a scandalous
headline is assumed to sell more, meaning more advertising revenue. It is
therefore of utmost interest to competing media organizations to break such
news before competitors, as this will give the organization with the scoop
the upper hand in relation to sales, clicks, and views. In a situation of
stagnant or dropping audiences, if advertising is central to the business case
of the newspaper or network, then scandals will especially be in demand.
Dropping audiences, common to the majority of print media systems today,
also threatens media industry of Brazil. Like the editors, media owners and
shareholders thus have a weighty incentive to invest in scandals, although in
their cases it is actual financial investments that are at stake in this economy
of attention.
To editors as well as owners, the question of sources, credibility, and
individual professional careers is usually of less concern than it is to the
journalists. Journalists, on the other hand, invest time, reputation and
goodwill from their sources in bringing new aspects of the scandal to the
table. What they potentially gain from this is of course professional status,
and the perks that come with it. What they might lose, in the case of false
allegations, is the trust of their sources, access to corridor talk and a dent in
their professional image. Various techniques are used to write up news items
deriving from information in corruption cases so as to protect journalists
from status loss.
Principally, the risk-free asset of the scandal is the well-researched and
objective news item. Unfortunately for normative theories of the public
sphere, there is a number of ways in which the journalist might jump the
gun, overstate the case, trust an untrustworthy source, or frame the issue
beyond fairness. This constitutes a risky investment of credibility in such a
less-than-optimal news construction (Liebes and Blum-Kalka 2004:
1154ff.). In some media systems, of course, the repercussions may be minor,
either to the media outlet that can wash the hands by firing the reporter and
hiring another journalist, or to the reporter him- or herself who might
weather a few months of bad standing with the powerful, as long as the
position, payroll, and byline is still secure. Sources that blow the whistle
have even more to lose, no matter if their story catches on or not.
The short-term investment of covering scandals holds a potentially very
large promise of returns for the journalist. Since the Watergate case, the idea
of the investigative journalist that brings about the downfall of a president
has hovered as the fixed star of the journalistic profession (Schudson 2004:
1234). The ideal incorporates within it the notions of the media as the fourth
estate, media as the watchdog of society, and the guardian of public truth.
Scandals, when they have political repercussions, confirm this idealized role
of the journalist. The ideal of unmasking power also leads to many frenzy-
like situations, barrage-questioning politicians and corporate bosses in full
run, in the hope that the stellar exposés written by one heroic journalistic
protagonist may secure democracy, along with Pulitzer prizes.
 In most media systems, legal frameworks exist to govern and guard
against slander, unfounded allegations and fanciful denunciations. However,
the worth of a scandal, in terms of sold copies and air-time, may very well
surpass the cost of a lawsuit to the newspaper or broadcasting network.
Especially in cases where evidence is sparse, but likely to trigger formal
investigations, the investment in the scandal promises great returns. Thus,
editors with economic incentives and journalists with ideological and
prestige-oriented incentives constitute the two intertwined sets of actors in a
media system that combine to produce the bandwagon behavior: Cascades
of news items that take their cue from one another lead to a circular
orientation to the original allegation of the scandal (which may or may not
be true). The cascade may in turn feed into the decisions of other actors
(say, in the political system), and eventually media bandwagon effects may
emerge across an entire media system: Everybody wants a piece of the
action and therefore needs to join the fray, truth be damned, if nothing else
then just to cover the resonance of the media hype. The bandwagon effect
generated by media covering media, results in a feedback loop and more
amplification, magnification, and stimulation of coverage (Vasterman 2005:
511).

The Centripetal Recycling of News

At the systemic or organizational level, the vertical ties between Brazilian
media outlets are likely to amplify a cascade, which results in the
reproduction of certain information and the tendency to ignore other events.
As mentioned before, Brazilian news media are organized in a few powerful
conglomerates, with much vertical as well as horizontal integration between
different regional outlets and different media types. A few newsrooms
control the gates of the political system, and the information admitted
through these gates propagates through networks of news agencies and
inter-media shoveling (Paterson and Domingo 2008).
The incipient research field of gatekeeping in digital media spheres
furthermore suggests that when news dissemination moves online, fewer
news gatekeepers become more influential than before (Bro and Wallberg
2015, Welbers et al. 2018). In a vertically integrated media system like
Brazil, where only a few news agencies distribute most of the news (Aguiar
2014: 55), the fragmentation and multiplication of online news sources has
its counterweight in this inwards spiral of attention of news media. Fewer
issues, touching only selected societal bases, stay on the agenda of
mainstream media for longer periods and are replicated by online news
media, whether associated or independent. The centripetal pattern of
shutting out minor news issues and focusing intensely on nation-shaking
crisis news was evident in the Lava-Jato case online as well as offline.
The selective recycling of only a handful of corruption cases – in a
scandal that seems to have capillaries reaching almost every corner of
Congress and most state governor seats – is a democratic problem. The
mainstream media, Folha, Estado, and Grupo Globo and their affiliated
media groups downstream, in particular, have been accused of an
ideological slant. The charges of anti-PT bias might be true in some
individual instances, most evidently in the weekly magazines, but the
problem of recycling information about a not-representative portion of the
corruption cases might also be a systemic problem:

the mostly noble intentions of journalists can result in a media agenda

that is far from most normative ideals. These noble intentions get
filtered through a complicated system of organizational and
marketplace incentives and disproportionate information processing. As
a result, such intentions are not always reflected in the news and its
patterns.
(Boydstun 2013: 209)

A major marketplace incentive is the audience profiling of newspapers

and the resurgence of the party press strategy mentioned in the second
chapter (Lattman-Weltman and Chagas 2016). The disproportionality
introduced by the media system integration and the newsrooms responding
and to scoops of other newsrooms by recycling the scoops and then
covering the same item from other angles is another filtering effect behind
the recycling.
Finally, the group-think mentality of editors in scandals may be caused
by the fact that the news values of a scandal story are usually given an equal
assessment by editors, leading them to locate the stories at the same level of
importance in newspapers or in newscasts. This is a pattern generally found
(Shoemaker and Vos 2009: 52) – the most prominent news items are
reported in essentially the same way – and it is especially true for Brazilian
mainstream media reporting on scandal cases. The observation that certain
events, dealing with political corruption at the national level, are
unanimously considered “good news,” leads me to the problem of
overvaluation: When the assessment of news values is distorted by
bandwagon effects among editors and cascade effects between media
outlets, democracy cannot rely on media as producer of accountability. In
the following section, I will explore the notion of value of information in
news media.

Value of Information and Overvaluation

The information flow of a scandal involves not just disclosure of
information, but also the anticipation of value, and the a priori realization of
this value. This flow is shared between media and the political system,
broadly speaking. By value, I designate the possible usage of some
information in accountability processes, such as the various judicial actions
mentioned in the previous chapter. This usage is obviously political in
nature, but is reflected and shaped by the media.
What I term value here goes beyond the usual sense of “news values” in
the literature (a classic term, see Galtung and Ruge 1965). News values
“reflect ideologies and priorities held in society” (Bell 1991: 156), and are
thus not neutral constructs outside of the particular journalists and editors
who decide to include or exclude information. Rather, the attribution of
value varies contextually. In corruption news, the factors for determining
value include the relevance of disclosed information to institutional
processes or accountability processes (what can be included, what will count
as evidence, etc.), and the eventual trajectory of the processes triggered by
disclosure (what is the ultimate outcome, e.g., how much damage can
disclosed information inflict on the careers of the denounced?). This is
linked to the ideal of watchdog journalism discussed in the sections above.
The potential repercussions of a scandal are also determining factors in the
editorial processes and the redissemination after publication; editors need to
prioritize what disclosed information merits a follow-up story, and consider
what will likely be retweeted, liked, and shared, etc. All of these questions
are strongly oriented to the future horizon of political and public action. In
other words: An information, as element in the scandal, is only as
interesting as its potential political value. This value derives from its
disruptive potential, but more importantly, the value is relative to the
perceived center of political agency.
The proximity to perceived centers of power is an important factor in
determining what information will be transmitted in mainstream media’s
information flow. It is common to the news criteria of most national media
systems to value coverage of power-holders, although certain positions of
power might be excluded from scrutiny (Thompson 2005: 43). News
production in national politics is strongly centralized, and this is mirrored in
the allocation of journalistic attention: “in a presidential system[,] media
tend to reinforce the emphasis on the administrative aspects of government,
rather than on party politics” (Albuquerque 2012: 91). With unclear party
lines, indistinguishable ideological stances between parties, and
heterogeneous government coalitions, political agency is often made a
matter of persons rather than parties – and the person of the president is
viewed as the central political force, dwarfing party leaders and governors.
Thus, the presidential political system also influences the relative value of
news items concerning political actors.
Media attention to information is thus modulated by the perceived
potential value of that information, just like in markets of other assets.
Unlike most tradeable assets, however, information does not lose value by
being copied. On the contrary, in scandals, given greater attention, the
dissemination of the original scandal will instead create increased demand
for new information on the topic. This feedback structure of information
demand in scandals is essential to creating the conditions for a bubble-like
situation, where certain information is valued greatly and other information
that would normally be of public interest, is systematically undervalued.
Following the impeachment and successive indictments of the PMDB
politicians and Michel Temer, it is obvious that the usage value of
information is neither necessarily nor sufficiently determined by its
belonging to categories of misinformation or disinformation. Disinformation
and misinformation, even when understood to be such, might still have
scathing potential. Therefore, in scandals, verifiably true information is not
guaranteed to be as interesting and therefore demanded as disinformation or
misinformation. Rather, assumptions of value are based on the potential
impact of the news, and this impact is partly determined by the position of
the denounced actor. These assumptions play out somewhat like speculation
on a stock market, where actors move resources in expectation of future
returns. In the same way, the anticipation of the value of information is part
and parcel of how media operates in scandals.
Interpreting the editorial selection processes of Brazil in these terms, we
may imagine that journalistic production dealing with the impeachment of
Dilma Rousseff was in high demand by editors because this process was
perceived to actually produce consequences and likely to change the
political landscape. Meanwhile, the processes which stalled (e.g., legal
actions against PMDB leaders and the ex-president Collor) were less in
demand, as they were either perceived or actively framed to be dead-end
stories, with no likely consequences and little relation to the center of
power.
The same reasoning could be used to explain the apparent contradictory
overvaluation of news items on Lula’s corruption cases: Although Lula was
not in public office, the fact that he was considered a likely presidential
candidate in 2018 made his trials much more interesting than any other part
of the Lava-Jato investigations. The disruptive potential, as described in the
previous chapter, of denying Lula the possibility of running for president,
would have considerable consequences for all parts of the political
spectrum. Although other ex-presidents – Collor, Cardoso, and Sarney –
were also implicated by testimonies, denying Lula’s political agency in
2018 would have much more impact. Thus, despite allegations against
Cardoso, the PGR requesting Sarney’s arrest and indicting Collor, it was
Lula’s sentence in 2017 that came to be framed as “historical.”

Further Stages of the Informational Cascade

Having discussed the intra-media stages of the information flows in the
Lava-Jato and impeachment scandals, I now turn to a number of societal
spheres where disclosed information trickled down and influenced
collective behavior and decision-making processes. The particular cases
analyzed in the previous chapter demonstrated that news on corruption in
the media may activate various accountability mechanisms, but the
activation of institutionalized accountability processes usually requires
sustained media scrutiny, and various political influences may obstruct
accountability processes. The cascade of information documented in this
chapter, meanwhile, may trickle down in other contexts and dimensions of
the public sphere than accountability processes. The dimension of electoral
accountability cannot be directly connected to the scandals of 2015–2017,
due to the electoral cycle of Brazil (and the fact that Brazil is anyway a
country with limited electoral accountability; see the previous chapter).
However, I have picked out several examples that may illustrate the
ramifications of the media’s informational cascade in other contexts.
The assumption of this section is that media attention is correlated (and
maybe causally related) to other spheres where information is valued, and
that the media agenda may drive the processes of information flows there. I
will briefly review four such spheres, using four secondary sources of data,
to establish how and to which extent the information emerging in the Lava-
Jato scandal reached Brazilian publics. In each of these spheres, information
pathologies may arise.
First, while collecting metadata of the sampled news items, the social
media statistics of each web-published article were also harvested (where
available). Such data gives a glimpse into the afterlife of online news items;
usually ephemeral, but in some cases brightly flaring. While the most
scathing leaks and allegations obviously get attention by social media users,
that attention is not consistently reflected in editorial decisions of follow-up
stories. The inconsistency of editorial selection and audience sharing of
news speaks volumes concerning the uptake and viral potential of news
items on specific political actors. Second, as several public opinion
institutes regularly conduct surveys in Brazil, the impact of corruption news
upon approval ratings, audience perceptions of societal problems, and the
political agenda can be explored. Third, the fluctuations of the stock markets
of Brazil provide yet another set of interesting data. The concept of
information flows seems particularly apt when noticing how smoothly and
rapidly information disclosed and leaked to the media was fed into the
financial and monetary flows of the Brazilian and international finance
market. Events on the political scene deemed important in the media
triggered immediate reactions from actors on the financial markets, and
those steps of the cascade sequence were visible in aggregated financial
indicators, such as stock markets and exchange rates. Finally, since social
mobilization and political participation is linked to the (social) media
agenda (Boulianne 2015), it is reasonable to compare the sampled coverage
of corruption to the street protests occurring in the sample period.
Social Media
While collecting metadata of news items, I also collected the social media
statistics of each article provided by embedded widgets on the news site
(where available). This is a data source that needs to be treated with
reservations in order to generalize, because the online audience of each of
the prestigious newspapers sampled is not at all representative of Brazil in
general. But the distribution of likes and shares certainly does provide some
hints as to which corruption cases attracted the most (or the least) amount of
attention on a short-term basis. Because the data provided by the media
platforms regarding social media distribution continuously varies for each
article, the data is rather unreliable: Likes are retracted, profiles are deleted,
and the scripts of the embedded widgets on the news sites are developed and
changed without any transparency to the researcher. Furthermore, script
errors and deliberate manipulation of bots can introduce false data or
manipulate the counters.
The most shared news item in the sample period on Folha’s website,
from October 25, 2015, revealed that Lula’s son was being investigated in
the Zelotes investigation. This item was shared 293,000 times, according to
the embedded counting script. That this item should be top-ranked seems
plausible, since this was the first news item in Folha directly associating
Lula’s family to corruption – albeit not the last. The most shared news item
in the sample period on Estado’s site Estadão was an item published on
November 11, 2015, reporting on ex-STF judge Joaquim Barbosa (the judge
rapporteur of the Mensalão case). In the article, Barbosa commented on the
situation of Petrobras, foreseeing Brazil’s “humiliation” in the legal
processes running in the legal system of the US. This item was shared
132,300 times according to the counter.
Among the top-ten shared news items of both outlets, there is surprising
diversity: One denunciation of Aécio Neves’ corruption reached 42,100
shares on Estadão, while Folha had two reports, dealing with two other
denunciations implicating Neves, that reached 110,000 and 87,000 shares.
Estadão’s other most-shared items dealt with Cunha’s cases (two items
reaching around 40,000 shares) and Lula (55,000 shares of another article in
the Zelotes case). An Estadão item about the hostile reactions to Neves and
São Paulo governor Geraldo Alckmin during street protests in March
reached almost 59,000 shares. Apart from Neves’ and Lula’s cases, the most
shared items of Folha’s site in the sample period detailed two other inquiries
targeting PSDB (Alckmin and the late national president of the party, Sérgio
Guerra), both items reaching 93,000 shares. Outside of the sampling period,
other breaking news on Folha’s site dealing with arrests (of PT leaders such
as Dirceu and Vaccari Neto) reached around 60,000 shares. Many other
politically intense moments in the sample period were described in series of
articles, which would likely divide the frequency of shares among several
news items, meaning that these moments are not visible as outliers in the
dataset on the parameter of social media shares.
Despite the attention of social media users, the three cases relating to
Neves and the two other cases linking PSDB to political corruption did not
become objects of more front-page news and little investigative journalism
afterwards. These stories are one-shot items on the graphs presented in the
previous sections, in contrast to the sustained media attention to Lula.
Somehow, the social media interest did not translate into editorial
prioritization of these sub-scandals emerging from the Lava-Jato
investigations. The data cannot reveal whether this is due to political bias in
newsrooms, or the sub-politics of information flows and source relations in
the accountability institutions, but it is nonetheless significant that some
storylines which gained much traction online were promptly disregarded in
the media.

