Silencing Critics: Why and How Presidents Restrict Media Freedom in Democracies
Marisa Kellam
Associate Professor
Waseda Institute for Advanced Study
Waseda University
Tokyo, Japan
[email protected]
Elizabeth A. Stein
Assistant Professor
The Institute of Social and Political Studies
The University of the State of Rio de Janeiro
Rio de Janeiro, Brazil
[email protected]
Working Paper
Last updated October 2014
Prepared for Presentation at the
Fundação Getúlio Vargas, Escola Brasileira de Administração Pública e de Empresas,
(FGV/EBAPE)
October 29, 2014
Abstract
The media hold elected leaders accountable by exposing corruption and policy failures. Although
many politicians accept media criticism as intrinsic to liberal democracy, some politicians rein in the
freedom of the press or intimidate media outlets in order to silence their critics. We identify
circumstances that motivate and enable presidents to curb media freedom in presidential
democracies. We argue that (1) presidents who hold ideological positions contrary to those of the
mainstream media adopt the media as viable opponents in the absence of an effective partisan
opposition; and (2) the media are vulnerable to presidential infringements on their freedom where
legislatures and judiciaries hold weak powers relative to presidents and are therefore unable to
constrain presidents' actions against media freedom. We support our argument with statistical
analyses of press freedom ratings in Latin American democracies from 1993-2013.
1. Introduction
In 2011, Ecuadorian President Rafael Correa sued Ecuador’s largest paper, El Universo, for criminal
libel in response to an editorial column that criticized Correa’s handling of a police uprising that
turned deadly, and referred to him as “the dictator.”1 The owners of the paper and the opinion
editor were sentenced to three-year jail terms and assessed fines of $40 million – sufficient to
bankrupt the newspaper. Ecuador’s highest court upheld the criminal libel convictions, eliciting a
barrage of international condemnation from press freedom organizations:
Since becoming president, [Correa] has used all the powers at his disposal – state-controlled
media, rubber-stamp lawmakers, the courts and the presidency itself – to intimidate the news
media and put an end to criticism of his government. Ironically, in view of his autocratic
record, the president feels he’s been libeled by a newspaper column that called him a
dictator.2
Although President Correa ultimately pardoned the convicted men and granted them, in his own
words, “remission of the sentences they rightly received,” he used the opportunity to declare victory
over Ecuador’s “media dictatorship.”
3
During Correa’s presidency, the government also has
confiscated the computers of a critical magazine, alleging bad labor practices; declared a state of
emergency ordering all television stations to transmit official state programming; ordered the
preventative detention of a reporter on terrorism charges; and forced television news programs to
defer some of their time to vitriolic rebuttals by government officials (HRW 2011, Lauría 2011b).
Furthermore, while silencing his critics in the private media, President Correa has simultaneously
constructed a network of state-operated media, including more than 15 television, radio and print
outlets (Lauría 2011b).
The decline in media freedom in Ecuador, though more extreme than many Latin American
countries, represents part of a regional trend. Freedom House, a self-described independent
watchdog organization, monitors economic, legal and political impediments to press freedom
worldwide and rates countries annually on a 100-point composite Press Freedom Index (PFI). 4
1
2
3
4
The column is available online: “NO a las mentiras.” El Universo. February 6, 2011,
http://www.eluniverso.com/2011/02/06/1/1363/mentiras.html.
“The press under siege: Dangerous precedent for press freedom in Latin America.” InterAmerican Security Watch. July
19, 2012 (reprint of material from The Miami Herald): http://interamerican securitywatch.com/the-press-undersiege-dangerous-precedent-for-press-freedom-in-latin-america/.
Correa’s comments were reported in English translation in “Ecuador President Correa pardons paper in libel case.”
BBC News. February 27, 2012: www.bbc.co.uk/news/world-latin-america-17177646, and Neuman, William. 2012.
“President of Ecuador to Pardon Four in Libel Case.” New York Times, February 27, 2012:
www.nytimes.com/2012/02/28/world/americas/president-of-ecuador-to-pardon-four-in-libel-case.html?_r=0.
We do not refer to “the press” in the strict sense, meaning only print news outlets, but rather use it interchangeably
with “the media.”
1
According to the PFI, press freedom deteriorated in most Latin American countries between 2001
and 2013, with 12 countries experiencing a net drop of 10 or more points. During this same period
of time, democratic institutions in Latin American countries, as measured by the Polity score, have
remained strong, with few exceptions. The Polity score evaluates countries annually based on the
institutionalization of authority with regard to executive recruitment, constraints on executive
actions, and political competition; higher scores indicate stronger democratic institutions. Only three
Latin American countries experienced declines in their polity score from 2001 to 2013. Figure 1
illustrates average trends in these two indicators across the region.
Figure 1. Average Democracy and Media Freedom Ratings in Latin America, 1993-2013
Despite a vastly improved institutional environment in Latin America, which might lead us to
anticipate improved protections for the media, average press freedom ratings for most countries in
the region have dropped precipitously over the last decade. The stark decline of the region’s average
media freedom masks differences across countries. Nevertheless, the diverging trends in average
media freedom and institutionalized democracy were not driven by a few cases dominating average
values, but rather by similar trends occurring across the majority of countries.
2
This trend of eroding press freedom despite consolidated democratic institutions raises
theoretical and empirical puzzles for the study of democratic accountability. Recent developments in
some Latin American countries also raise normative concerns; sustained degradation of freedom of
the press has the potential to take its toll on the quality of governance and lead to the deterioration
of democratic institutions. Amartya Sen (1999) highlights the role of a free press in democracy in his
famous study that showed that where democracy exists, famines do not, which he attributed to the
ability of free media to draw attention to citizens’ needs. Absent media freedom Sen’s maxim might
not hold up over the long run. A subdued press cannot carry out its ideal democratic functions of
informing the citizenry, which is a critical component of vertical accountability (Waisbord 2002).
Research demonstrates that freedom of the press is associated with less corruption and better
governance (Norris 2010, Camaj 2013, Stein and Kellam 2014). While many scholars see media
freedom as essential to democracy (e.g. Dahl 1998, Norris 2000), other scholars contend that factors
such as political culture and institutions may supersede the media’s importance (e.g. Gunther and
Mughan 2000, Graber 2003).
While scholars disagree about the media’s importance for strengthening democracy, existing
research consistently demonstrates that democracy is insufficient to create free media. Scholars who
study media in transitional democracies around the globe suggest a range of impediments to media
liberalization, from limited ideological diversity, independence and professionalization to outright
government repression (Lawson and Hughes 2005, Voltmer 2013, Young and von Doepp 2013).
Many other well-known issues threaten journalists’ ability to report freely in democratic countries.
Especially where rule of law remains tenuous, impunity for perpetrators of physical attacks against
journalists – committed by drug cartels, organized crime syndicates and paramilitary forces –
perpetuates these actions aimed at intimidating the press (Collings 2001, Wasibord 2000, Hughes
and Lawson 2005). We believe that these types of anti-press violence – despite their prevalence in
Latin America – cannot account fully for recent deterioration in media freedom in the region, since
journalists who have covered these dangerous beats have suffered anti-press violence for decades
before the recent declines in press freedom ratings.
Declining press freedom in Latin American countries reflects the influence of politicians who
have turned the mainstream media into their principal political rival and, as such, have taken actions
against the media to stifle news coverage (Boas 2012). Yet, most existing research overlooks
restrictions to press freedom commonly initiated by presidents who “only sporadically resort to
coercive means,” instead using “legislation to chill critical coverage” (Waisbord 2002, 91). While
3
political scientists have shown that autocrats sometimes have incentives to permit partially or almost
free media (Egorov, Guriev and Sonin 2009, Whitten-Woodring 2009), we know less about
democratic leaders’ incentives and abilities to constrain media freedom.
To explain the presidential component of declining media freedom under democracy, we
identify motivating, and constraining or enabling conditions that underlie presidents’ intrusions on
the media. Presidential harassment of the media is, in part, episodic; presidents are more likely to
exert greater control over information in the context of street protests, violent conflict or political
crises that pose serious threats to government stability (Young and von Deopp 2013). Presidential
media harassment is also a function of the institutional context; we argue that in the absence of an
effective partisan opposition the media become viable opponents of the president, and are
particularly vulnerable targets of presidential actions where the legislature and judiciary hold weak
powers relative to the president.
We propose that where and when presidents come to power in landslide victories – often due to
the breakdown of traditional party systems – they must create an opposition in order to rally and
sustain their base. Deinstitutionalization of political completion follows party-system collapse,
resulting in more contentious forms of politics (Flores-Macías 2010, Morgan 2011). In countries that
have experienced the breakdown of the party system and that also lack a diverse ideological array of
media, presidents who hold contrary ideological preferences take on the media establishment as their
political opposition, accusing them – often for good reason – of ideological biases and conscious
efforts to distort the truth when reporting on the government. Since the media have an incentive to
cover controversy (Baum and Groeling 2009), they often take the bait and engage presidents in
rhetorical battles, providing fodder for presidents to take more severe actions, beyond mere antimedia rhetoric, to silence their critics.
In this paper, we examine political and press freedom trends in Latin American democracies
from 1993 through 2013, in order to assess the validity of our theory. Latin American countries
during the last 20 years provide a good sample of countries for examining this phenomenon. The 18
democratic countries have exclusively presidential systems, though other institutions vary. Previous
research has described media ownership in Latin America as “oligarchic,” mostly privately owned by
a small number of wealthy, primarily conservative families (Djankov et al 2001, Hughes and Lawson
2005, Boas 2012, Waisbord 2010). The problematic concentration of media ownership and lack of
ideological diversity remain the worst for television, the medium that reaches the greatest number of
people in Latin America (Hughes and Lawson 2005). At the same time, the left has resurged in
4
several countries throughout Latin America (Castaneda 2006, Weyland, Madrid, and Hunter 2010;
Levitsky and Roberts 2011), creating potential for increased ideological antagonism between the
president and the media (Boas 2012). The degree of press freedom and its degradation varies greatly
not only across countries, but also within some countries over time.
We proceed with a brief review of the literature on state-media relations in Latin America and a
description of trends in press freedom since democratization. We then develop our theory regarding
the conditions under which presidents take actions that limit press freedom. We elaborate on the
conditions that constrain or enable presidents from doing so. In testing our argument empirically,
we employ dynamic time-series, cross-sectional regression analyses to assess the political
circumstances and institutional arrangements under democracy that are associated with lower levels
of media freedom and with larger changes in media freedom, both annually and over the course of
presidents’ terms. Following the discussion of our empirical results, we conclude with a brief
commentary on the implications of our findings for the sustainability of liberal democracy
throughout the region.
2. State-Media Relations in Latin America
Research on state-media relations after Latin American countries transitioned to democracy points
to varied outcomes with regard to freedom of the press in Third-Wave democracies. Under
authoritarian rule in Latin America, much of the media had been complicit with the regimes
(Skidmore 1988; Smith 1997, Samples 2010); editorials in many newspapers actively called on the
military to intervene to remove leftist presidents in Chile and Brazil. Yet, in these same places, a
vibrant opposition press thrived (Kucinski 1991, Stein 2008). With the rebirth of democracy in the
region, some major media outlets used (or were forced by) the democratic transitions to redefine
themselves as democratic entities (Porto 2007, Matos 2008). New economic conditions of media
organizations, which had previously depended on state subsidies and advertising revenue, drove
them to become watchdogs in order to compete for readership with the alternative press that carried
out that role during authoritarian rule (Pinto 2008). Enjoying unprecedented freedom after
democratic transitions, the Latin American press uncovered several corruption scandals that led to
resignations and impeachment hearings against high-level government officials including against
several presidents (Waisbord 2004, 2006; Pérez Liñán 2007, Balán 2011).
Yet democratization has not prevented unwarranted lawsuits and criminal prosecution against
journalists who cover corruption and political scandals in Latin American countries. Several
5
countries in Latin America maintain desacato laws, which criminalize the publication of information
considered embarrassing to the president or government (Hughes and Lawson 2005). According to
international law, desacato laws “grant an unjustified, special protection to public officials and
therefore limit free expression” (HRW Ecuador Report 2011). Even where laws governing media
content mandate just legal protections for journalists, such laws only safeguard journalists if judges
remain autonomous from politicians and have the will to prosecute those who attack journalists
(Waisbord 2002).
Although the combination of political liberalization and media commercialization increased
investigative journalism in Latin America, the literature also points to opportunities for governmentmedia collusion after (re-)democratization. Scholars of Latin American media contend that despite
political liberalization concentrated media ownership in these countries has impeded the
democratization of the media 5 (Hallin and Papathanassopoulos 2002, Hughes and Lawson 2005,
Boas 2012). Relative to other world regions – both more and less-developed – a higher percentage
of media outlets in Latin America, and of the press in particular, is privately held (Djankov et al.
2001). In several countries where “a handful of conglomerates” have controlled multiple media
outlets, “politicians only had to collude with a few major players to ensure overwhelmingly favorable
coverage” (Boas 2012, 3). Though some improvement in ownership concentration has occurred in
Latin America’s largest markets, Brazil and Mexico’s top television networks still dominate their
closest competitors by more than 30 percentage points on the Herfindahl Hirschman Index, a
measure of media concentration (Boas 2012, 4-5). Due to weak laws regulating monopolies and
oligopolies, along with historical links between politicians and media owners, most Latin American
countries have not developed a tradition of vibrant, depoliticized public broadcasting (Fox and
Waisbord 2002; Matos 2011).
Whether it stems from corporate ownership or government assumption of media outlets, media
concentration generally reflects limited interpretations of government policies and politicians and
provides unfair advantages to whoever controls the media. Studies have shown that publicly run
broadcasting (at the state and federal level) produces news biased toward the incumbent or his party
(Hughes and Lawson 2004); that ownership of regional media outlets by local elites strengthens their
political influence to such a degree that some willingly operate at a loss to maintain their influence
(Amaral and Guimarães 1994, Hughes and Lawson 2005; Cañizález and Lugo-Ocando 2008); and
5
Based on the literature, we interpret “media democratization” to imply increasing pluralism of media ownership and
content and/or expanding access to a wider audience.