Public Opinion Polls

Opinion and agenda polls registered a doubling of the amount of
respondents that noticed and remembered negative news concerning the
Brazilian government more than doubled in the six months following the
2014 elections. The survey respondents registering negative news about the
Rousseff government reached 72 percent in March 2015, and that amount
was steady in the following surveys (Indicadores CNI 2015a, 2015b).
During the same period, President Dilma Rousseff’s approval ratings in
polls fell from around 50 percent during the 2014 elections, 40 percent
around Christmas 2014, to 12–13 percent (Indicadores CNI 2016, Datafolha
2016b) in March 2015, to 8–10 percent in July–August 2015 (Datafolha
showing the lower figure, surveyed in early August 2015; ibid., 14).
Government approval ratings for the months corresponding to the media
sample presented here is very similar to the fluctuations of corruption news:
 Baptista (2017) has demonstrated that the government approval ratings
and surveys of trust in the Rousseff administration, in the various polls
conducted during 2015 and 2016, correlated significantly with the negative
coverage of corruption on the front pages of the weekly magazines – more
negative news covariates with lower trust in government and lower approval
ratings. The only moment where this was not true was immediately after the
2014 elections, when both approval ratings and negative coverage grew.

Reactions on the Stock Market

The stock market fluctuations to the political events in the sample period
provide a third source of data, which could be interpreted as stages of an
informational cascade. Specifically, the initiation of each step in the
impeachment process triggered immediate reactions from actors on the
financial markets, visible in aggregated financial indicators such as stock
markets and exchange rates. Just like other actors observing the signals of
the political events, traders and shareholders reacted to the publicly
available signals: When news indicated impeachment of Rousseff as likely,
Brazilian stocks increased in value, and inversely decreased in periods
without news on the impeachment, because the Rousseff administration was
depicted as unable to implement economic reforms. Although theoretically,
the sudden concerted moves of shareholders scattered across the world
could be mere correlation, I will argue that the events that gradually brought
about the ousting of Dilma Rousseff, because they were internationally
visible and of obvious consequence to the economic policies of Brazil, were
in fact causing these fluctuations.
Figure 5.7 Approval ratings of government and coverage of scandals,
2014–2016.

Source: Baptista 2017.

As an example, a spike in Brazilian stock value is visible on the day

when the impeachment petition was accepted (December 3, 2015). The São
Paulo stock market Bovespa rapidly reacted, breaking a steady decline,
rising 3.29 percent on average within the day, with Petrobras preferred stock
rising 6.12 percent and the Banco de Brazil leading with an 8.4 percent rise.
As the Supreme Court intervened in the impeachment process shortly
thereafter, and it became clear that the initiation of proceedings would only
take place after the Carnival, the Bovespa index fell into decline again
(while the dollar-real exchange rate climbed). However, coinciding with
news that heralded the impeachment of Rousseff (e.g., the imprisonment of
Rousseff’s marketing strategist João Santana on February 23, and new plea
bargains alleging corruption in Rousseff’s campaign on March 4), that trend
was broken again.
Figure 5.8 Bovespa index, November 2015–March 2016.

Source: BM&F-Bovespa 2018.

Figure 5.9 Bovespa index, April 18–May 12.

Source: BM&F-Bovespa 2018.

Zooming in on the last days of Dilma Rousseff’s presidency, Figure 5.9

shows the Bovespa index from the day of the vote in the Câmara to the day
of the vote in the Senate.
In response to the Câmara ratification, the Bovespa index rose with a
brief interruption until April 27, continuing the trend of March and April.
On the 28th, the Central Bank chose to maintain the interest rate at 14.25
percent and the market cooled off for a fortnight, only to rise again with the
two-day debate in the Senate heralding the removal of Rousseff on May 12.
The collective of actors known as “the stock market,” in the aggregate,
viewed the impeachment optimistically and responded positively to each
step of the impeachment process.
The exchange rate between the Brazilian real and the US dollar also
reflected the market activity and the possibility of impeachment. In Figure
5.9 the tendency is therefore the same (although graphically inverted): The
real steadily lost value vis-à-vis the dollar throughout the second half of
2015, and only recuperated with the string of negative news for Rousseff
that started in February and became plainly visible from March onward.
Not surprisingly, the actors of the stock market react on possibilities – in
this case, on the possibility of impeachment, trying to buy stocks cheap
before other actors realize that their value may increase after the
impeachment. This explains the rapid spikes, rather than prolonged value
increases, of the market: Short bursts of speculative activity and no real
changes in the economic realities of Brazil, because the Congress stalled
economic reform during the whole period while waiting for the
impeachment.

Figure 5.10 Exchange rates BR$–US$, August 2015–August 2016

Source: Banco Central do Brasil 2018.

 The media system and the financial markets are linked empirically, but
the underlying phenomenon of information value also links them
theoretically: The value of (scandal) information on the media market, just
like on the stock market, is anticipated (that is, potential) value.

Street Protests
Although the sample period of the media data presented here comprises
seven months of coverage, it is worthwhile to consider both the street
protests of the sample period and the period preceding it. Immediately after
the presidential elections of 2014, right-wing groups voiced concern with
the election result, and the leading opposition party PSDB filed several
actions in the electoral court, as described in the preceding chapter. The
same political spectrum, galvanized around three online movements,
mobilized crowds for street protests during 2015 and until the vote of
impeachment in the Câmara on April 17, 2016. While the mobilization of
these protests and the size of the crowds are obviously not only related to
the news on corruption, but determined by many factors, it is fair to assume
that the degree of success in mobilizing large groups against political
corruption would be influenced by news on exactly that topic. Therefore, I
will venture a few observations concerning the relation between news on
political corruption and the street-level support of the various protests in
Brazil in 2015–2016. The majority of these were cast as anti-governmental,
while a number of street rallies supported Dilma Rousseff and PT,
denouncing the impeachment process as a coup d’état. The anti-government
street protests were planned some time ahead, mostly by the same three
groups Movimento Brasil Livre, VemPraRua and Revoltados Online (as
well as minor participation by SOS Forças Armadas and the right-wing
party Solidariedade) (Chaia et al. 2017: 9). Because of the identical
organizational setup combined with the varying information flows of media,
we can think of the news as the independent variable that, to some
unverifiable extent, has influenced the dependent variable of street rally
turnout.
In 2015, the largest protests took place on March 15 in a range of
southeastern Brazilian state capitals and larger cities. The figures reported
for the number of participants varied wildly: While the organizing groups
claimed that millions protested in total, and the police of São Paulo claimed
that 1 million had participated in the protest in downtown São Paulo, the
survey institute Datafolha more modestly counted 210,000 participants at
the peak that day (Datafolha 2015a). In Rio de Janeiro, the police gave an
initial estimate of 15,000, while the organizers claimed 50,000 participants
were present. In Brasília, the police estimated 40,000 protesters, and in Belo
Horizonte, 24,000. The most covered events of the news that immediately
preceded these protests against government corruption were the
imprisonment of the PT treasurer and the inquiries being opened into 47
high-level politicians from the government coalition. Furthermore, the week
before these protests, several affluent neighborhoods of Rio and São Paulo
had “panelaços”: People slamming pots and pans together from their
windows in protest – a protest targeting the televized speech of Dilma
Rousseff commemorating the International Women’s Day.
Later anti-government protest of 2015 did not reach the same levels of
mobilization. Although, again, estimates disagreed, they were all lower
relative to the estimates of the March protests – in São Paulo on April 12,
Datafolha counted 100,000, and in August, 135,000. Police reported 25,000
in Brasília on August 16, while their colleagues in Rio refrained from
reporting numbers on the same day. On December 13, the lowest turnout of
the anti-government protests was registered: Estimates clocked in at 30–
40,000 protesters in São Paulo and some thousands in Brasília, Belo
Horizonte, and Rio. The low turnout could be correlated with a string of bad
news for the main forces moving the impeachment along: The disciplinary
process against Eduardo Cunha was underway, Michel Temer was being
ridiculed for a leaked letter to the president complaining about his lack of
duties as vice-president, and the STF had overturned the procedure of
impeachment initiated by Cunha. The counter-protests on December 15 had
more success, gathering 55,000 in São Paulo, according to Datafolha, and
100,000 according to the organizers (which here consisted of trade and
student unions as well as left-wing movements). Earlier pro-government
rallies had gathered on April 7 and August 20 with less support, but with the
narrative of a looming coup d’état firmly grounded in the coverage of the
pending impeachment proceedings, the left-aligned movements and unions
had more success on the streets in December.
The groups Movimento Brasil Livre and VemPraRua once again called
for street protests in March 2016, and these turned out to muster massive
support. This time around, the call for impeachment overshadowed the anti-
corruption agenda and the support for the Lava-Jato investigations; or
perhaps more precisely, these topics were mixed together now. Both Folha
and Estado dedicated their front pages to illustrating the crowds massing in
São Paulo, reporting that the protests were the largest seen in Brazilian
history, with estimates of the police reaching 1.4 million and the Datafolha
counting 500,000 protesters downtown (Datafolha 2016a). The Rio police
force again did not give an estimate, while the organizers claimed one
million protesters marched in the protests. More than 200 other cities had
registered protests. Preceding this successful mobilization was three weeks
of the largest build-up of negative news for Rousseff and Lula especially. As
seen in Figure 5.2, the attention to the Lula cases exploded with his
testimony in early March, more Petrobras revelations, and the arrest of the
PT marketing strategist João Santana in February. News on impeachment
also climbed in this period, with PMDB leaving the governing coalition
wholesale and paving the way for a supermajority against Rousseff in the
Câmara. However, as mentioned above, the mood of the great street protests
in São Paulo was hostile to the political establishment in general, and the
PSDB leaders Neves and Alckmin had to retreat from the public protest.
The circulation of corruption news can be interpreted as one stage of an
informational cascade that feeds into the mobilization of street protests. I
would argue that the mobilization of protests is possibly correlated with (but
not solely determined by) the media agenda, and I can demonstrate
empirically that the coverage of the protests amplified the informational
cascade, as the sampled media outlets produced massive amounts of
coverage of the largest anti-government protests. The coverage of the
protests, of course, also tied in to the coverage of corruption and the
impeachment. On the other hand, the anti-impeachment protests of the
following week, which still got mentioned on the front pages of Folha and
Estado, saw much less article coverage, and the layout and photo editors of
both newspapers chose to run much smaller images of the anti-impeachment
protests. In size, frequency, and staying power, the anti-impeachment was a
lesser item on the media agenda.

Conclusion
I will summarize the argument of this chapter by highlighting an effect of
the media coverage that is distinct from the agenda setting or salience
transfer effects discussed in classic media studies.
In the attention economy of modern media organizations, agenda space is
a limited resource, as is the resource of journalists in newsrooms. The
Brazilian media had an exceedingly diverse menu of corruption cases to
look into, and, as demonstrated in Chapter 3, many agents were interested in
pushing information pertaining to particular cases. Such conditions result in
heterogeneous pressures for selection of information. In addition to this, the
same media organizations are pressed to maximize their audience by
providing information relevant to a national public, rather than a local one.
The outlets, furthermore, have strategic goals of keeping certain kinds of
news consumers as paying clients. While the input side of the information
flows was perhaps diverse in 2015, the journalistic and editorial product of
Folha, Estado, and O Globo increasingly came to focus on scandals
implicating Lula in early 2016. As demonstrated above, the virtual
audiences were apparently attentive to many other cases in addition to
Lula’s, and the large street protests of March 2016 called for large-scale
political renovation as well as impeachment of Rousseff. The combination
of large-scale street protests and intense media coverage of Rousseff’s
increasing isolation, Lula’s failed nomination, and the imprisonment of PT’s
marketing expert, however led the government coalition partners to jump
ship. Sensing that impeachment was a real possibility, the media coverage
intensified while the markets reacted positively.
Since the very start of the Lava-Jato investigations, key non-PT
politicians had been the objects of sporadic media exposés: Vice-President
Michel Temer, Senate President Renan Calheiros, and a number of other
senators and congressmen, all from PMDB. The name of the 2014
presidential runner-up Aécio Neves, likewise, had cropped up in the
denunciations from time to time, with short-lived media attention. These
politicians became the key actors in the formation of a new government.
With the leaks of conversations between Lula and Rousseff, government
coalition parties found a convenient excuse to jump ship, and news on
impeachment immediately flooded the landscape of media. In the two
months from the great street protests in March to the preliminary removal of
Rousseff on May 12, the media outlets were engrossed in the proceedings
and the spectacle of the impeachment. Ignoring the investigations
implicating the parties coming into power, the intense media focus paved
the way for the ousting of Rousseff, and legitimized her impeachment.
 The informational cascade in the media had not just agenda-setting
effects, but more insidiously, the informational cascade had the unfortunate
effect of eclipsing accountability mechanisms and confusing the political
scenario of impeachment with corruption investigations. Rather than
concerning themselves with the increasingly real possibility of corrupt
politicians taking power in order to save their own hides, the collapse-of-
government frame dominated the political news with increasing salience.
Failing to hold the future government accountable, the informational
cascade of the media reproduced the same signal and, on top of that, created
a media event around the ousting of Rousseff. In the following chapter, I
will review the ways that the media event of the impeachment came to be
intertwined with the Lava-Jato investigations and the discourse of
corruption, despite the fact that the effective legal ground of the
impeachment was neither corruption nor related to Petrobras.