6
that incumbent candidates for municipal elections who operate community radio stations
substantially benefit from them at election time (Boas and Hidalgo 2011). The ideal of stateownership or regulation of broadcast media as a mechanism to minimize political biases and serve
the public interest underlies recent calls for media reform by civil society. However, with regard to
Latin American public broadcasting “it is clear that the state is still very politicized, more than three
decades after the fall of the dictatorships in the region” (Matos 2011, 185).
In addition to media concentration in Latin America, the tenuous state of rule of law in many
countries has driven the deterioration in press freedom under democracy. Journalists assigned to
beats covering crime (drugs in particular), corruption, politics and human rights have the highest risk
of being victims of violence stemming from their jobs (Lauría 2013). Journalists’ fear of repression
perpetrated by drug cartels, organized crime syndicates and paramilitary forces limits their ability
and/or willingness to report freely (Collings 2001). Additionally, ineffective judicial systems – with
judges who themselves may be threatened by drug-related violence or bought off by the cartels –
provide impunity to perpetrators of crimes against the press, leaving journalists vulnerable and less
likely to pursue such stories in the future (Waisbord 2002, Boas 2012).
At the same time, local public officials use criminal and civil defamation lawsuits to protect their
reputations, and occasionally turn to physical intimidation to thwart critical reporting or in
retribution for past publications, strongly deterring watchdog journalism in many areas of Latin
America (Boas 2012). In Brazil, judges issue injunctions against the press preventing them from
publishing material; CPJ cites a study by the University of Texas’ Knight Center for Journalism
saying that 21 such injunctions occurred in the weeks before Brazil’s 2010 elections. One of the
most famous cases “judicial censorship” against the press remained in place more than four years
after an injunction was first issued against one of the nation’s top newspapers, O Estado de São Paulo,
in which a judge in the Brazilian capital prohibited the paper from reporting anything further on a
story the paper broke about corruption and nepotism on the part of former Brazilian President and
then-President of the Senate José Sarney (Lauría 2011a, Freedom House 2010, Otis 2014).
While previous research highlights important impediments to press freedom in democratic Latin
America, such as concentrated ownership and weak rule of law, the comparative political science
literature has not sufficiently addressed systemic declines in press freedom that can be attributed, in
large part, to the actions of the presidents themselves. Yet, international newspapers and press
freedom monitoring agencies regularly report on the actions taken by Latin American presidents to
7
intimidate journalists and stifle or mold news coverage. We highlight some of these presidential
actions below.
2.1 Presidential Actions against the Press in Democratic Latin America
Presidents and their communications teams can incentivize and reward positive coverage or can
impede and penalize negative coverage using a repertoire of legal, political and economic tools at
their disposal. Latin American presidents in contemporary democracies have resorted to a broad
range of these mechanisms to rein in contentious and purportedly biased media.
Like President Correa of Ecuador, as discussed above, former Venezuelan President Hugo
Chávez actively chiseled away at media freedom using both legal and economic measures and
regularly engaged in verbal harassment of the press. In Peru, under President Alberto Fujimori,
presidential advisors, hoping to stifle press criticism, resorted to systemic bribery to control the
media. Government strategies to tamp down critical coverage also have contributed to a decline in
investigative journalism in Argentina (Pinto 2008). Presidents in these and other countries use
government advertising, bank loans and tax credits to incentivize positive reporting or, conversely,
employ sanctions by withdrawing such benefits, applying political pressure, raiding publication
offices, and confiscating publications determined to be unfavorable to the president ex post and by
utilizing other forms of “soft censorship” to limit press scrutiny of their office.
Recently a few Latin American presidents have signed new laws that potentially allow them to
regulate media content. These presidents argue that the laws they endorse are aimed at
“democratizing the media” (Brockwehl 2013). Yet, Latin American governments have demonstrated
limited success in establishing media outlets oriented toward public service and education. Carolina
Matos (2011, 187) states that “(s)tate intervention in South America has had the main aim of
reinforcing governmental powers rather than promoting democratic forms of communication.”
Even if the primary intent of such legislation is to increase media pluralism, such as Argentina’s 2009
Law on Audiovisual Communication Services and Ecuador’s 2013 Communication Law, or inhibit
discrimination against sub-populations, as in the case of Bolivia’s 2010 Law Against Racism and all
Forms of Discrimination, these laws increase government oversight of media (Brockwehl 2013).
Several Latin American presidents have expanded state-run broadcasting. For example,
ownership of media, once reserved for the Venezuelan opposition as evidenced during the “mediaorchestrated” coup attempt in 2002, has been “reversed in the incumbent’s favor by a wave of cash.
The government invested more than $40 million in upgrading the state-owned television station and
8
the government news agency, established three more television stations, acquired more than 145
local radio stations and 75 community newspapers, and created dozens of administration-friendly
websites” (Corrales and Penford 2007, 110). Likewise, President Rafael Correa expanded Ecuador’s
state-owned media empire; Correa took advantage of the 2008 financial crisis that sent many media
organizations to the brink of bankruptcy, taking “public” ownership of such outlets after the
government purchased these companies’ debt (Freedom House 2011). Though the government of
Argentine President Cristina Fernández de Kirchner did not assume direct control of media outlets,
as in the cases above, she pushed through legislation that threatened to take over control of the
country’s sole newsprint manufacturer, Papel Prensa, from Argentina’s two largest newspapers by
declaring its supply a public interest concern. Additionally her government threatened the news
outlets with legal action for purchasing Papel Prensa illegally during the military dictatorship (Rafsky
2012). Given the sordid history of many news outlets during the periods of dictatorship, presidents
pursue these actions under the banner of pursuing justice. Nevertheless, given contemporary
contentious relations between leftist presidents and large media outlets in several Latin American
countries, it can be difficult to discount those who portend that political motivations also underlie
even the most noble of these efforts.
Other Latin American presidents’ attempts to regulate the press have met resistance, such as
former-Brazilian President Luiz Inácio “Lula” da Silva’s effort to create a social council to oversee
the media at the National Conference on Communications in 2009, which his government
sponsored. Scholars, business people and members of civil society debate the benevolence of Lula’s
efforts to reform the media environment (Matos 2011). Fear of the potential for state censorship
ultimately doomed Lula’s attempt to establish the social council for media oversight, a proposal that
prompted many media organizations and news associations’ refusal to participate. Lula reaffirmed
his commitment to the “sacred” principle of media freedom while at the same time using the
opportunity to discredit the media when he proclaimed, “I’ve learnt to live peacefully with all the
slander and abuse against me from certain media.” 6
Though some presidential actions against the media have received significant international media
coverage, the recent deterioration of press freedom is not unique to any one Latin American
country. Nearly all countries have witnessed some backsliding on guarantees and enforcement of
media freedom. Nevertheless, some countries have witnessed far more concerning changes in their
6
The English translation is reported by MercoPress in “Lula da Silva unhappy with the media sponsors ‘social
control’ of content.” December 28, 2009. http://en.mercopress.com/ 2009/12/28/lula2.
9
media environment than have others in the region. These trends motivate our theory for explaining
what occurrences and which governmental attributes account for pronounced declines in some
countries, whereas in other countries presidents either have been unable to impinge on media
freedom or unmotivated to initiate such efforts. Given considerable anecdotal evidence of
presidents trying to silence “watchdog” media in Latin America, we seek to address a gap in the
scholarly literature on threats to media freedom in democratic presidential systems. In the next
section we describe our theory in greater detail, highlighting testable implications of our argument.
3. Why Do Presidents Try to Silence the Press and What Enables Them to Do So?
In this section we propose a president-centered explanation for declining press freedom in
democracies. We focus on the motivating, and constraining or enabling conditions that lead or impede
presidents’ attempts to suppress political criticism or to coerce positive media coverage.
3.1 Motivating Factors
Where and when there is little ideological diversity of the media, presidents with ideologies opposed
by the media establishment likely anticipate that the editorial line coming from ownership will
contradict and possibly impede their policy agendas. The media may thwart the presidential agenda,
in part, through their agenda-setting powers that help shape public opinion (Iyengar and Kinder
1987) and, in part, by pressuring legislators who need access to the media during the campaign
season to oppose the incumbent administration’s policies. In the words of one elected leader from
Latin America, “The press has a decisive influence over Congress […]. If the press moves against a
law, it is very difficult to pass it” (UNDP 2004, 164). Presidents who face opposing media have
greater motivation to constrain the media than do presidents who have the media on their side.
In countries with concentrated, conservative, private ownership of the media – the norm in Latin
America – we expect leftist presidents to be more motivated to limit press freedom than presidents
who fall elsewhere on the left-right ideological continuum. Leftist presidents advocate for greater
equality, oppose traditional hierarchies and concentrated private ownership, and aim to broaden
political participation (Levitsky and Roberts 2011). Where wealthy, conservative families typically
own the large media conglomerates, a natural tension arises between presidents on the left and the
editorial line of media owners (Kitzburger 2010). Leftist presidents and their supporters presume
that:
media and journalistic institutions are, despite their claims of neutrality, powerful social actors
linked to the upper classes, social elites, or powerful corporations…Hostility on the part of the
10
media is therefore interpreted as part of the resistance by the upper classes and established
powerful interests to the reformist and democratizing agendas of progressive governments
(Kitzburger 2010, 8).
The resurgence of leftist presidents in much of Latin America thus has intensified conflict
between these presidents and the media. Starting with the 1998 election of Hugo Chávez as
Venezuela’s president, leftists went on to win the presidencies in more than half of the countries in
the region during the decade that followed. Steven Levitsky and Kenneth Roberts argue that unlike
candidates’ leftist posturing in the 1980s and 1990s that did not extend to policy, “the post 1998
wave of leftist victories ushered in a new era of policy experimentation in which governments
expanded their developmental, redistributive, and social welfare roles. The ‘left turn,’ therefore,
changed not only who governed in Latin America, but also how they governed” (2011, 2, emphasis in
the original).
Philip Kitzberger (2010) claims that leftists in Latin America stand out from other governments
because many of them have gone public in their confrontations with the media as they appeal to and
mobilize their base of support. For example, the Bolivarian Revolution that Hugo Chávez
continually promoted “translated into an open confrontation with the commercial media” (Cañizález
and Lugo-Ocando 2008, 191). According to Kitzberger’s study, Chávez addressed the media
negatively in 45 percent of his public appearances (2010, 8).
Though we focus on the left in the context of Latin America, we believe that were media
ownership in other countries to be concentrated historically among a few private, liberal owners,
conservative politicians likewise would take on the media as their primary opponents in the absence
of a partisan opposition. The motivating nature of ideological conflict between presidents and the
press leads to the following expectation:
H1a:
Media freedom is more likely to decline when and where presidents hold ideological
positions contrary to the dominant ideological leaning of the media establishment
[e.g in Latin American countries we refer to leftist, rather than centrist or
conservative, presidents].
Several populist leftist presidents, including Venezuela’s Hugo Chávez, Bolivia’s Evo Morales,
along with Rafael Correa in Ecuador, rose to power when traditional political parties had imploded
amid corruption scandals, persistent policy failures, and eroding party-citizen linkages (Corrales and
Penfold 2007; Flores-Macías 2010; Levitsky and Roberts 2011; Morgan 2011; Searight 2012).
Scholars have shown that in the wake of party-system collapse, which can weaken democratic
accountability and expand the boundaries of political contention, deinstitutionalization and
11
personalism often emerge (Flores-Macías 2010; Morgan 2011). Having benefitted from the
implosion of traditional political parties or having succeeded in dissolving the political establishment,
these newly elected populist presidents lack a political villain to target with rhetoric they use to
sustain the enthusiasm of their base. Since free and fair competition defines democratic regimes,
politicians’ rhetoric frequently involves accusations and comparisons among opposing parties and
politicians. The void that develops in the absence of an organized, potentially threatening, partisan
opponent leaves presidents with no obvious target for political contention. To fill the void, populistleftist presidents transform other well-established institutions, the media in particular, into their
primary foe.
Absent a unified opposition to challenge the president or keep him in check, we contend that
presidents thrust media organizations into the role of the president’s political opposition, a role the
media establishment often accepts willingly. From owners’ perspective, they need to publish stories
that sell news. Incentives drive journalists to favor conflict and seek usable sound bites that a strong
political opposition usually provides (Baum and Groeling 2009). While journalists might prefer to
quote opposition legislators criticizing the president or even cite a co-partisan of the president
revealing intraparty conflict, absent partisan contention, many journalists in Latin America opt to
directly challenge presidents. As the media become the opposition to the regime, they provide more
fodder for bombastic statements by presidents about media organizations’ anti-government bias.
The confrontational dynamic of state-media relations in this context turns presidents’ rhetoric into
actions against the press and leads to deteriorating conditions under which the press must operate.
This reasoning leads us to the expectation that leftist presidents who come to power under
conditions of limited political competition and face only feeble opposition from the center or right
will have the greatest motivation to confront the press in Latin America. These minimally
competitive environments tend to follow landslide victories in presidential elections. Hence we offer
a qualifying condition to the expectation regarding leftist presidents that we outlined above in
Hypothesis 1.
H1b:
Media freedom is more likely to decline when and where presidents hold ideological
positions contrary to the dominant ideological leaning of the media establishment
and lack an effective partisan opposition [e.g. when and where presidents win
elections in landslide victories].
Regardless of ideology and the extent of political competition they face, all presidents must
worry about their political support both from the public and from their inner circle. Hence all
12
presidents control the flow of information from and about their office to some degree. Presidents
hold controlled press conferences, for example, where they can limit questions and select friendly
reporters. When countries experience domestic crises, broadly defined, presidents become
personally vulnerable as the most visible politician. Particularly in times of crises, presidents often
take measures to “manage” the news. However, some presidents go further and harass and
intimidate journalists and media outlets. They also may sue members of the media so that the media
remain embroiled in legal battles rather than focused on investigative reporting.
Daniel Young and Peter von Deopp argue that government attacks on the media increase during
periods of government insecurity when incumbents face “clear and present threats to their ability to
retain power” (2013, 39). The authors argue that these circumstances motivate governments to
minimize negative accounts that could undermine their support and encourage governments to
control information that might enable opposition coordination. Analyzing government-initiated
attacks on the media in African countries, Young and von Deopp show that media harassment
increases in periods following protests and revelations of coup plots.