Notes
1 Floridi argues (2011: 104) that derived information is really meta-information, and that a
definition of semantic information, strictly speaking, cannot include indirectly emerging data, but
must inherently be truthful. Floridi would therefore reject this categorical scheme used here on
ontological grounds. For the purposes of this chapter it is necessary to encompass the (derived or
direct) variants of information that are not truthful, although this, to Floridi, would merely be
communication which does not entail truth value (ibid.: 96).
2 Due to the diverse codings of the cases involving Vaccari and Dirceu, neither politician appears in
the graph of the most-reported topics.

References
Abramo, C.W. (2007) “Brazil: A Portrait of Disparities” Brazilian Journalism Research, vol. 3(1),
93–107.
Aguiar, P. (2014) “Marx explica a Reuters: economia política das agências de notícias.” In J.A. Silva
Jr., M.C. Esperidião, and P. Aguiar (eds.), Agências de Notícias: perspectivas contemporâneas
(38–78). Recife: EdUFPE.
Albuquerque, A. de (2012) “On Models and Margins.” In D. Hallin and P. Mancini (eds.), Comparing
Media Systems beyond the Western World (72–95). Cambridge: Cambridge University Press.
Altheide, D. and Snow, R. (1979) Media Logic. Los Angeles, CA: Sage.
Banco Central do Brasil (2016) Cotações de fechamento de Dólar dos Estados Unidos. Retrieved
from www4.bcb.gov.br/pec/taxas/port/ptaxnpesq.asp?id=txcotacao.
Baptista, E. (2017) “Corrupção e opinião pública: o escândalo da Lava Jato no governo Dilma
Rousseff.” PhD Dissertation. Belo Horizonte: UFMG.
Beale, S. (2006) “The News Media’s Influence on Criminal Justice Policy: How Market-Driven News
Promotes Punitiveness” William and Mary Law Review, vol. 48(2), 397–481.
Bell, A. (1991) The Language of News Media. Oxford: Blackwell.
Bikhchandani, S., Hirshleifer, D., and Welch, I. (1998) “Learning from the Behavior of Others:
Conformity, Fads, and Informational Cascades” Journal of Economic Perspectives, vol. 12(3),
151–170.
Bird, E. (1997) “What a Story! Understanding the Audience for Scandal.” In J. Lull and S. Hinerman
(eds.), Media Scandals: Morality and Desire in the Popular Culture Marketplace (99–121). New
York: Columbia University Press.
Biroli, F. and Miguel, L.F. (2013) “Meios de comunicação, voto e conflito político no Brasil” Revista
Brasileira das Ciências Sociais, vol. 28(81), 77–95.
BM&F-Bovespa (2018) Cotações. Retrieved from www.bmfbovespa.com.br/pt_br/servicos/market-
data/cotacoes/.
Boulianne, S. (2015) “Social Media Use and Participation: A Meta-Analysis of Current Research”
Information, Communication and Society, vol. 18(5), 524–538.
Boydstun, A. (2013) Making the News. Politics, the Media, and Agenda Setting. Chicago, IL:
University of Chicago Press.
Bro, P. and Wallberg, F. (2015) “Gatekeeping in a Digital Era” Journalism Practice, vol. 9(1), 92–
105.
Cabral, E.D.T. (2017) “Mídia Concentrada no Brasil: Até Quando?” Revista Latino-Americana de
Ciencias de la Comunicación, vol. 13(24), 49–59.
Carey, J. (1986) “The Dark Continent of American Journalism.” In R. Manoff and M. Schudson
(eds.), Reading the News (146–196). New York: Pantheon.
Chaia, V., Maranhão, C., and Martinho, S. (2017) “O Brasil da telenovela e do jogo político: um
estudo de caso do impeachment de Dilma Rousseff e a estética das telenovelas brasileiras no ano
de 2016.” Paper presented at ANPOCS 41st conference at Caxambu, Minas Gerais, October 24,
2017.
Damgaard, M. (2015) “Multiple Margins and Mediatized Transgression” Ephemera –Theory and
Politics in Organization, vol. 15(2), 411–434.
Datafolha (2015a) Manifestação na avenida Paulista. 15/03/2015. São Paulo: Datafolha.
Datafolha (2015b) Avaliação da presidente Dilma Rousseff – novembro de 2015, PO813824. São
Paulo: Datafolha.
Datafolha (2016a) Manifestação Avenida Paulista. 13/03/2016. São Paulo: Datafolha.
Datafolha (2016b) Avaliação da presidente Dilma Rousseff – abril de 2016, PO813859. São Paulo:
Datafolha.
Datafolha (2017) Avaliação do presidente Michel Temer – novembro de 2017, PO 813942. São Paulo:
Datafolha.
Dretske, F. (1981) Knowledge and the Flow of Information. Cambridge, MA: MIT Press.
Entman, R. (1993) “Framing: Toward Clarification of a Fractured Paradigm” Journal of
Communication, vol. 43, 51–58.
Entman, R. (2010) “Media Framing Biases and Political Power: Explaining Slant in News of
Campaign 2008” Journalism, vol. 11(4), 389–408.
Fallis, D. (2009) “A Conceptual Analysis of Disinformation.” Preprint of iConference 2009.
Retrieved from www.ideals.illinois.edu/bitstream/handle/2142/15205/fallis_disinfo1.pdf?
sequence=2.
Fallis, D. (2011) “Floridi on Disinformation” Etica & Politica/ Ethics & Politics, vol. 13(2), 201–
214.
Floridi, L. (2002) “On Defining Library and Information Science as Applied Philosophy of
Information” Social Epistemology: A Journal of Knowledge, Culture and Policy, 16(1), 37–49.
Floridi, L. (2005) “Semantic Conceptions of Information.” In Stanford Encyclopedia of Philosophy.
Retrieved from https://plato.stanford.edu/entries/information-semantic/.
Floridi, L. (2011) The Philosophy of Information. Oxford: Oxford University Press.
Galtung, J. and Ruge, M. (1965) “The Structure of Foreign News” Journal of Peace Research, vol.
2(1), 64–91.
Ghanem, S. (1997) “Filling in the Tapestry. The Second Level of Agenda Setting.” In M. McCombs,
D. Shaw, and D. Weaver (eds.), Communication and Democracy: Exploring the Intellectual
Frontiers in Agenda Setting (3–14). Mahwah, NJ: Lawrence Erlbaum Associates.
Grice, H.P. (1957) Studies in the Way of Words. Cambridge, MA: Harvard University Press.
Hendricks, V. and Hansen, P.G. (2016) Infostorms. Why Do We “Like”? Explaining Individual
Behavior on the Social Net. 2nd edition. London: Springer.
Indicadores CNI (2015a) CNI/Ibope Avaliação do governo, septembro 2015. São Paulo:
Confederação Nacional da Indústria.
Indicadores CNI (2015b) CNI/Ibope Avaliação do governo, dezembro 2015. São Paulo: Confederação
Nacional da Indústria.
Indicadores CNI (2016) CNI/Ibope Avaliação do governo, março 2016. São Paulo: Confederação
Nacional da Indústria.
Iyengar, S. (1990) “Framing Responsibility for Political Issues: The Case of Poverty” Political
Behavior, vol. 12(1), 19–40.
Kristensen, N.N. and Mortensen, M. (2015) “Amateur Sources Breaking the News, Metasources
Authorizing the News of Gaddafi’s Death: New Patterns of Journalistic Information Gathering and
Dissemination in the Digital Age” Digital Journalism, vol. 1(3), 352–367.
Landerer, N. (2013) “Rethinking the Logics: A Conceptual Framework for the Mediatization of
Politics” Communication Theory, vol. 23, 239–258.
Latinobarómetro (2017) Informe Latinobarómetro 2017. Santiago de Chile: Corporación
Latinobarómetro.
Lattman-Weltman, F. and Chagas, V. (2016) “Mercado Futuro: A Economia Política da
(Re)Partidarização da Imprensa no Brasil” Dados, Revista de Ciências Sociais, vol. 59(2), 323–
356.
Lazarsfeld, P., Berelson, B., and Gaudet, H. (1944). The People’s Choice: How the Voter Makes Up
His Mind in a Presidential Campaign. New York: Columbia University Press.
Liebes, T. and Blum-Kalka, S. (2004) “It Takes Two to Blow the Whistle: Do Journalists Control the
Outbreak of Scandal?” American Behavioral Scientist, vol. 47(9), 1153–1170.
Lima, V.A. da (2004) “Sete teses sobre mídia e política no Brasil” Revista USP, São Paulo, no. 61,
48–57.
McCombs, M. and Shaw, D. (1972) “The Agenda-Setting Function of Mass Media” Public Opinion
Quarterly, vol. 36, 176–187.
McCombs, M. (2005) “A Look at Agenda-Setting: Past, Present and Future” Journalism Studies, vol.
6(4), 543–557.
Matos, C. (2008) Journalism and Political Democracy in Brazil. Lanham, MD: Lexington Books.
Obelitz Søe, S. (2014) “Information, misinformation og disinformation. En sprogfilosofisk analyse”
Nordisk Tidsskrift for Informationsvidenskab og Kulturformidling, 3(1).
Obelitz Søe, S. (2016) The Urge to Detect, the Need to Clarify. Gricean Perspectives on Information,
Misinformation and Disinformation. PhD Thesis. Copenhagen: University of Copenhagen.
Paterson, C. and Domingo, D. (eds.) (2008) Making Online News: The Ethnography of New Media
Production. New York: Peter Lang.
Pring, C. (2017) People and Corruption: Latin America and the Caribbean. Berlin: Transparency
International.
Reese, S. (2007) “The Framing Project: A Bridging Model for Media Research Revisited” Journal of
Communication, vol. 57, 148–157.
Russell, A. and Waisbord, S. (2017) “The Snowden Revelations and the Networked Fourth Estate”
International Journal of Communication, vol. 11, 858–878.
Schudson, M. (2004) “Notes on Scandal and the Watergate Legacy” American Behavioral Scientist,
vol. 47, 1231–1238.
Sniderman, P. and Theriault, S. (2004) “The Structure of Political Argument and the Logic of Issue
Framing.” In W. Saris and P. Sniderman (eds.), Studies in Public Opinion: Attitudes, Nonattitudes,
Measurement Error, and Change (133–165). Princeton, NJ: Princeton University Press.
Thompson, J.P. (2000) Political Scandal: Power and Visibility in the Media Age. Cambridge: Polity
Press.
Thompson, J.P. (2005) “The New Visibility” Theory, Culture & Society, vol. 22(6), 31–51.
Tuchman, G. (1978) Making News. A Study in the Construction of Reality. New York: The Free Press.
Tumber, H. and Waisbord, S. (2004) “Introduction” American Behavioral Scientist, vol. 47, 1143–
1152.
Van Dijk, T. (1988) News as Discourse. Mahwah, NJ: Lawrence Erlbaum.
Vasterman, P. (2005) “Media-Hype. Self-Reinforcing News Waves, Journalistic Standards and the
Construction of Social Problems” European Journal of Communication, vol. 20(4), 508–530.
Waisbord, S. (2000) Watchdog Journalism in South America. News, Accountability, and Democracy.
New York: Columbia University Press.
Walgrave, S., Boydstun, A., Vliegenthart, R., and Hardy, A. (2017) “The Nonlinear Effect of
Information on Political Attention: Media Storms and U.S. Congressional Hearings” Political
Communication, vol. 34(4), 548–570.
Welch, M., Prica, E., and Yankey, N. (2002) “Moral Panic over Youth Violence: Wilding and the
Manufacture of Menace in the Media” Youth and Society, vol. 34(1), 3–30.
Wien, C. and Elmelund-Præstekær, C. (2009) “An Anatomy of Media Hypes: Developing a Model for
the Dynamics and Structure of Intense Media Coverage of Single Issues” European Journal of
Communications, vol. 24(2), 183–201.
Wiewiura, J. and Hendricks, V. (2017) “Informational Pathologies and Interest Bubbles: Exploring
the Structural Mobilization of Knowledge, Ignorance, and Slack” New Media and Society, Article
first published online: January 10, 2017 – https://doi.org/10.1177/1461444816686095.
Wortham, S. and Rhodes, C.R. (2012) “Narratives across Speech Events.” In A. De Fina and A.
Georgakopoulou (eds.), The Handbook of Narrative Analysis (160–177). Malden, MA: Wiley-
Blackwell.
6 Coup d’état or Constitutional Act?

 
At the outset of this book, I asked how the failures of accountability
processes in Brazil could happen in plain sight of the population and the
media, and how the checks and balances of the Brazilian democracy, that
supposedly had become more and more consolidated, so quickly could give
way. How could the initial success of the Lava-Jato investigations,
apparently ending the rule of impunity for the business elite and the
politicians benefitting from corruption and cartelization, turn into a media
crusade against corruption that undermined the rule of law?
I will retrace the steps of this book in order to answer that question. By
doing so, this chapter can also shed light on the question frequently raised
in debates during and after the impeachment: Was the impeachment of
Dilma Rousseff really a golpe, or coup d’état, or was it a congressional act
within the limits of the Constitution – an exit to the deadlock of Brazil’s
faltering coalitional presidentialism under Rousseff?
First, I will revisit the theoretical line of this book. In scandals, media
matters in several ways: because media attention signals importance to
society and sets the agenda, and because this agenda-setting effect may
trigger accountability processes (Peruzzotti 2006: 255, Smulovitz and
Peruzzotti 2000: 152, Waisbord 2000: 230ff.) and call decision-makers to
action (Walgrave et al. 2017). While media agenda setting in scandals can
have electoral consequences and also lead to the mobilization of street
protests, it is the ability to set the stage for horizontal accountability that is
most interesting here, because exposés in agenda-setting media may partly
determine who is to be held accountable (Porto 2011: 122).
There are many factors that determine what gets covered by media and
what does not. Media agendas are usually characterized by skew (Boydstun
2013), generated by the incentives and restraints of media institutions.
Media agenda congestion, policy-maker attention and the institutional set-
up of news media covering political life combine into two feedback
mechanisms – internally in particular media organizations, and between
media actors. Such feedback is especially likely in a vertically integrated
and concentrated media system such as Brazil’s (Cabral 2017), because the
same news items propagate down a chain of media outlets within the same
conglomerate. Though scandals are likely to also be characterized by
explosiveness, sudden spikes of attention may still turn into a sustained
media storm (Walgrave et al. 2017). This happens when an informational
cascade (Hendricks and Hansen 2016) emerges between various actors
(such as media outlets, political parties, and financial actors), repeating and
recycling the same signal many times over. All of this will exacerbate the
bias (Feres Júnior and Sassara 2016) of an already skewed media agenda.
While the scrutiny of political corruption and improbity in media
exposés normally triggers accountability processes (Porto 2011), a radically
skewed media agenda may counteract accountability, by exposing only a
select few of many different corruption cases. Some cases then linger in
court dockets and on police desktops, while others are hastened through
without due consideration. In a judicial system where corruption trials may
drag out for a decade or more (Taylor 2011: 172, Falcão et al. 2014),
rapidly executed processes are perceived as exceptional, obtain more
coverage in the short run, and stand out on the media agenda. Insiders of
investigations and trials may leak or push information to media outlets to
exacerbate this pattern. Leaks that can scandalize members of the political
elite are more likely to make it past the gates of a media system for three
reasons; because this kind of news deals with actors in the eye of the public,
because it ticks all the boxes of news values (Galtung and Ruge 1965), and
because media outlets (even without access to the leaked information itself)
cover the repercussions as well as the original coverage; “in this way, the
coverage becomes in part coverage of the coverage” (Kristensen and
Mortensen 2013: 361).
In sum, by continuously leaking information of corruption investigations
on a large scale, these actors may drive the media agenda, which in turn
triggers accountability processes and supports mobilization of discontent.
Even when journalists create or yield space to competing frames, the leaky
sources hold power to initiate cascades between media actors, if these
choose to recycle the leaked information, and thereby to create the
momentum for accountability processes. This book has aimed at exposing
and discussing this systematic link between leaks from concealed sources,
the web of accountability institutions, and the Brazilian media system.
The link between leaked disclosure of corruption, media, and
accountability processes, influences politics in three ways. Reforms and
bills are proposed to remedy the democratic deficits, but this might merely
be a strategy of signaling and maintaining legitimacy (Seligson 2002).
Political actors can also utilize information from leaks and media exposés to
initiate further legal actions. If they are able to influence courts’ decisions
and jurisdictions, they have the option of not only waging lawfare
(Comaroff and Comaroff 2006), but also reinforcing the media attention to
those cases from which the legal actions derived. In addition to this,
political actors may capitalize on the momentum of informational cascades,
skewed media agendas, and accountability processes, by strategically
jumping ship (Balán 2011). The use of denunciations of corruption as an
intra-government instrument to claim more power may thus look like a
genuinely ethical act on the surface, but in reality cover an intra-elite power
struggle where political dominance is sought through non-democratic
means (Waisbord 2000). If media actors fail to acknowledge this and act
upon it, they are contributing to the construction of manipulative publicity
in the public sphere (Habermas 1989 [1962]: 178), to the detriment of true
accountability.
In Chapter 4, I described the failure of the Brazilian horizontal
accountability processes since 2015 to produce equal and timely
interventions into politics, although the Lava-Jato probe was frequently
represented as exceptionally effective. Nonetheless, the successes of the
Lava-Jato eclipsed several prominent failures of accountability institutions,
and the horizontal accountability failures were thereby compounded by the
media-borne vertical accountability deficits in equilibrating the overall
attention to the actors coming into power, as analyzed in Chapter 5. Chapter
3, meanwhile, described the moments where various political and judicial
processes were kick-started by leaks in the media, most notably the gradual
disintegration of the governing coalition in 2015 and early 2016. The net
result of these stacked failures, leaks, and scandals was the impeachment of
Dilma Rousseff and the quickening of court trials involving Lula (relative
to the delays of legal processes involving non-PT politicians).
To demonstrate that this cascade of leaked information and political
responses actually contributed to the ousting of Rousseff and the
inauguration of Michel Temer, I will now outline the process leading to the
impeachment. I can then conclude the book by considering how the
impeachment and the array of other consequences produced by leaks and
lawfare affected the general condition of the Brazilian democracy. Finally, I
will consider the lessons of Lava-Jato and the impeachment in theoretical
and practical terms.

Fiscal Delays and the Impeachment of Dilma Rousseff

The impeachment petition, submitted in October 2015, that eventually led
to the ousting of President Rousseff in August 2016, was based on several
charges, many of them related to the corruption of Petrobras. However,
according to the law defining impeachment proceedings (law 1079 of
1950), acts anterior to the mandate of the incumbent cannot legally ground
an impeachment process. That clause meant that even if conclusive
evidence of Rousseff’s active or passive acceptance of corruption in
Petrobras before 2015 had surfaced, she would have been shielded from
such charges (although the allegations or evidence could still be used in
TSE trial). Despite this restriction of the impeachment law’s scope, the
arguments concerning Rousseff’s role in the graft of Petrobras contracts
were still present in most of the stages of the process: In the initial petition,
in the public debates and the votes of the Câmara and Senate, as well as in
the final accusation, the libelo acusatório (Wink 2017). So, in order to
provide a firmer ground for initiating the process of impeachment, the
opposition had to frame the impeachment petition around Rousseff’s
personal responsibility for certain fiscal delays. These were dubbed
pedaladas fiscais, and a description of the impeachment demands a brief
exposition of the history of those delays. The following section cannot do
the complexity of the constitutional and fiscal intricacies justice, but may
give the reader some sense of what the delays came to signify in public
discourse.
During 2014, the Finance Ministry had expanded an existing and
common practice of delaying transfers of funds between the State Treasury
and public banks. Three public financial institutions (the Caixa Federal, the
Banco do Brasil, and the FGTS) transferred money to citizens in several
conditional cash transfer programs, as well as to companies through several
lines of credit for industrial companies and agribusinesses, but the State
Treasury would then “pedal” for a while, only dispensing funds some days
later. Shifting temporarily the financial burden of the programs onto other
public institutions, the delays somewhat masked the fact that the state
accounts were in bad shape after years of plummeting oil prices and the
great slow-down of the Chinese economy, the most important export market
of Brazil. The expansion of this practice, reaching at one point R$3.5 billion
in delayed funds from day to day, was revealed by journalists from the
financial beat of Estado in April 2014. While the secretary-general of the
State Treasury, Arno Augustin, denied that this constituted an illegal budget
maneuver, the government still stopped that practice in 2014. The Lava-Jato
scandals and the elections pushed the case of the pedaladas from the
headlines. Therefore, the topic lay dormant since August 2014, save for one
front-page headline on January 11, 2015, in Estado, but it had been on the
agenda of an audit inquiry internally in the TCU, the Tribunal of Union
Accounts. On April 17, Estado and Globo highlighted the leaked report of
Julio Marcelo de Oliveira, Prosecutor of TCU. Oliveira then presented the
tribunal with an audit report that pointed out that pedaladas infringed the
Law of Fiscal Responsibility (Villaverde 2016: 177ff.). Meanwhile,
impeachment re-emerged in the media in February, March, and April 2015
(e.g., Zalis 2015), and opposition parties gladly commented on the issue.
José Mucio, one of nine TCU judges, was assigned to the case. He
agreed with the auditor and recommended that the court plenary decide on
the culpability of all involved in the fiscal maneuvers of 2014. Despite
strong protests from the government, headed by the Attorney-General,
protests from a number of ministers from the 2014 government and
presidents of public banks testifying in the court, the other TCU judges
concurred. The maneuvers and the defenses of government occupied
journalists covering the financial beat in the daily newspapers in June, July,
and August 2015 (Villaverde 2016: 181ff.). However, in 2015, the practice
of pedaladas had already been stopped. The only delayed payments
between the State Treasury and the banks were transfers to back up an
agribusiness line of credit promoted by the government called the Plano
Safra. This credit line, totaling R$5.6 billion, had nominal payments in
January and June 2015 that were only eventually paid in December 2015.
While the case was under scrutiny in the press and in the TCU, on July
17, the first crack in the government coalition appeared when Eduardo
Cunha stated that he considered himself in opposition to PT, and that he
wanted all of PMDB to leave the government with him. Cunha had suffered
two days of journalistic barrage because a leaked testimony in the Lava-Jato
case implicated him. Cunha criticized the PGR, and claimed the
government was covertly attacking him through the corruption
investigations. Cunha also began to undermine the stability of Rousseff’s
position by signaling the viability of the impeachment petitions he had
received as president of Câmara. Through his position, he was vested with
the power to deny or proceed with the impeachment process. Removing
Rousseff would bring Vice-President Temer, also national president of
PMDB, to the presidency, and in the event of the vice-president’s removal
as a result of the ongoing TSE trial, Cunha would be the next in line to
claim the presidency (see Chapters 3 and 4 concerning Cunha’s corruption
cases).
Despite Cunha urging his party to jump ship, PMDB remained in the
coalition and even gained more posts: In the cabinet reshuffle on October 2,
the party gained one additional minister position for a new total of seven
(Almeida 2016: 89), but almost half of the party’s members in the Câmara
began to stray from the government line in votes. Soon after Cunha broke
ranks, Michel Temer declared at a press conference that “someone” would
now have to ensure political unity to govern the country. The vice-president
implied that the president could not be entrusted with this responsibility,
underscoring the growing tension between PMDB and PT (ibid.: 58).
Nonetheless, PT could not govern without PMDB, but many PMDB
Congressmen were voting for conservative bills, spearheaded by Eduardo
Cunha. In late July, Cunha also added two new elements to his pressure.
First, a range of expensive bills (called the pautas-bomba) that would force
the government into a budget deficit; and second, Cunha stated that he was
now analyzing one particular petition filed by Hélio Bicudo, former PT
member and founder, and Miguel Reale Junior, former PSDB Minister of
Justice in the last year of Cardoso’s government. The two senior lawyer-
politicians (aided, from August onwards, by the younger professor of law
Janaína Paschoal) marshalled three overall arguments in three different
versions of their impeachment petition:

• The corruption in Petrobras (linked to the Mensalão case).

• The pedaladas (eventually interpreted to be legally identical to the
delayed payment for the Safra credit line in 2015).
• The state budget of 2015, then moving toward a deficit.

The final argument was included when the lawyers realized that
expenditure totaling R$2.5 billion in presidential decrees had been
announced by the government in July and August 2015. However, at the
moment of issuing the decrees, the Congress had not ratified a state budget
deficit, and spending in excess of the state budget could therefore be
considered an unauthorized loan and a breach of the Law of Financial
Responsibility, article 36, just like the pedaladas. A revised petition penned
by the trio of lawyers was handed in to Cunha on September 17, and
resubmitted in October 15, in a final revision. Both submissions became
media events through the coverage of the large print and broadcast media.
In the final petition, the president was deemed responsible for the delays,
which according to the petition were crimes against the state. Both pieces of
information were contested: The practice of delaying payments is wide-
spread at both federal and state level in Brazil, and Rousseff’s hand in
cooking up quite specific fiscal transactions was debatable (but see
Villaverde 2016).
Meanwhile, TCU had scheduled the verdict of the 2014 state accounts
for October 7. The report of Oliveira, the TCU auditor, concerning the
pedaladas, became the grounds for rejecting the accounts, and with this
sentence, TCU reinforced Reale’s, Paschoal’s and Bicudo’s interpretation of
delays, ruling pedaladas to be in violation of the Law of Fiscal
Responsibility. Only once before, in a crisis of the Vargas government in
1937, had the tribunal rejected the state accounts. To deny that line of
argument in an eventual impeachment process, the government also tried to
remedy the situation of delays in 2015, including the paying back of interest
to the public banks, adding R$55.8 billion to the deficit.
With the rejection of the 2014 state accounts in the TCU, the outlooks
for the accounts of the state in 2015 turned even bleaker, and Minister of
Finances Joaquim Levy was struggling to find a way of ensuring that the
state budgets of 2015 would not end up with a deficit, in spite of declining
tax income. The noticeable problems for the government of ensuring a
budget surplus conspired to make Standard & Poor’s lower the investment
grade of Brazil (Almeida 2016: 69ff.). This was seen as a blow to Levy. The
economic goal of the state, the meta fiscal, was revised several times, for an
extraordinary deficit at the end of 2015. The revisions also contributed to
the lowering of the country’s international investment grade with the other
major international grading agencies.
The approval of the state budget only passed the floor of the Congress
very late in the year. With the hope of pushing the Rousseff administration
into another breach of the Law of Fiscal Responsibility, the vote for
approving the budget was postponed several times in November 2015 by
the opposition and Cunha at the helm of Congress. With the unexpected
arrest of the senator Delcídio do Amaral, the budget wound up on the
agenda in the last week of the parliament working year. The Congress
ended up ratifying a total state budget deficit of R$119.9 billion, and Levy
left office shortly. The media represented his resignation as a response to
the congressional deadlock.