We have similar expectations for Latin American presidents who anticipate that popular
challenges have the potential for “snowball” effects, such as waves of anti-government streetprotests, possibly threatening their rule by inviting legislative or judicial investigations and
international pressure. These circumstances likely increase presidents’ motivation to restrict the
publicity of mass dissent (or the government’s response to protests). On the one hand, in cases
where presidents anticipate crises and thus can pre-empt coverage, we expect they will resort to
tactics to suppress reporting of the crisis event; where crises emerge unexpectedly, on the other
hand, presidents may resort to punitive rather than pre-emptive measures to shorten the duration of
exposure and lessen their vulnerability. Thus, we hypothesize the following:
H2:
Media freedom is more likely to decline when and where economic crises occur.
H3:
Media freedom is more likely to decline when and where presidents face persistent
contentious events.
3.2 Enabling and Constraining Factors
While some presidents may have greater motivation than others to curb press freedom, among those
presidents motivated to do so, some will face greater constraints that inhibit their ability to limit
media freedom. Though Latin American countries tend to have strong executives as compared to
the United States, the degree to which presidents can successfully take unilateral action against the
13
press varies depending on the presidents’ constitutional powers and the relative strengths of other
branches. Some presidents’ have direct legislative powers, which may include decree and veto
authority, the ability to declare legislation urgent, or the exclusive right to set the budget and to
initiate legislation in certain areas (Payne 2007). In particular, the constitutionally allocated authority
to legislate by decree enables presidents to directly alter the status quo. The strength of decree
powers varies with the extent of restrictions placed on the policy area or conditions under which
decrees may be issued, and the permanency of the decree (UNDP 2004). When empowered with
discretionary decision-making and unilateral policy-making, some presidents may change media
policy without the possibility of contestation or overrule from other branches.
H4:
Media freedom is more likely to decline when and where presidents have stronger decree
powers.
Some constitutional designs potentially limit presidents’ abilities to persecute the press and
control news coverage through the legislative process. Presidents lacking decree powers, having
limited veto and/or other agenda-setting powers, would need legislative cooperation to succeed in
curbing press freedom (Cox and Morgenstern 2001, Alemán and Tsebelis 2005). The configuration
of partisan veto players at a given time can affect whether other institutions help or impede
presidents who wish to limit press freedom. For example, if presidents’ parties hold a large share of
congressional seats or if the opposition parties are highly fractionalized, then the president will be
better able to take decisive actions against the press. These conditions reduce the president’s
accountability because neither the president’s party nor a fragmented opposition is likely to check
questionable presidential actions (Shugart and Carey 1992, Negretto 2006, Morgenstern et al. 2008,
Samuels and Shugart 2010). Conversely, legislatures that more effectively balance executive power
constrain the policy-making process and limit changes to the status quo, including the state of media
freedom.
Where presidents face strong, cohesive legislative oppositions, presidents will find it difficult to
take action against the press, particularly a press that might be sympathetic to the opposition. Thus,
legislative configurations that strengthen democratic accountability in general – beyond limited press
freedom – should constrain the president from restraining the press. Likewise, the judiciary also
potentially acts as veto players in the policy-making process. Where judiciaries are not subservient to
the president and have the power to review executive and legislative decisions, the judicial branch
should better protect the media against presidential infringements on liberty of the press (PérezLiñán and Castagnola 2009). If empowered to do so, legislators and judges should have incentive to
14
defend a free press that reports on inter-branch relations and thus helps to maintain the balance of
power through monitoring and information-revelation.
H5:
Media freedom is less likely to decline where institutional constraints on executive
power are stronger.
H6:
Media freedom is less likely to decline where judiciaries have greater independence
from the executive branch.
Finally, above and beyond the constraints of political institutions, where presidents benefit from
significant natural resource wealth they may be less accountable to otherwise powerful economic
interests or to their constituents. As rents derived from minerals, natural gas, oil or timber increase,
governments rely less heavily on income taxes, consequently making presidents less accountable to
their electorates and enabling presidents to “buy” political support (Bratton and van de Walle 1994,
Ross 2001). Because resource-rich presidents are less concerned with bureaucratic performance than
their peers who do not benefit from resource rents, they also have less need for the monitoring and
information that free media provide (Egorov, Guriev and Sonin 2009, Stockmann and Gallagher
2011). Additionally, because resource wealth minimizes presidents’ dependence on the support of
economic elite, including those families who frequently own media conglomerates in Latin America,
these presidents have greater leeway with regard to their incursions on the media.
H7:
Media freedom is more likely to decline the greater a country’s resource
wealth.
In summary, we argue that presidents who are in opposition to the dominant ideology of
mainstream media outlets, especially those presidents who lack an effective partisan opposition, have
the strongest motivation to suppress media freedom. In the case of Latin America, we expect leftist
presidents who come to power in landslide victories will be the most motivated. Yet, presidents’
motivation to curtail freedom of the press is not a sufficient indication that such freedom will
inevitably decline. Several contextual factors may hinder or facilitate presidents’ attempts. All else
equal, presidents will be less able to obstruct the news media where they face substantial institutional
constraints in the legislative process and/or where the judiciary can exercise its independence and has
broad powers of judicial review. Nevertheless, strong legislative majorities, decree powers, or access
to substantial rents derived from natural resources may insulate presidents from mechanisms
designed to check presidential power, thereby enabling presidents’ incursions on press freedom, given
that other factors remain the same. In the next section we present our empirical analysis in which we
15
assess whether the factors that we posit motivate, abet or discourage presidents from silencing their
critics affect media outcomes as hypothesized.
4. Methodology
We test our hypotheses using annual data on media freedom, political institutions, and other political
and economic factors in Latin American countries from 1993 or beginning with democratization
through 2013. While the above hypotheses specify conditions when media freedom is more, or less,
likely to decline, we observe both static and dynamic implications of these hypotheses in our data. If
presidents actively curb media freedom, then we will observe lower levels of media freedom, on
average, when and where they do. Hence, we construct empirical models to account for variation in
the level of media freedom across countries and time, and also to assess whether the estimated effects
produce contemporaneous and/or long-term changes in media freedom over time.
We conduct two sets of analyses, one using annual observations and the other using presidential
terms as the unit of analysis. For models with annual data, we include countries receiving a polity
score of 5 or higher in a given year from 1993 through 2012, giving us between 11 and 20 yearly
observations in the 18 countries. When analyzing terms, our model includes 80 presidential terms
that reached at least 2.5 years in countries that had polity scores of at least 5 throughout the entire
term. We start terms in the year in which a president comes to power so long as she assumed office
prior to July 1; we use the subsequent year as the first year in office when a president assumes power
July 1 or later. Likewise, we end a president’s term in the year she leaves office if it occurs on or after
July 1; where a president’s term ends prior to July 1 we count the prior year as her final year in
office. In the term-level analysis, we measure the level of media freedom in the final year of the
term, or 2013. Except where noted otherwise below, we calculate the average value of the
independent variables for the years of the presidential term up to but not including the final year in
office.