A Skein of Scandal and Legitimacy

As described in Chapter 4, in November, Cunha repeatedly stated that the

petition for impeaching Rousseff was almost cleared for take-off in
Congress. This was interpreted by journalists and pundits as a bargaining
chip in his own struggle to avoid a disciplinary process in the Conselho da
Ética. The PT representatives in the Conselho voted for opening a process
against Cunha, on December 2, and the same afternoon, Cunha declared
that he would accept the petition filed by Bicudo, Reale, and Paschoal. In
the absence (at that time) of any investigations targeting Rousseff
specifically in the ongoing corruption probe, the negative evaluation of her
administration’s fiscal policies and management of state accounts became
the kernel of the impeachment process.
A number of maneuvers masterminded by Cunha dominated the agenda
of Congress in the week that followed. Cunha established a set of rules that
allowed dissenters of government coalition parties to be elected for the
special Congressional commission that would be installed to analyze the
impeachment petition. Through secret vote-casting for the candidates for
this commission, the Congressmen could ignore party discipline (and the
normal rules of Congressional commissions), by voting for lists consisting
of dissenters. Protesting against this, several ballot boxes were smashed in
the Congress by Rousseff allies during the tumultuous vote. Aiming to stop
the chaos of the Câmara, the STF judge Barroso preliminarily revoked
Cunha’s rules and urgently put the question of due process in the
impeachment on the agenda of the STF for the following week. This helped
Rousseff to a certain degree, as the intervention of STF on December 16
and 17 annulled the special commission installed by Cunha and made the
coming stages of the impeachment process less volatile and subject to his
artifices. However, even with a fixed set of rules defined by STF, the
government was still facing the fact of the crumbling coalition.
What was less visible at the time of the intervention was the STF’s stake
in securing the constitutionality of the impeachment process. Fully aware
that with an indicted politician running the process, the impeachment
looked less than legitimate, STF’s president (at the time, Ricardo
Lewandowski) agreed to hear Cunha’s complaints the day before
Christmas. The meeting between the president of the Câmara and the
president of the STF was made open to journalists, however, and
Lewandowski denied every issue taken up by Cunha. This had the double
effect of showing that the STF was both committed to securing the
constitutionality of the process and that the Court did not confide in Cunha,
who left the audience chamber of the Court embarrassed. Crucially, the STF
had also unanimously denied the government’s request for stopping the
entire process due to Cunha’s indictment. The initiation of the impeachment
might be a part of Cunha’s grab for power and hope of immunity against
prosecution, but the STF maintained the position that it was entirely
legitimate.
Following leaks of conversations between Lula and Rousseff (see
Chapter 3), and the arrest of their marketing strategist João Santana, dissent
grew in most parties of the government coalition. The special commission
that was eventually installed in March 2016 contained a majority of
Congressmen in favor of impeachment. The dissent was probably also
motivated by Realpolitik: At this moment, the federal government was
unable to guarantee much influence to their allies, and with another state
budget deficit looming, the gains from pork-barrel politics were expected to
be minimal. Furthermore, the possible removal of Rousseff would result in
a new cabinet, meaning entirely new horizons of influence. An eventual
new governing coalition could pick out new persons for the many positions
of confidence in state companies, agencies, and entities, and continue to
meddle in public tenders.
No matter the motivation, the special commission approved a report in
favor of impeaching Rousseff on April 11, with the votes 38 for and 27
against. The registration for the final meeting of the commission even saw
physical infights between substitutes of absent commission members. The
commission’s concluding report (Arantes 2016) repeated that government
credit lines without backing to the banks doling out the money were
effectively unauthorized loans, like the pedaladas, and the report affirmed
that Rousseff was ultimately responsible for this practice. An addendum to
the impeachment law 1079/50 links this law to the Law of Fiscal
Responsibility, so that budgetary movements not ratified in the annual
Budget Law are defined as crimes of the Executive, and this meant that
Rousseff transgressed both laws. The major scandals of the PT government
(now defined as crimes of administrative improbity) were again mentioned
in the special commission’s report, despite the delimitation of the temporal
scope in the impeachment law. Thus, the Lava-Jato investigations, the
Zelotes case, and PT’s and Rousseff’s involvement in the corruption of
Petrobras anyway snuck into the process (ibid.: 10ff.).
Cunha then scheduled the vote of the Câmara plenary for the following
Sunday, April 17, and 73 percent of the Chamber of Federal
Representatives voted in favor of the petition for impeaching Dilma
Rousseff. Of the 511 representatives voting, virtually none cited the explicit
justification of the petition – the delays of transfers from the Treasury to the
public banks. However, scores of representatives stated that by voting for
impeachment, they voted against corruption, frequently reproducing the
exact narrative about Rousseff that first emerged in 2014 with the initial
disclosure of corruption in Petrobras.
Cunha was removed as president of the Câmara shortly thereafter (as
described in Chapter 4), and his substitute, Waldir Maranhão (of the PP),
revoked the vote of April 17, but then retracted that revocation within 24
hours. The Senate ratified the Câmara’s vote on May 12, removing Rousseff
for a maximum of 180 days. Between the vote in the Senate that
temporarily removed Rousseff and the final vote in August, a technical
report made by Senate clerks demonstrated that no internal document
corroborated the theory of Rousseff’s personal involvement in masking
budget deficits through the fiscal delays (Pederiva et al. 2016: 212ff.), but
that the decrees for extra expenditure in July and August 2015 should have
been previously ratified in Congress. The technical report pointed out that
in the process of executing the decrees, no decision-makers in government
were made aware that delays of payment for these decrees would impact
upon the budget goal.
Nonetheless, the special commission of the Senate produced a report in
favor of ousting Rousseff permanently during June and July, again bringing
in arguments about Rousseff’s role in the graft of Petrobras, and the
commission ratified the report with the votes 14 to 5 on August 4. Ricardo
Lewandowski, finishing his period as president of the STF with this
process, could then conduct the final stage of the impeachment process in
the Senate in the end of August, sitting beside Senate President Renan
Calheiros of the PMDB, also investigated by the Lava-Jato task-force. The
Senate voted on August 31, after two days of political-juridical
proceedings, with 61 senators in favor and 20 against impeaching Rousseff.
In spite of the fact that the two issues were legally separate, the Lava-
Jato scandal fed into the impeachment process in no small way, tangling
together questions of political responsibility, hidden motives, and
corruption. Despite these extra-processual intrusions along the way, from
Cunha’s maneuvers to the final charge, the STF maintained a shroud of
legitimacy under the auspice of then-President Lewandowski. Although no
other executive or governor in Brazil has been removed for fiscal delays,
despite doubts about Rousseff’s personal responsibility in the transfer of
funds between state entities, and despite the fact that Rousseff was never
charged in a civil or criminal suit for administrative improbity, STF secured
the execution of impeachment according to the letter of the constitution.
Perhaps because the judges viewed this as a viable exit from the political
deadlock of the Congress, or maybe because they believed in the
accusations floating concerning Rousseff’s role in the graft of Petrobras, the
STF became a stunted instance of checks and balances: By merely allowing
the Legislative to demand accountability of the Executive, without
bothering to hold accountable the Congressmen playing the jurors of the
impeachment, the STF deliberately unbalanced the equilibrium in the
Brazilian separation of powers. This was obvious in the way that the
impeachment was conducted as a solely political trial, without the ambition
of upholding the rule of law beyond the letter, and it became even more
obvious in situations involving STF and TSE Judge Gilmar Mendes the
following year. Here, impunity for Temer (in the TSE trial), his cabinet
(e.g., Moreira Franco or Romero Jucá), and supporters (such as Aécio
Neves) was prevalent, while Lula, Rousseff, and their ministers enjoyed no
protection through similar interventions.
The vertical dimension of accountability mattered little in relation to the
shroud of legitimacy bestowed by the STF, as the scholars of law and the
bar association paid more attention to the impeachment itself than to the
role of STF in securing the process. Furthermore, media attention worked to
the detriment of the vertical dimension of accountability, as clarified in the
previous chapter: The agenda-setting outlets systematically over-exposed
the impeachment process and reduced attention to the cases involving every
PMDB leader but Cunha, paving the way for replacing Rousseff with her
vice president, Michel Temer. At that point, Temer himself had been
implicated by five different individuals in testimonies of the Lava-Jato
probe, and, like Rousseff, he was indicted by the PGR in September 2017
for his role in the corruption of Petrobras.
Although Rousseff may someday be found guilty in the Petrobras case,
for passively or actively maintaining the model of political control over
directorates (and tacitly or expressly accepting the derived kickback
payments to PT and the coalition partners), the impeachment process rested
on other legal grounds. These grounds were fragile, but the STF judges
upheld a façade of legitimacy, and they went out of their way to secure the
execution of the impeachment, while never stopping to question whether
the culpability for personally attributed crimes, defined by the impeachment
law as crimes de responsibilidade, actually extends to acts committed down
the chain of command. Was Rousseff really personally ordering the delays
and the specific technical realization of credit lines of the Plano Safra? If
there were protests against the delays in the State Treasury, had she herself
demanded the silencing of these protests from her subordinates? And if
these quasi-pedaladas of 2015 were deliberately executed to cook the books
of the state budget and uphold the country’s investment grade, should it be
considered a crime against the state, or a defense of the nation’s financial
assets? And if Rousseff actually did all of this maliciously, did the penalty
of losing the presidency equal the gravity of the crime?
Given these deficiencies, the impeachment was arguably a coup, in the
sense of an illegitimate take-over of government with no consulting of the
electorate, although the process was effectively dressed up as a
constitutionally legitimate act. While the legality of the proceedings was
observed, I will argue that the legitimacy of the impeachment was
thoroughly impaired as Temer and his party were participants in the graft of
the state companies and benefitted from the corruption. To add insult to
democratic injury, grabbing the reins of the presidency also provided Temer
with the instruments for stopping the TSE trial and curtailing other
corruption probes. In a way, the greater scandal was really the skein of
legitimacy and scandal, which permitted the maintenance of the façade. In
this entanglement of media coverage of impeachment, lawfare,
contradictory judicial rulings, faltering checks and balances, and openly
failing accountability mechanisms, it was never consistently questioned
whether prosecutors and judges, while dredging information up, speeding
up or dragging out cases, were also dragging the Brazilian democracy onto
thin ice. With Temer’s narrow escape from the TSE trial and two
indictments in 2017, impunity was restored, leaving the ice to thaw and
closing around the hole left by the sinking legacy of PT’s ex-presidents.

Crusade Against Corruption in a Crumbling Democracy

The two preceding sections have given a partial answer to the question
posed in the introduction, concerning the power relations between the
Brazilian Executive and Legislative branches. A decade ago, Brazilian
democracy seemed well-consolidated, competitive, and increasingly
accountable (Avritzer 2016: 27, Power and Taylor 2011: 266), orbiting
around the office of the president. The strength of the presidency made
scholars of Brazilian politics (e.g., Amorim Neto 2007, Pereira and Mueller
2000) argue that the Executive branch of the Brazilian government in fact
held too much power relative to the other branches. After the impeachment
of Rousseff and the two obstructed indictments of Temer, this perception
needs to be reformulated. Now, what matters is really the power of veto-
players in the Congress to form alternative majorities. In other words, the
lesson of the Lava-Jato case and the impeachment to political science was
that a Brazilian president is dependent on Congress, and not the other way
around.
The president, moreover, is only to a limited extent able to influence
judicial affairs. In the analyses presented here, we see that the STF acts
incongruently but independently of the executive branch (even when a
majority of the bench is appointed by the prevailing party’s presidents),
while the TSE remains heavily influenced by the president. The Ministério
Público and the PF act independently, albeit with a possibly vested political
agenda, but a Congress (super-)majority can shield presidents from
investigations – and this is a central bargaining chip in the power game
determining Executive prerogatives. The checks and balances of Brazilian
politics is a complex game, and the Congress – not the president – is
situated centrally to fix the weights of the balances, as the arbiter of
impeachments, deadlocks, and veto powers against accountability
mechanisms initiated by the MPF.
The considerations of checks and balances above leads me to the fourth
and final question posed initially. Could it be that independence of the
judiciary is not beneficial to democracy? In the introduction to this book, I
remarked that we generally perceive the independence and impartiality of
the judiciary as a bedrock of the rule of law. The independence of the
Brazilian public prosecutors and the courts has grown tremendously since
the 1990s, so much that the debate has now seen a sea change: Could
independent courts and prosecutors turn out to be a problem for democracy?
This is a complex debate best left for another full-length study, but certain
elements of this answer have cropped up in the course of this book.
First, when key actors of Congress and the Executive work toward
ending the combat against corruption, independence of the judiciary is
obviously of utmost import. No more than two weeks after the
impeachment, it became clear that substituting Temer for Rousseff would
not bode well for the Lava-Jato investigations. Leaked recordings published
by Folha in May 2016, produced before the impeachment became a reality,
featured a key witness in the Petrobras case discussing the removal of
Rousseff with major political actors of PMDB. In the conversations, the
politicians (Renan Calheiros and ex-President Sarney, for instance) agree
with their interlocutor that it is necessary to end the corruption probes into
the political establishment, which threatened to “put an end to the political
class.” Another key actor, PMDB Senator (and later Minister) Romero Jucá,
affirmed on the tapes that it was necessary to “stop the bleeding” of the
political class by changing government and ending the PGR’s and the Lava-
Jato task force’s project of cleaning Brazilian politics:
 We have to stop this shit … We have to overthrow the government in
order to stop the bleeding. To stop [the project seeking to] finish off the
political elite, and construct a new caste of politicians, pure, that aren’t
corrupt … they would fall. All off them. Aloysio, Serra, Aécio.
(Cited in Valente 2016, translation mine)

The growing suspicion that a group within the political elite of Brazil
was interested in removing the President Dilma Rousseff with the intention
of saving their own hides was thus confirmed. With this leak and others to
follow, the exertion of sustaining the façade of legitimacy was negated by
events taking place between May 2016 and 2017. In consequence of the
above-mentioned leak, Jucá and two other ministers of the Temer cabinet
were forced to resign. The testimonies of Odebrecht leaders that gradually
leaked in November and December 2016 made it clear that many decision-
makers in Congress and in governor’s offices would also be interested in
“stopping the bleeding.” In May 2017 the Temer administration and the key
supporting senator Aécio Neves became embroiled in scandal again,
through yet another leak of the Lava-Jato investigations. Temer was
discussing hush money to stop Eduardo Cunha from negotiating a plea
bargain in his prison cell, while Neves had requested bribes for paying his
lawyer’s bills in the Lava-Jato case. This time, with the dissemination of
recorded conversations and the tracking of carry-on bags of cash delivered
to Temer and Neves’ aides, there was no denying that kickbacks and bribes
continued to play a role in the backstage politics at the highest level, and
that political actors were more than willing to obstruct justice.
However, the flip-side of independent prosecutors and judges is the
willingness to leak and prosecute politics in ways that actively disrupt
democracy and deliberately distort the distribution of media attention.
Holding politicians accountable is a good thing, but when the information
game of lawfare and media hunt becomes more viable than court trials, the
electorate is not served, and neither is the sense of justice.
Despite a strong professional identity, the prosecutors aligned at different
moments with various political forces, undermining first Lula, then
Rousseff, and later Temer, by leaking information at critical junctures. So,
with the strategic option of leaks available, judicial actors can bypass the
institutionally designed spaces for resolving or mediating conflicts, relying
on media hype to create the necessary legitimacy for the very procedure of
bypassing. But where medical bypass surgery necessitates delicacy, the
effects of this targeted transparency is rarely surgical in its incisions into
institutional settings. Ignoring due process and individual rights to privacy
for extended periods can result in dysfunctional institutional relations and
constitutional crisis, as seen at various moments.
The fact that the STF, TSE, and MPF in conjunction have been calling
the shots during the unfolding scandals meant that general elections of 2018
were determined by judiciary actors at the outset. The outcome of the 2018
elections – unknown at the time of writing this book – will also be shaped
by the climate of the Lava-Jato investigations and impeachment, in addition
to the campaign donation reform of 2015 plus the historically abnormal
situation of presidential candidates from both PT and PSDB being
discredited in the media. At a more tangible level, however, judiciary
interventions and the timing of legal processes may keep some corrupt
politicians in the running for presidency and Congress while stopping
others due to the tardiness of legal processes and the Ficha Limpa law.
Presidential candidates and the scores of investigated federal
representatives and senators alike maintained or lost their eligibility at the
mercy of the timing of the accountability institutions, while a polarized and
volatile political context resulted in many voters looking to outsider
candidates perceived to be “clean.”
Another problematic consequence of judiciary independence is the
speculation flowing from leaks. Setting the stage for conspiratorial media
interpretations, leaky accountability serves political interests by tarnishing
images and disregarding due process; this was the case of Judge Sérgio
Moro’s online publishing of phone conversations that documented the
anxiety of Rousseff and Lula in relation to Lula’s appointment as minister,
but hardly showed criminal masterminds at work. The publication of the
Rousseff-Lula dialogue was however swiftly picked up by the mainstream
media to reinforce the narrative of a political crisis.
The combat of corruption became a crusade with vested party interests,
always unknowable to the electorate despite the apparent increased
transparency of legal processes. This great innovation of the Lava-Jato case
– the unceasing publication and pushing of information – has certainly
shown its political worth. While still largely ineffective at the individual
level of Congress, the Lava-Jato scandal created the conditions for the
impeachment of a president, and with it, a regime change – not just a
change of governments, but, as a corollary of that change, a regime change
in crucial policy aspects. These policy changes have included a redrawing
of the role and the size of the state, education, reforms of social security and
pensions, and a wave of privatizations, to name a few. Thus, the power of
publicity does not necessarily reside at the individual level, where a scandal
might mean the end of a particular political career. At the institutional level,
the scandals have changed the framework of the intersection of public
information flows in media systems and political systems, and this has
repercussions, as described in the previous chapter, for public opinion, stock
markets, and the mobilization of political resistance. In the wake of the
Lava-Jato scandals, impeachment, and leaks, a tale often told about Brazil
persists: 97 percent of the population thinks the government governs for
personal gains and only 13 percent of citizens are satisfied with democracy
(Latinobarómetro 2017: 9–10), political corruption is perceived to be
rampant and increasing (Pring 2017), and the disapproval of Congress is the
highest since the return to democracy (Datafolha 2017). Trust in
government and political institutions is very low, and the percentage of
voters declaring their intention to vote blank in the 2018 general elections
was record-high less than a year before the elections.

Post-Politics in the Light of Leaks and Scandals

The profound effects on Brazilian democracy of the Lava-Jato probe may
generate insights into the effect of political scandals and leaks beyond the
Brazilian border. First, based on the case studies of this book, I suggest that
leaks should be considered a part of a broader post-political climate. I
would argue that post-politics comprise not just the projection of alternative
facts or the viral spread of damaging (fake or real) news, but also the use of
bulk information to produce institutional rupture. Like the leak of US
Democratic National Committee emails prior to presidential nominating
convention in 2016, or the devastating allegations made against François
Fillon in the French presidential elections in 2017, the Brazilian case shows
how political or institutional actors become pitted in antagonistic relations
through leaks – and for sure, neither the Lava-Jato leaks nor the Fillon
allegations cropped up randomly. Both are intentional, designed to wreak a
certain amount of havoc in a political arena. This havoc, however, rapidly
undermines the stability of the political institutions and settings in which
they work. Since the leaks of Lava-Jato have made every politician a
possible target, each politician is prepared to transgress the normal rules of
democratic conduct if need be. These transgressions have become weapons,
pure and simple, in an ongoing institutional conflict, and we may expect
more conflicts of the sort in the future.
The second consequence of leaks in post-political environments is the
ever-increasing suspicion cast upon political actors. In cases such as Brazil,
that suspicion arises for very good reasons, to be sure. But leaks, through
the technological and organizational setup of whistleblowers, data
journalism, servers, etc., simultaneously configure disclosure and secrecy in
order to protect the sources. The leaking is naturalized as the morally
correct thing to do, maintained and protected as a private, heroic act, and
rarely questioned in its political intent. The new mediated visibility
(Thompson 2005) of political actors does not extend to the individuals or
collectives who work through leaks, and the transparency demanded of
politicians is not reflected in the accountability of leaky organizations
(Wahl-Jørgensen 2014). But public information is not necessarily
transparent or innocent, and media professionals partaking in the leaks-and-
journalism assemblage should be wary of mindlessly recycling leaked
information. In this theater of shadows, the power of leaks not only consists
in the bypassing of normal media gatekeepers and fact-checking, but also
provides a slippery entry point for the post-politics of allegations,
denunciations, and leaks thereby to underpin and strengthen certain kinds of
defamatory speech about politicians easily detectable in social media.
Finally, while leaks do not necessarily contribute to the post-truth
problematic of contemporary democracy, they do in fact play into the
distinction of knowledge types by projecting the leaked information as “a
superior mode of knowledge, a cultural signifier of unmediated objective
information” (Hansen and Flyverbom 2014: 2). Leaks, in the public sphere
of media disseminating these leaks, occupy a privileged space relative to
the claims made by political actors, who are rendered suspect a priori by
virtue of their presence in the leaked information. The apparently “hard”
facts found in leaked files command the top position of the hierarchy of
public knowledge, and leaks thereby create the conditions of qualifying,
disqualifying, and delegitimizing information.
Theoretical Implications and Remaining Questions
The post-political climate of leaks eclipses ideology, but also challenges
key theoretical terms. Where earlier literature saw the positive contributions
of leaks to public accountability, especially in global context of fiscal
havens and international networks of corrupt and secretive offshore shell
companies, I would like to draw attention to the dynamic of leaks in terms
of information flows. Leaks may shake a web of accountability institutions,
but the media limelight can be attractive not just to kick-start investigation
but also for political instrumental reasons. This means that the
accountability theory model (Mainwaring and Welna 2003, Porto 2011)
needs to be reinterpreted as inserted into Realpolitik and struggles for
symbolic capital (Thompson 2000). The vertical and horizontal dimensions
are intersected by vectors of political force that bend and stretch the web of
accountability. An investigation or a trial, with solid evidence or without
substance, can be of utmost value – not for democracy nor for
accountability, but to eclipse other cases or as a stepping stone for
launching political candidatures. Information flowing from such legal
actions can drive the media agenda and contribute to the formation of
informational cascades between the agenda-setting media outlets. Watchdog
media becomes a hollow term, then, not because of political or ideological
bias, nor because of singular instances of recycled fake news, but because
the emphasis of the infostorm redirects attention and precipitates pluralistic
ignorance in public spheres (Wiewiura and Hendricks 2017).
With the emergence of such infostorms, the actual content of news items
may be subsumed in the avalanche of signals. For theoretical and
methodological purposes, this means a reorientation of research from media
texts to media spheres, from news frames to agendas, or from micro to
macro. While particular texts – such as initial disclosures or leaks – may
have much agenda-setting influence, derived social proof only becomes
manifest when the same signal is retransmitted, disseminated, and recycled
again and again. A political crisis may lurk on the horizon, only to flare up
when herds of political actors conjure the crisis into being, and the self-
fulfilling prophecy of crisis becomes reality. This was the case in Brazil,
when declining economic indicators such as export revenue and tax revenue
became exacerbated by the cascades of news on political turmoil and the
derived effects on investors, stock markets, and exchange rates. A political
crisis, in this sense, could be conceptualized as a bubble: a rapidly growing,
bootstrapping informational phenomenon.
Scandals and leaks about corruption are effective catalysts for the
phenomena discussed here; infostorms, cascades, and bubbles can be set in
motion by when media actors flock around a case of political transgression
and act in a herd-like manner (Hendricks and Hansen 2016). Departing
from this observation, the case studies of this book may generate an
important question for media system theory: What does the uniformity (or
pack-mindedness) of a country’s media agenda mean to political issues, and
does media research have adequate conceptual tools to establish the degree
of group-think in media systems? These questions cut across newsroom
ethnography, journalism, and agenda-setting studies, and could in time open
up new avenues of research on agenda fragmentation, pluralization, or
polarization of public spheres, and democratic participation in a wider
sense. If democracies are founded upon the premise of access to transparent
and dependable information on affairs of state, then it is vital to establish
how media institutions funnel and filter information, and to what extent
such filtered information trickles down into the blogosphere, social media,
and ultimately civic society.
Perhaps media institutions should not be conceptualized (as
accountability theory commonly does) as one institution among others in
the web of accountability? When the attention and visibility staged in the
media generate distorted policy-maker agendas, driven by backstage
interests, I fear that media act rather as political stakeholders. The same
doubts could be raised for the new variety of indignant online movements
that capitalize on the information flows about political transgression to
produce hateful speech toward entire segments of populations and
ideologies associated with these segments.
Aside from theoretical implications for accountability and scandal
theory, media system research, and agenda-setting studies, several practical
implications of the Brazilian case are left at the end of this book. First, it is
important to find models that can ensure and qualify the separate
independence of society’s watchdogs – understood here to be not only
media but also courts and prosecutors – since the combination of lawfare
and scandal creates a toxic and opaque political atmosphere. How to avoid
that the barking of one watchdog instills a frenzy in the rest of the pack,
without restricting freedom of speech and Constitutional procedures? In less
poetic terms, the question remains as to whether and how the feedback
mechanisms of Brazil’s accountability and media institutions could be
dismantled and reassembled into something more democratic, equilibrated,
and informative. Second, the imbalance of the Brazilian system’s checks
and balances is on display in the cases presented here, but a viable way to
improve the checks and balances seems a long way off as long as the
Congress is situated comfortably in a dead zone of accountability; largely
untouched by the corruption investigations and largely unwilling to let go of
elite privileges and the possibility of obstructing interventions from the
judiciary branch.
Finally, the data presented in this book can be mined and interpreted in
many more ways, and I welcome suggestions and attempts to do so. As
most parts of the previous chapter’s dataset on media coverage and leaks
are published in the public domain, scholars from diverse fields can perhaps
think of novel ways of reassembling the complex and remarkably skewed
interplay of information flows, media dissemination, and corruption
allegations in Brazil. To be sure, the possibilities of detecting patterns and
gaining insights are not limited to the brief content analyses provided in the
previous chapter, and so I invite scholars to interrogate and expand the data,
as well as to introduce new questions and approaches that can expand the
scope of the present inquiry.