We conduct time-series, cross-sectional analyses of media freedom scores, recognizing that such
analyses pose well-known challenges for statistical estimation. To address differences across
countries in levels of media freedom unaccounted for by the explanatory and control variables, we
include random effects for countries in all of the statistical models we estimate (Clark and Linzer
2012). Also, according to a test appropriate for panel data we can strongly reject the null hypothesis
of no serial correlation in our model specification when applied to the annual data (Wooldridge
16
2002, Drukker 2003). Therefore we estimate the annual model using a GLS method for random
effects in panel data with autocorrelated [AR(1)] disturbances (Baltagi and Wu 1999).
In addition to explaining variation in the level of media freedom across countries and time, we
also seek to account for variation in the amount of change in media freedom from one year to the
next and over the course of presidential terms. Thus to model the time series processes using the
annual data, we estimate an autoregressive distributed lag (t-1) model (ADL), which includes the
lagged value of the dependent variable and lagged values of some independent variables. Our theory
implies that change in some of the explanatory variables should produce change in media freedom
and this model allows us to examine the extent to which such change occurs immediately or in the
future, and to predict the total amount of change that will play out over time (De Boef and Keele
2008). To assess change across presidential terms, we also run a term-level analysis that controls for
the “inherited” media freedom score, which is essentially a lagged dependent variable measuring the
media freedom score in the final year of the previous presidential term. Once we include the
inherited media freedom score, the explanatory variables then account for variation in the net
change that occurs during presidential terms.
4.1 Dependent Variable
Our dependent variable is the Press Freedom Index (PFI) from Freedom House. 7 This is a
composite evaluation of obstructions to press freedom in three areas: legal, political (including
repression), and economic. The legal component evaluates the extent to which laws and regulations
– and the government’s inclination to use these laws and legal institutions – effectively restrict media
freedom. The political component describes the government’s degree of political influence and
control over news content as well as journalists’ ability to cover the news free from harassment,
intimidation, arbitrary detention, violent attacks, etc. The economic component refers to various
aspects of media ownership, state advertising and subsidies, the impact of corruption and bribery on
reporting, and the extent to which the country’s economic situation influences the sustainability of
the media. To create the annual ratings, Freedom House combines assessments from their in-house
analysts with evaluations carried out by other press freedom monitoring agencies that belong to the
International Freedom of Expression Exchange network.
Freedom House (2014) makes two key points emphasizing comparability across space and time,
in the description of their methodology:
7
The codebook in the appendix provides sources, definitions and measurement for all variables used in the paper.
17
The ratings are compared with the previous year’s findings, and any major proposed
numerical shifts or category changes are subjected to more intensive scrutiny. These reviews
are followed by cross-regional assessments in which efforts are made to ensure comparability
and consistency in the findings.
And,
Recent modifications are intended to capture changes in the news and information
environment without altering the comparability of data for a given country over the 34-year
span or the comparative ratings of all countries over that period.8
Though we believe that Freedom House provides a generally reliable and comprehensive
measure of press freedom across countries, we are somewhat more apprehensive of the index’s
comparability over time, due not only to changes in the weighting of the component parts, but also
because of changing societal norms regarding which factors most limit journalists’ freedom. To
address the reweighting for time series purposes, we adjust the weights for the years 1993-1995 to
match the weights used since 1996 and recalculate the scores. To account for possible shifts in
journalistic norms that could influence coding over time, we include the average Freedom House
score for all democratic regimes (i.e. those regimes with Polity scores of 5 or greater) worldwide –
excluding Latin America – in our statistical models as a control for any contagion effects from global
events or systematic shifts in coding norms.
As described earlier, the PFI theoretically ranges from 0 to 100; we have reversed the scale in
our statistical models so that higher numbers indicate greater media freedom. Although the index as
constructed by Freedom House is bounded from 0 to 100, no cases included in our analysis
approach the index’s upper or lower limits. All observations included in our analysis average 59 on
the reweighted and reversed PFI, with a standard deviation of approximately 13.
4.2 Independent Variables
We include several variables that capture conditions that we hypothesized would increase presidents’
motivation to restrict press freedom. First, we use a dummy variable for Leftist President.
Accepting Levitsky and Roberts’ (2011) definition of the Left, we identify presidents whose rhetoric,
support bases, and policy platforms prioritize socio-economic and political equality. 9 Landslide
Win, also a dummy variable, indicates that the president won with a margin of victory 20 percent or
8
9
http://freedomhouse.org/sites/default/files/FOTP2014_Methodology.pdf
Leftist presidents are indicated in Table A3 of the appendix. Our coding is largely consistent with those of Levitsky
and Roberts (2011), Murillo, Oliveros and Vaishnav (2010), Baker & Greene (2011), Wiesehomeier and Benoit
(2009) and Coppedge (1997).
18
greater of total votes over his or her nearest competitor in the presidential election.10 To analyze the
effect of Leftist President, conditional on the level of political competitiveness that they face, we
also include the interaction term of Leftist President x Landslide Win.
As additional motivating factors, we include variables that indicate policy failures or threats to
presidential legitimacy or power. We include the total number of General Strikes and of Antigovernment Demonstrations recorded for each country in each year. In our term model, we
include the count of the total number of strikes or protests that occurred during each presidential
term. We use a dummy variable to note years, or terms, in which a Banking Crisis began.
A second set of variables captures basic political and economic conditions that likely enable
presidents to silence media critics, should they be motivated to do so: (1) the extent of Decree
Powers, which indicates presidents’ abilities to unilaterally enact policy change, and (2) Natural
Resource Rents as a percent of GDP, which accounts for presidents with the capacity to
orchestrate rentier states, purchasing sufficient support to sustain presidents while giving them
greater leeway to trample on the media.
With a third set of variables, we measure the extent of institutional and political checks on
presidential action, which we suggest will limit presidents’ abilities to silence media critics: (1)
effective Constraints on the Executive, which indicates the degree to which other accountability
actors – primarily opposition legislative parties in presidential democracies – constrain presidential
behavior (Marshall, Gurr and Jaggers 2014), and (2) the degree of Judicial Independence, an
indicator of the judiciary’s ability to act authoritatively and autonomously, particularly relative to the
executive (Cingranelli and Richards 2008).
4.3 Control Variables
Finally, to better isolate the influences of the factors described above on media freedom, we include
four control variables. First, we include the United Nations Development Programme’s Human
Development Index (HDI) to account for potential differences in media audiences and news
sources related to demographic characteristics of the populations (e.g. levels of education,
urbanization, and poverty). Second, we control for the number of Journalists Killed as a
consequence of their reporting or while on assignment, as reported by the Committee to Project
Journalists. While journalists’ deaths concern us in principle, our primary research interest concerns
10
When run-off presidential elections take place, we code the election as a “landslide” only if the president’s margin
of victory was 20 percent or more in both the first and second rounds.
19
the aspects of media freedom that (ostensibly) democratic presidents willingly manipulate, which to
our knowledge rarely includes the murder of journalists. The inclusion of this second control
variable helps filter out the variation in the PFI that occurs as a result of anti-media violence
committed by gangs, drug cartels, and other organized crime syndicates in a country. It also captures
the effect on press freedom ratings associated with journalists murdered as a consequence of locallevel politics or corruption that have no direct link to the president. Third, we include a dummy for
Federalism, indicating the five countries that have effective political decentralization, either
resulting from constitutionally federal systems or from the power of elected governors in practice.