References
Almeida, R. de (2016) A Sombra do Poder. Os Bastidores da Crise Que Derrubou Dilma Rousseff.
Rio de Janeiro: Editora LeYa.
Amorim Neto, O. (2007) “O poder executivo, centro de gravidade do sistema político brasileiro.” In
L. Avelar and A.O. Cintra (eds.), Sistema político brasileiro: uma introdução (123–133). São
Paolo: UNESP Press.
Arantes, J. (2016) Relatório. Brasília: Câmara dos Deputados. Retrieved from
www2.camara.leg.br/atividade-legislativa/comissoes/comissoes-temporarias/especiais/55a-
legislatura/denuncia-contra-a-presidente-da-republica/documentos/outros-
documentos/ParecerDep.JovairArantes.pdf.
Avritzer, L. (2016) Impasses da democracia no Brasil. Rio de Janeiro: Civilização Brasileira.
Balán, M. (2011) “Competition by Denunciation: The Political Dynamics of Corruption Scandals in
Argentina and Chile” Comparative Politics, vol. 43(4), 459– 478.
Boydstun, A. (2013) Making the News. Politics, the Media, and Agenda Setting. Chicago, IL:
University of Chicago Press.
Cabral, E.D.T. (2017) “Mídia Concentrada no Brasil: Até Quando?” Revista Latino-Americana de
Ciencias de la Comunicación, vol. 13(24), 49–59.
Comaroff, J. and Comaroff, J. (eds.) (2006) Law and Disorder in the Postcolony. Chicago, IL:
University of Chicago Press.
Datafolha (2017) Avaliação do Congresso Nacional, novembro 2017 – PO 813942. São Paulo:
Datafolha.
Falcão, J., Hartmann, I.A., and Chaves, V. (2014) III Relatório Supremo em Números: o Supremo e o
tempo. Rio de Janeiro: Escola de Direito do Rio de Janeiro da Fundação Getulio Vargas.
Feres Júnior, J. and Sassara, L. (2016) “O cão que nem sempre late: o Grupo Globo e a cobertura das
eleições presidenciais de 2014 e 1998” Revista Compolitica, vol. 6(1), 30–61.
Galtung, J. and Ruge, M. (1965) “The Structure of Foreign News” Journal of Peace Research, vol.
2(1), 64–91.
Habermas, J. (1989 [1962]) The Structural Transformation of the Public Sphere. Cambridge, MA:
MIT Press.
Hansen, H.K. and Flyverbom, M. (2014) “The Politics of Transparency and the Calibration of
Knowledge in the Digital Age” Organization, vol. 22(6), 872–889.
Hendricks, V. and Hansen, P.G. (2016) Infostorms. Why Do We “Like”? Explaining Individual
Behavior on the Social Net. 2nd edition. London: Springer.
Kristensen, N.N. and Mortensen, M. (2015) “Amateur Sources Breaking the News, Metasources
Authorizing the News of Gaddafi’s Death: New Patterns of Journalistic Information Gathering and
Dissemination in the Digital Age” Digital Journalism, vol. 1(3), 352–367.
Latinobarómetro (2017) Informe Latinobarómetro 2017. Santiago de Chile: Corporación
Latinobarómetro.
Mainwaring, S. and Welna, C. (eds.) (2003) Democratic Accountability in Latin America. Oxford:
Oxford University Press.
Pederiva, J., Alves, D., and Rincon, F. (2016) Laudo Pericial. Brasília: Senado Federal. Retrieved
from www12.senado.leg.br/noticias/arquivos/2016/06/27/laudo-da-junta-pericial.
Pereira, C. and Mueller, B. (2000) “Theory of Executive Dominance: The Committee System in the
Brazilian Congress” Revista Brasileira de Cie^ncias Sociais, vol. 15, 45–67.
Peruzzotti, E. (2006) “Media Scandals and Social Accountability. Assessing the Role of Scandals in
Argentina.” In E. Peruzzotti and C. Smulovitz (eds.), Enforcing the Rule of Law. Social
Accountability in the New Latin American Democracies (249–271). Pittsburgh, PA: University of
Pittsburgh Press.
Porto, M. (2011) “The Media and Political Accountability.” In T. Power and M. Taylor (eds.),
Corruption and Democracy in Brazil (103–126). South Bend, IN: University of Notre Dame Press.
Power, T. and Taylor, M. (eds.) (2011) Corruption and Democracy in Brazil. South Bend, IN:
University of Notre Dame Press.
Pring, C. (2017) People and Corruption: Latin America and the Carribbean. Berlin: Transparency
International.
Seligson, M. (2002) “The Impact of Corruption on Regime Legitimacy: A Comparative Study of
Four Latin American Countries” The Journal of Politics, vol. 64(2), 408–433.
Smulovitz, C. and Peruzzotti, E. (2000) “Social Accountability in Latin America” Journal of
Democracy, vol. 11(4), 147–158.
Taylor, M. (2011) “The Federal Judiciary and the Electoral Courts.” In T. Power and M. Taylor
(eds.), Corruption and Democracy in Brazil (162–183). South Bend, IN: University of Notre
Dame Press.
Thompson, J.P. (2000) Political Scandal: Power and Visibility in the Media Age. Cambridge: Polity
Press.
Thompson, J.P. (2005) “The New Visibility” Theory, Culture & Society, vol. 22(6), 31–51.
Valente, R. (2016) “Em diálogo, Jucá fala em pacto para deter avanço da Lava Jato” Folha de S.
Paulo, p. A4, May 23.
Villaverde, J. (2016) Perigosas pedaladas. São Paulo: Geração Editorial.
Wahl-Jørgensen, K. (2014) “Is WikiLeaks Challenging the Paradigm of Journalism? Boundary Work
and Beyond” International Journal of Communication, vol. 8, 2581–2592.
Waisbord, S. (2000) Watchdog Journalism in South America. News, Accountability, and Democracy.
New York: Columbia University Press.
Walgrave, S., Boydstun, A., Vliegenthart, R., and Hardy, A. (2017) “The Nonlinear Effect of
Information on Political Attention: Media Storms and U.S. Congressional Hearings” Political
Communication, vol. 34(4), 548–570.
Wiewiura, J. and Hendricks, V. (2017) “Informational Pathologies and Interest Bubbles: Exploring
the Structural Mobilization of Knowledge, Ignorance, and Slack” New Media and Society, Article
first published online: January 10, 2017 – https://doi.org/10.1177/1461444816686095.
Wink, G. (2017) “Judicialização da política ou politização da justiça? Uma Análise Crítica do
Discurso do Libelo Acusatório contra a presidenta Dilma Rousseff no processo de impeachment”
Moara, vol. 47, 152–177.
Zalis, R. (2015) “E o governo mal começou …” Veja, edition #2417, March 18.
Afterword1
Lula and Lava-Jato

 
The current political crisis Brazil is facing has a multifaceted nature, which
can only be grasped through a multidisciplinary approach. The coup against
Dilma Rousseff itself is a point of contention among political scientists, and
between them and law scholars, most supporting the idea of a political coup
while a few others resist employing the term, claiming that everything was
done according to legal procedure and that no major democratic institution
was harmed – a rather doubtful opinion. The event was termed by some
commentators a judicial-media-parliamentary coup because it was brought
about through the synergetic endeavors of both houses of the legislative
branch – the Chamber of Deputies and the Senate –, the judicial system,
which promoted the Lava Jato Operation on one side while allowing the
case against Rousseff to progress despite its lack of substance, and the
media, which fiercely supported the pro-impeachment movement and
initiate a campaign against Rousseff, her government, the Workers’ Party
(PT) and Luiz Inácio Lula da Silva.
From the perspective of media studies, the behavior of Brazil’s major
news outlets was truly staggering. Putting things in perspective, the fact that
the media acted against PT was not at all a surprise. Numerous academic
studies have shown, since the return of the democratic regime in the 1980s,
that Brazil’s big media has shown a marked bias against the left,
particularly PT and Lula (Aldé, Mendes, and Figueiredo 2007, Azevedo
2017, Feres Júnior and Sassara 2018, Kucinski 1998, Miguel 1999, Miguel
and Biroli 2011, Rubim and Azevedo 1998). In election after election, they
have employed a variety of means to undermine left-wing candidacies and
promote its conservative opponents, from the manipulation of television
debates to heavily loaded framing of key political themes, to agenda-setting
that exploits bad news for the left while hiding information that could harm
conservative candidates, to highly speculative news that border what today
we call fake news, to the outright fabrication of scandals.
Most of the academic work done in the field of media and politics in
Brazil focused on elections. Thus, the impression one might get by
becoming acquainted with this literature is that the media behaves in such
manner partly because of the highly politicized climate of the election
period. That is certainly no excuse for the general behavior of Brazilian
media during elections, but there hadn’t been much evidence about its
behavior in non-electoral years. This all changed with project
Manchetômetro,2 which was established to analyze the media coverage of
the 2014 presidential election but then continued until this day, producing
sentiment analysis of the big media coverage of chief political figures,
parties, and themes, and of the economy.
Our data shows, for example, that after her 2014 victory in the
presidential election, Dilma Rousseff enjoyed the opposite of a honeymoon
(Feres Júnior and Sassara 2018), a concept some American authors define
as a period of truce between the newly elected president and the political
opposition, including the media (Dominguez 2005, Johnson 1983). The
number of negative news items about her doubled in January, the first
month of her second term, and practically tripled two months later, while
her neutral coverage declined. Rousseff’s negative coverage kept on
peaking in September and December 2015, when it reached around 300
negative news items per month in the three newspapers and Jornal
Nacional – in other words, an average of ten negative items a day.3 In
March 2016, the same month Lula was unlawfully forced to testify by
Judge Sergio Moro, and when the same judge, Moro, illegally leaked to
Grupo Globo a wiretap containing a phone conversation between Rousseff
and Lula, Rousseff’s negatives reached the historical mark of 373. We can
see here what this volume’s author, Mads Damgaard, termed lawfare in
Chapter 4, borrowing the concept from anthropology and from the writings
of Lula’s attorney, Cristiano Zanin: the strategic use of politically loaded
legal procedures combined with biased media coverage in the pursuit of
political ends. President Dilma Rousseff was actually the first target. The
intensity of her negative coverage only subsides when she is removed from
power.
 A similar pattern of coverage is shown when we examine the variable
federal government. The figures are even larger. In March 2015 the
government received almost 500 negative items. This massacre continued
on a similar level until Dilma Rousseff was removed from power in May
2016. Proving a point Mads Damgaard makes quite well in this book,
despite the numerous accusations of corruption against Michel Temer and
the meager popular support he enjoys, the media gave his government a
perennial honeymoon, with a brief interruption in May 2017 when a
recorded conversation between him and Joesley Batista, the owner of the
meat-packing giant JBS, in which the president condones bribery, and a
video showing one of his assistants receiving a suitcase full of cash were
leaked to the media. After a couple of months, however, the scandal was
practically forgotten by the news outlets, which continue to treat Temer in a
very benign manner, fully supporting his agenda of neo-liberal economic
reforms.
But the replacement of Dilma Rousseff by the neo-liberal Temer is just
half of the story. The other half, probably the most important one in
political terms, is the persecution of Lula by the same judicial/media-
organizational dispositive that produced the impeachment. The behavior of
public prosecutors, judges, and Supreme Court justices has been so flawed
and politically loaded that it is impossible to speak of it here without
extending this text beyond what is advisable. It suffices to say that Moro
sentenced Lula even without sound material proof of his guilt, that the
verdict was affirmed by the appellate court, which extended his sentence
just enough to force his imprisonment, and that the Supreme Court refused
to grant him a habeas corpus even though the Constitution expressly
forbids imprisonment until all possibilities of appealing are exhausted,
which is Lula’s case.
Lula’s political stature can hardly be overstated. During the last 30 years
he has been a chief protagonist in Brazil’s presidential elections. He passed
to the second round of the 1989 election, to be defeated by Fernando Collor
the Mello after a massive media campaign that accused PT of being
involved in the kidnapping of a famous businessman, which turned out to
be false, and the broadcasting of a heavily edited and manipulated version
of the last debate of the campaign by TV Globo’s Jornal Nacional, which
portrayed Lula losing the debate to Collor. Lula was also the runner-up in
the following two presidential elections, 1994 and 1998, and finally won in
2002 and 2006, defeating PT’s main political adversary, PSDB. With his
popular approval rating approaching 80 percent, he managed to elect Dilma
Rousseff, his hand-picked successor, twice, in 2010 and 2014, again
defeating PSDB candidates. Now he keeps a clear lead in 2018’s
presidential race but will most probably be unable to run given his
conviction and detention.
In other words, the coup did not stop at Rousseff, cancelling the popular
mandate transferred to her through popular vote: it is going all the way to
block the majority of the people to choose their favorite candidate in the
next presidential election. But lawfare is not complete if we do not factor in
the role of the media. And in the case of Lula it has exceeded itself. The
numbers of his news coverage, produced by Manchetômetro, are
staggering. Without holding any office, Lula started to become the target of
a fierce media smear campaign about the time Dilma Rousseff was starting
her second term. In January 2015 he received 50 negative news items in our
database. This figure almost doubled in the next month and by October it
had already tripled. But nothing compares to the peak of March 2016, when
Moro had Lula forcefully conducted to testify. That month Lula received
352 negative items from the three newspapers and Jornal Nacional.
The numbers can hardly be more extreme. If we take into account only
the opinion items (editorials, op-ed articles, and signed columns) of the
three newspapers, Lula received from January 2015 to April 2018 an
average of 66 negative pieces per month. This is almost eight opinion
pieces per day per newspaper. Since his prosecution was controversial,
given that besides all its political import it also involved important
constitutional matters, one would assume that the media cared also to
publish pieces that would comment on the facts from perspectives other
than those of the accusation. But that is not the case. Positive pieces, that is,
texts that put into question the dealings of Moro, the prosecutors, and were
keener to the arguments of Lula’s defense averaged 3.5 throughout the
period from January 2015 to April 2018: almost 20 times less frequent than
negative texts. Just to keep the comparison similar, their frequency was
close to one piece per newspaper per month.
Quantitative analysis such as the one I have been expounding here gives
the reader a rather cold and detached image of what really goes on. So, we
will delve a little bit more into details to flesh out the analysis. Right after
Lula was arrested, his attorneys filed a habeas corpus petition at the
Supreme Court claiming that the arrest was illegal because the constitution
forbade it before the case went through the three stages of Brazil’s judicial
system: first instance, regional appellate courts, and the federal Supreme
Court. How did the big media approach this event? Let’s see how the two
most important Brazilian newspapers, O Globo and Folha de S. Paulo,
covered the Supreme Court ruling on the day of the verdict, April 4, 2018.
The cover of O Globo, the newspaper of the family Marinho, owners of
Grupo Globo, the largest media conglomerate in Latin America, makes
quite an impact. Three-fifths of its area is covered by items referring to the
Supreme Court ruling. The headline, in huge letters on the top of the page
reads: “For Dodge, the excess of appeals annihilates justice.” It is a
reference to a declaration given by the Chief Federal Prosecutor Raquel
Dodge. Below the headline there is a title that says that the Supreme Court
will rule over intense pressure. At the bottom of the column there is another
title saying that Commander-in-Chief of the Army, General Vilas Bôas,
tweeted that impunity should be rejected at all costs, a clear threat against
the Supreme Court in case they decide to suspend Lula’s arrest. Then, on
the right-hand side of the page there is a column containing snippets of text
pointing to opinion pieces inside the newspaper. The top one is the editorial
entitled “Lula at the Supreme Courts represents impunity.” This is followed
by a teaser for the column of Miriam Leitão, a known detractor of President
Lula, with a threatening tone, saying that the Court will give Lula the
habeas corpus. The next snippet is authored by Merval Pereira, another
self-declared enemy of Lula and his party, saying that Justice Rosa Weber
will have the responsibility to decide upon such an important matter, that is,
putting pressure on her to decide against Lula. The next snippet is by Elio
Gaspari, which has the clear intent of delegitimizing Lula’s habeas corpus
by saying that poor convicts do not have access to so many appeals.
Underneath these columns there is a large photograph that takes the entire
center area of the first page, showing people in São Paulo protesting against
Lula’s habeas corpus. The title reads “pressure from the streets,” followed
by the information that “thousands went to the streets in at least 23 states
and the Federal District to ask the Supreme Court to confirm the arrest of
president Lula.” This is the whole coverage of the event on the first page:
That is, O Globo did not allow in its first page a single mention of an
opinion in favor the habeas corpus.
 The opinion pages are no different. The editorial, as its title announces,
is fiercely against the habeas corpus, arguing that its acceptance by the
court will benefit all corrupt politicians. In the same page there is the article
by O Globo’s columnist Élio Gaspari saying that the habeas corpus is a
privilege of the elites, not enjoyed by the poor. The following opinion page
contains a text by Zuenir Ventura, which despite its vague tone argues that
there are reasons to believe that if the habeas corpus is accepted, the fight
against corruption will be endangered. At the bottom of the page, an article
by Jorge Maranhão, the director of a right-wing NGO Instituto de Cultura
de Cidadania, who argues that the Supreme Court cannot overrule the will
of the people in the streets. This phrase is in line with the picture on the first
page of the newspaper, and both are examples of how mainstream media
often approaches fake news. Simply, there were not masses of people in the
streets demanding the Supreme Court to withhold Lula’s imprisonment.
Finally, there is a cryptic text by Marco Lucchesi which does not touch
directly the subject in question but suggests that there is a new ethical
Brazil rising from the wreckage of the old society, a soteriological and
messianic discourse very much in line with that of the evangelical
prosecutors of Lava-Jato.
The first page of Folha de S. Paulo is quite similar to that of its
counterpart from Rio de Janeiro. The space dedicated to the topic is a little
smaller, but still quite sizable: roughly half the page. The titles are inverted.
The highlight of the headline is given to the tweet of the commander-in-
chief of the Army, who warned the Supreme Court against accepting the
habeas corpus. To Rachel Dodges’s declaration against the excess of
appeals, the newspaper gives a smaller box of text. Also similar to the first
page of O Globo is the disposition of a column on the right side of the page
displaying snippets leading to opinion pieces inside the newspaper. The first
one comments on the reaction of the general’s comment in the Supreme
Court. The second states that the government of Paraná is already finding a
proper place to incarcerate Lula. The third snippet, by columnist Hélio
Schwartsman, is a cryptic note saying that the Supreme Court’s decision
will not alter the political context. Next is the same text by Élio Gaspari,
who published in both newspapers. And finally the editorial of Folha de S.
Paulo, entitled “It is not about Lula,” is announced.
The first opinion page is topped by the same editorial. The point is clear:
if the Court accepts the habeas corpus, corrupt politicians would avoid
prison. Thus, the justices must reject it. On the right-hand side of the
editorial, there are three columns. The top one by Schwartsman replicates
the cryptic tone of the snippet, but is clearly ironical toward Gilmar
Mendes, a STF judge who switched his previous opinion to vote in favor of
the habeas corpus. The second text, by Bruno Boghossian, accuses the
court of privileging Lula, suggesting that it should reject the habeas corpus.
And finally, columnist Ruy Castro openly advocates the imprisonment of
the ex-president in his article, calling him an ingrate. The second opinion of
Folha de S. Paulo contains a section called Trends and Debates that often
pitches contrasting opinions on the same topic. This time the topic was
Lula’s request to the Supreme Court. The text on the top is authored by the
Prosecutor-General Raquel Dodge herself, defending the position that
Brazil should follow other advanced democracies around the world and
adopt the imprisonment of people convicted after the first appeal. The text
below is authored by a lawyer named Augusto Arruda Botelho, who
defends the Constitution against the innovation forcefully advocated by the
Lava-Jato supporters. There is one important detail, however. Dodge’s text
is not only on top of his, but it is three times larger.
Completing its second opinion page, Folha de S. Paulo has another fixed
section, the Readers’ Panel, which reproduces emails sent from alleged
Folha’s readers. Four of them deal with the habeas corpus case. The first
simply states that a 2016 statute interprets the Constitution allowing for
Lula to be arrested after the sentence if confirmed at the appeals court. The
second is supportive of STF President Cármen Lúcia, who not only voted
for repealing the habeas corpus but also arranged the Court’s agenda in a
way to harm Lula and his defense. The third email echoes the main
argument of the editorial and of Gaspari’s article, that the right to a third
instance appeal in the Constitution only benefits the rich. And finally, a
fourth email is a short invective against Lula, saying that he only deceives
intellectuals that are not able to think.
This brief summary of what the two main newspapers in the country
have published in their first pages and opinion pieces in the day of a
momentous ruling in the Supreme Court reveals how profoundly slanted
Brazil’s big media is regarding Lula and PT and how it is deeply invested in
promoting certain legal views in tandem with sectors of the judiciary and of
the MPF, a practice that is the core of what Damgaard has correctly
identified as lawfare. The Brazilian media promotes a massacre in the realm
of public opinion and in order to do that it disregards the most basic
principles of professional journalism, first of all to give a fair and balanced
account on facts and opinions, and to allow for a plurality of opinions to be
expressed.
Brazilian democracy is nowadays tremendously sick, mostly because its
institutions of control have been colonized by politics, while the political
institutions have been severely devalued in the forum of public opinion –
the presidency and Congress are the least trusted institutions in the country
(see Chapter 6). The media behaves like a political party, the judiciary and
public prosecutors are politically motivated while politicians and political
parties seem stunned and unable to provide solutions that would re-energize
the polity. The judicial institutions will not fix themselves: They are an elite
of highly paid state bureaucrats that enjoy legal protections of all sorts and a
great deal of institutional freedom without having to be accountable to
almost any institution. The media also enjoy extreme freedom from libel
laws, that practically do not exist, and from regulations in general: Brazil
has one of the most concentrated and unregulated media systems in the
world. The correction must come from outside these control institutions. It
must come from the politics itself.
–João Feres Júnior, PhD
Director of Instituto de Estudos Sociais e Políticos at Universidade do
Estado do Rio de Janeiro