We include this variable because in addition to presidents’ influence on media freedom, subnational
politicians and institutions in these countries likely influence the media freedom environment.
Fourth, as mentioned earlier, we include the Global Average of the Press Freedom Index, among
democracies outside of Latin America, to control for potential changes over time in the criteria used
in Freedom House’s evaluations of media freedom.
5. Results and Discussion
We present the results of our analysis using annual data to predict the level of press freedom across
both countries and time in Table 1, which provide support for most of our hypotheses. The
coefficient estimates in the first column represent a static model, which uses current values of the
independent variables to account for variation in the PFI across country-year observations. In line
with our conditional hypothesis regarding the motivating nature of ideological conflict between the
president and the press outlined in H1b, we find considerably lower media freedom during leftist
presidencies when the president faced less electoral competition. Holding all else constant at their
observed values, the PFI is 5.73 points lower, on average, under leftists who came to office as the
result of a landslide win than leftists who did not win in landslides; the 95% confidence interval for
the marginal effect of landslide for leftist presidents is [-8.75, -2.71].
In terms of motivating events featured in H2 and H3, our results indicate that overall media
freedom is slightly lower in Latin American countries during times of financial crises and popular
mobilization. The start of a banking crisis is associated with a PFI that is about 3 points lower on
average. Five percent of the country-year observations included in this analysis had 2 or more
general strikes; we estimate media freedom to be 1.3 points lower when 2 strikes take place as
compared to cases of no strikes, all else equal.
20
21
Table 1. The Determinants of Annual Press Freedom in Latin America, 1993-2012
21
Our findings confirm that presidential powers and institutional checks and balances are important
determinants of press freedom (H4-H6). Where presidents hold the strongest decree powers media
freedom is, on average, 7 points lower than where presidents do not have decree authority. Where
democratically-elected presidents dominate other branches of government (scoring a 4 on the
executive constraints measure), indicating political systems with limited horizontal accountability
such as Ecuador under Correa and Venezuela under Chávez (2006-2008), the model predicts a PFI
that is 7.8 points lower than under the most constraining institutions found in Latin American
democracies, such as contemporary Chile, Costa Rica, Nicaragua, Paraguay, Peru or Uruguay (all
scoring 7 in 2012), holding all else constant. According to our estimates, the highest level of judicial
independence tends to raise media freedom by 3 points as compared to country-years where the
judiciary is not independent.
Finally, countries with a larger share of their GDP originating from natural resource rents have
lower press freedom scores, all else equal, which is consistent with the enabling characteristic of
natural resource rents stated in H7. For an increase of 10 percent in the share of GDP that comes
Turning to the dynamic models shown in the second and third columns of estimates reported in
Table 1, we examine how the independent variables affect the level of media freedom across
country-years once we control for the PFI score of the previous year. Focusing on the autodistributed lag (ADL) model reported in the third column, we assess both immediate and long-run
effects of the variables that test our motivating hypotheses (H1-H3) with regards to annual change
in PFI scores within countries. In the final column of the table, we provide the long-run multipliers,
calculated based on our ADL model, and their standard errors, which we estimated using the Bewley
transformation (De Beof and Keele 2008).
When analyzing annual change in the PFI, we again find strong support for our motivating
hypotheses. In particular, the dynamic specifications show that a statistically significant decline in
media freedom is associated with the interaction of leftist presidents and landslide victories. We
estimate a coefficient of 0.88 on the lagged dependent variable, which indicates that changes in
media freedom persist to a considerable extent across subsequent years. According to the ADL
model, for instance, media freedom declines by 3.84 points in the year when a leftist comes to office
after a landslide win, and continues to decline during the years that follow. After four years in office,
the model predicts a net decline in media freedom of 5.15 points. In this model, we also find
statistically significant short- and long-term effects of leftist presidents who were not elected in
22
from natural resource wealth, the model predicts media freedom to be about 1.6 points lower.
landslides and of non-leftists who were elected in landslides. Although media freedom increases
immediately by 1.38 when leftist presidents take office, it declines in subsequent years. The results of
the ADL model also suggest that anti-government protests lead to long-term declines in the PFI, but
the immediate effect is minimal.
In the dynamic models, we do not find statistically significant effects for the institutional
variables, probably because the effects of these variables are already captured in the lagged
dependent variable. However, for the variables measuring constraints on the executive and judicial
independence we estimate small positive coefficients that come close to conventional levels of
statistical significance.
In Table 2, we present the results of similar static and dynamic models of media freedom
analyzed by presidential terms, rather than by year. Reinforcing the findings in the annual model
with regard to Hypothesis 1B, these results show that leftist presidents who came to office following
landslide elections are associated with lower levels of media freedom and substantial declines by the
end of their terms. The LDV model indicates that these presidencies typically lead to an 8-point
drop on the PFI, all else remaining constant. To put this drop in context of our earlier discussion,
the PFI for Ecuador declined by 6 points during President Correa’s first term and another 14 points
during his second term. Although the term model generally does not provide support for our other
hypotheses regarding motivating events, we find a borderline statistically significant negative effect
of anti-government protests on change in media freedom over the course of a presidential term.
The term-level analysis supports Hypotheses 5 and 6 related to institutional checks that
constrain presidential incursions on the press. According to the static model, the PFI tends to be
about 12 points higher at the end of terms in which the president was the most constrained than it
typically is at the end of terms where presidents face minimal constraints. Similarly, the model
suggests that media freedom is, on average, 10 points higher at the end of presidential terms that
faced independent judiciaries than it is when the judiciary was not independent. Combined, these
results demonstrate the importance of horizontal accountability for media freedom. The dynamic
model also shows that a one point increase in the constraints on the executive measure is associated
with an increase of 1.8 points in the PFI; conversely, where presidents face fewer constraints, press
freedom tends to decline during their term.
Finally, with regard to the robustness of our findings, we note that even after we exclude the
extreme cases, both in level and in change, our results hold in the annual and term models (see
23
24
Table 2. The Determinants of Press Freedom in Latin America Across Presidential Terms, 1993-2013
appendix). Also, our statistical results suggest that the control variables, apart from the federalism
indicator, are important in the model specification
6. Conclusion
While many factors lead to declines in press freedom, in this article we explicitly focus on
circumstances that encourage and allow presidents to pressure the media and restrain press freedom
in democratic regimes. We offer a comparative institutional explanation for differences in the extent
of freedom enjoyed by the press across Latin American countries and for changes in the media
environments within these countries over time. In doing so, we have also sought to advance
scholarly understanding of accountability in presidential democracies.
We contend that understanding presidential actions meant to limit press freedom under
democratic regimes requires theorizing about both presidents’ political motivations and their
institutional and economic constraints. With regard to motivation, we attribute contention between
leftist presidents and the media in Latin America to the historic concentration of media
conglomerates in the hands of wealthy conservative families. We argue that leftist presidents tend to
engage in heightened rhetoric against the press to sustain their base and are more motivated to take
actions that repress critical news coverage when they come to power in landslide victories, indicating
weak levels of political competition often arising from the collapse of the previous party system.