Notes
1 Editor’s note: This Afterword, provided by João Feres Júnior, reflects upon the Brazilian media
system’s coverage of the final steps of the trial that led to the incarceration of ex-President Lula
in Curitiba on April 7, 2018.
2 Manchetômetro, literally “headline meter,” is a research project conducted by the Laboratory on
Media and the Public Sphere (LEMEP), led by me, focused on the daily publication of data
concerning the news coverage of politics and the economy done by the most prominent Brazilian
media outlets: newspapers Folha de S. Paulo, O Globo, Estado de S. Paulo, and television news
program, Jornal Nacional, which is broadcast by TV Globo.
3 Manchetômetro’s dataset covers the first page and two opinion pages of each newspaper, which
contain editorials, signed columns, and op-eds, and the entire content of Jornal Nacional, the
most-viewed TV newscast in Brazil.

References
Aldé, A., Mendes, G., and Figueiredo, M. (2007) “Tomando partido: imprensa e eleições
presidenciais em 2006” Política e Sociedade, vol. 10, 153–172.
Azevedo, F.A. (2017) A Grande Imprensa e o PT (1989–2014). São Carlos: Editora UFSCar.
Dominguez, C.B.K. (2005) “Is It a Honeymoon? An Empirical Investigation of the President’s First
Hundred Days” Congress & the Presidency, vol. 32(1), 63–78. doi: 10.1080/07343460509507697.
Feres Júnior, J. and Oliveira Sassara, L. de (2018) “Failed Honeymoon: Dilma Rousseff’s Third
Election Round” Latin American Perspectives, vol. 45(3), 224–235. doi:
10.1177/0094582x18767429.
Johnson, K.S. (1983) “The Honeymoon Period: Fact or Fiction?” Journalism Quarterly, vol. 60(3),
869–876.
Kucinski, B. (1998) A síndrome da antena parabólica: ética no jornalismo brasileiro. 1a edn. São
Paulo, SP, Brasil: Editora Fundação Perseu Abramo.
Miguel, L.F. (1999) “Mídia e eleições: a campanha de 1998 na Rede Globo” Dados, vol. 42(2), 253–
276.
Miguel, L.F. and Biroli, F. (2011) “Meios de comunicação de massa e eleições no Brasil: da
influência simples à interação complexa” Revista USP, vol. 90, 74–83.
Rubim, A.A.C., and Azevedo F.A. (1998) “Mídia e política no Brasil: textos e agenda de pesquisa”
Lua Nova: Revista de Cultura e Política, vol. 43, 189–216.
Index

 
Page numbers in bold denote tables, those in italics denote figures.
accountability 3, 9–12, 53, 136, 188; horizontal accountability 10, 87, 188; societal accountability 10,
101; vertical accountability 10, 93, 188
agenda-setting 143, 154–155, 180, 186
Amaral, Delcídio do 56, 68, 107, 117, 124–125, 163, 165

Balán, Manuél 13, 43, 70, 187

bandwagon effect 148–150, 169–170
Banestado case 100
Benjamin, Herman 107–111
Bikhchandani, Sushil 15, 146, 149
Bovespa (stock market) 175, 176, 177
Brindeiro, Geraldo 90

Calheiros, Renan 115, 123, 165

Câmara dos Deputados 54, 63–64, 110–114
Cardoso, Fernando Henrique 5, 68, 120, 159
CartaCapital 26, 123
cascade see informational cascade
cassation of mandate 102, 105–108, 110–114, 128
checks and balances 12, 19, 31, 61, 94, 104–105, 110, 117, 136, 186, 194–195
coalition formation 70–72, 130
coalitional presidentialism 3, 70, 186
Collor de Mello, Fernando 4, 31–32, 64, 68, 98, 165, 172, 206
Congress 3, 8, 30, 32, 54–55
Conselho da Ética 110–114, 118
Constitution of Brazil (1988) 32, 69, 80n1, 89, 95, 117, 137n3, 137n5
Comaroff, Jean and John 132–133, 138n10, 187
coroneis eletronicos 28
Corporatism 90
corruption (definition) 34, 158
Costa, Paulo Roberto 7, 54
coup d’état 24, 112, 179–180, 186–195, 204–206
Cunha, Eduardo 55, 63, 110–114, 160, 161

Dallagnol, Deltan 52, 91, 101, 131

Datafolha 100, 118, 155, 175
delação premiada see plea bargain
Dodge, Raquel 90, 125, 137n1
doleiro 6

electoral reform 127, 187, 197

Época 16, 26–27, 33, 100
Estado de São Paulo 16, 24–27
Estadão (online news platform) 27, 174

Fachin, Luiz Edson 96, 125

fake news 144
Ficha Limpa (law of the Clean Slate) 88, 135
FIFA football leaks 44
Folha de São Paulo 16, 24–27
foro privilegiado 64, 66, 97, 127, 135
Foucault, Michel 75
framing 14–15, 18, 23, 33, 35–37, 50, 123, 148, 151

gatekeeping 34, 47, 53, 73, 79, 147, 169

Goffman, Erving 49, 75
golpe see coup d’état
Grupo Globo (broadcasting network, media conglomerate) 16, 29–30, 33, 123

Hansen, Pelle G. 13–16, 143, 145–146, 150, 187

Hendricks, Vincent 13–16, 143, 145–146, 150–155, 187, 200
Hoffmann, Gleisi 98–99, 125
hype cycle 168

impeachment 127–128, 159, 188–195

information 12, 150
informational cascade 13–15, 145–150, 169, 172, 187
information flow 9–13, 147, 149, 170–173, 179, 181
infostorm 143, 200
institutions 61, 93
interaction 62
Intercept, The 58
intertextuality 146, 153, 155, 158
IstoÉ 16, 26, 31, 33, 56, 100, 156

Janot, Rodrigo 36, 90–91, 98, 125, 137n2

JBS case 60, 116, 128
June 2013 protests (Jornadas de Junho) 6
Jucá, Romero 72, 123, 165
judicialization 86–87, 131, 133
judge rapporteur see ministro relator

Lava-Jato 6–9, 54–60

lawfare 86, 131–135, 187
leaks 51, 54, 73–77
Lucia, Carmen 96, 117
Lula da Silva, Luiz Inácio 5, 56, 118–126, 161, 162, 172, 206

Maia, Rodrigo 114, 128

Maranhão, Waldir 112–114, 193
media storm 119, 143–144, 154, 187
media system 23–33, 156
Mendes, Gilmar 67, 96, 99, 106–109, 120, 134–135, 137n2
Mensalão 5, 87, 99, 104, 148, 174
Miguel, Luis Felipe 9, 15, 26, 48, 148, 204
ministro relator 96–97, 106, 174
moral panic 144
Moro, Sérgio 52, 65–66, 99–101
MPF (Ministério Público Federal) 89, 134

narrative 18, 23, 37, 112, 136, 152–153, 180

Nassif, Luis 65
Neves, Aécio 60, 116, 138n8, 165, 174
new institutionalism 61, 95
news values 170–171, 187
newsworthiness 48–51, 152

Odebrecht (company) 57, 97, 120, 135, 153

Odebrecht, Marcelo 57, 107, 122, 137n6
O Globo (newspaper) 16, 24–27

Palocci, Antônio 94, 122, 125, 129

Panama papers 2, 44, 119
Paradise papers 2, 44
Pedaladas (fiscal delays) 104, 127, 166, 188–191
Peruzzotti, Enrique 11, 186
Petrobras 7–9, 54
PF see Polícia Federal
PGR see Procurador-Geral da República
Pinheiro, Leo 55–56, 64–65, 119–121
plea bargains 7–8, 92
PMDB 55–59, 70
Polícia Federal (PF) 91
Porto, Mauro 11–12
PP 70–72, 80n2, 98, 104, 110, 125, 129, 137n4
preventive arrests 8, 52, 93–94, 97, 99, 114, 119, 120, 134
print runs 24, 25
process tracing 62
Procurador-Geral da República (Prosecutor-General) 89–91
protests 100, 134, 178–180
PSDB 54, 59, 67, 97, 105–109, 117, 135, 174, 180, 206
PSOL 63–64, 111, 129
PT see Workers’ Party
public signal 14–15, 37, 73, 146–147, 152, 158, 187, 190

Rousseff, Dilma 5, 54, 186–194

Sarney, José 4, 30, 58, 68

scandal 12–13, 30–35, 167–168, 170–172, 186
Senate 70, 90, 110, 115
Serra, José 59, 89, 118, 135, 196
STJ, Supremo Tribunal da Justiça 102, 137n5
STF, Supremo Tribunal Federal (Supreme Federal Court) 95
source relations 12, 52–53

TCU see Tribunal das Contas da União

Temer, Michel 108, 127–128, 166
Tijolaço 27, 58
Thompson, John 76, 144, 171
Toffoli, Dias 64–65, 96, 106
transparency 77–80
Transparency International 77
Tribunal das Contas da União 103, 190–191
Tribunal Superior Eleitoral (Supreme Electoral Court) 88, 101–103
TSE see Tribunal Superior Eleitoral

UOL 27

Vaccari Neto, João 54, 119, 166

Valor 16, 27
vazamentos 45
Veja 4, 16, 119
veto players 10, 43, 130, 195
visibility 75

Waisbord, Silvio 10–11

Watergate 45–47, 78, 168
WikiLeaks 2, 45, 79
White, David Manning 47
Workers’ Party 2, 5, 32–33, 54–57

Yousseff, Alberto 6–8, 54, 100, 107, 118, 129

Zavascki, Teori 63, 96–97

Zelotes operation 104, 115, 122–124, 174