Under these conditions, presidents see the media as their most viable opposition and will directly
challenge the media with words, legislation, legal and economic sanctions, and other forms of
repression. To understand the extent of infringements, we consider enabling and constraining
factors on presidential action. We contend that the media’s freedom is more vulnerable to
presidential infringements where other political institutions neither counterbalance the power of the
executive nor serve as effective oversight mechanisms to sanction presidential actions.
Our empirical results identify several factors that influence press freedom in Latin American
countries, both its level at specific points in time and its change both across years and throughout
presidential terms. First, we find that the level of press freedom is lower across country-years
governed by leftist presidents who face limited political competition, as we hypothesized. When
these presidents take office, media freedom suffers immediately and continues to decline. By the end
of their terms, media freedom is substantially lower than when they assumed power. We also find
that mass mobilization and banking crises are associated with lower levels of press freedom. Our
findings show that institutional constraints from legislatures and from independent judiciaries, work
25
as intended in promoting accountability by protecting press freedom in Latin America. We also
provide evidence that media freedom is lower in countries where presidents have substantial decree
powers and where natural resources contribute substantially to the economy. In short, we confirm
that powerful presidents in Latin America pose a threat to media freedom.
Powerful presidents from across the region have employed a wide range of tactics to silence
their critics; yet different means lead to similar ends: where presidents face less scrutiny due to an
intimidated or incapacitated press, leaving citizens poorly informed, democracy fails to thrive in the
long term. While we recognize that the media in some Latin American countries have notorious
histories of collaboration with authoritarian leaders, we do not see democratically elected presidents
resorting to the same tactics of censorship and repression as a positive counterweight to the media’s
tainted legacy. We agree with Taylor Boas (2012, 8) that “a climate of conflict” between presidents
and the media diminishes democratic accountability. Highly politicized watchdog journalism may
lack credibility, but we find self-censorship more concerning. When intimidation, pressure or legal
sanctions work as intended, the media avoid investigative journalism and refrain from critical
reporting.
Increasing government control of media outlets in some Latin American countries is also
troubling. Simeon Djankov and his co-authors “find pervasive evidence of ‘bad’ outcomes
associated with state ownership of the media (especially the press)” (2001, 4). They show that high
levels of government ownership of the printed press negatively affect citizens’ rights, levels of
corruption and the efficacy of government. They also show that in countries dominated by stateowned media, indicators of health and educational wellbeing are inferior to those in countries with
privately owned media, even if concentrated.
Only when the media are free and independent can they serve as the “fourth estate,” regardless
of whether or not they actually fill this role. In this framework, the media play an essential part in the
checking and balancing that foster democratic accountability, such that the strength of the legislative
opposition increases the freer the press; the media’s influence weakens when the judiciary lacks
independence (McMillan and Zoido 2004). When John McMillan and Pablo Zoido (2004) compare
the value of bribes paid to congress members, judges, and news outlets during the presidency of
Alberto Fujimori in Peru, they find that Vladimir Montesinos, the head of Peru’s National
Intelligence Service, paid television, followed by newspapers, the highest bribes. The authors
attribute the difference in cost primarily to the media's key informative role: “while the checks and
balances are interdependent, the media are the crucial check” (21).
26
Based on our analysis, and supported by many press freedom organizations, it appears that
promoting diversity in ownership across mediums will best serve Latin Americans by strengthening
accountability, a key to fostering thriving democracies in the region. Diversification in ownership
should lead to higher quality journalism as various publishers and broadcasters compete with one
another for a limited audience and fill niche markets (Baker 2007). Taken as a whole, media systems
with diversified ownership are less biased. Diversification also should reinforce the media’s ability to
hold governments accountable because where more owners exist, the government is less likely to be
able to “buy” all media outlets and the likelihood that all media would willingly be bought declines
(Besley and Prat 2006, Baker 2007).
At the same time, our analysis also reveals the need for strong mechanisms of horizontal
accountability that can protect media freedom by preventing presidents from silencing media outlets
or seeking retribution against journalists who report unfavorably about them. Furthermore, our
research points to the importance of vibrant, balanced political competition among political parties.
Strong presidents and weak party systems do not bode well for media freedom nor democratic
accountability in Latin America; with no meaningful political-institutional opposition, the media may
become the last opposition standing – that is, until presidents silence them.
27
Bibliography
Alemán, Eduardo and George Tsebelis. 2005. “The Origins of Presidential Agenda-setting Power in Latin
America.” Latin American Research Review 40(2): 3-26.
Amaral, Roberto and Cesar Guimarães. 1994. “Media Monopoly in Brazil.” Journal of Communication 44(4): 2638.
Balán, Manuel. 2011. “Competition by Denunciation: The Political Dynamics of Corruption Scandals in
Argentina and Chile.” Comparative Politics 43(4): 459-478.
Baltagi, Badi H. and Ping X. Wu 1999. “Unequally Spaced Panel Data Regressions with AR(1) Disturbances.”
Econometric Theory 15: 814–823.
Baum, Matthew A. and Tim Groeling. 2009. War Stories: the Causes and Consequences of Public Views of War.
Princeton, N.J.: Princeton University Press.
Andy Baker and Kenneth F. Greene (2011). “The Latin American Left's Mandate: Free-Market Policies and
Issue Voting in New Democracies.” World Politics 63(1): 43-77.
Baker, C. Edwin. 2007. Media Concentration and Democracy: Why Ownership Matters. New York: Cambridge
University Press.
Besley, Timothy and Andrea Prat. 2006. “Handcuffs for the Grabbing Hand? Media Capture and
Government Accountability.” The American Economic Review 96(3): 720-736.
Boas, Taylor C. 2012. “Mass Media and Politics in Latin America.” In Constructing Democratic Governance in Latin
America, 4th edition, eds. Jorge I. Dominguez and Michael Shifter. Baltimore, Md.: The Johns Hopkins
University Press. Draft version, downloaded from http://people.bu.edu/tboas/media_LA.pdf
Boas, Taylor C. and F. Daniel Hidalgo. 2011. “Controlling the Airwaves: Incumbency Advantage and
Community Radio in Brazil.” American Journal of Political Science 55(4): 868-884.
Bratton, Michael and Nicholas van de Walle. 1994. “Neopatrimonial Regimes and Political Transitions in
Africa.” World Politics 46: 453-489.
Brockwehl, Alexander. 2013. “Ecuador’s New Law Dispels the Myth of `Media Democratization’ in Latin
America.” Freedom at Issue Blog: June 13, 2013. http://www.freedomhouse.org/blog
Camaj, Lindita. 2013. “The media's role in fighting corruption: Media effects on governmental
accountability.” The International Journal of Press/Politics 18 (1): 21-42.
Cañizález, Andrés and Jairo Lugo-Ocando. 2008. “The Media in Venezuela: The Revolution was Televised,
but No One was Really Watching.” In The Media in Latin America, ed. Jairo Lugo-Ocando. New York:
Open University Press, pp. 193-210.
Castaneda, Jorge G. 2006. “Latin America’s Left Turn: A Tale of Two Lefts.” Foreign Affairs May/June,
http://www.foreignaffairs.com/articles/61702/jorge-g-castaneda/latin-americas-left-turn, accessed May
20, 2014.
Cingranelli, David L. and David L. Richards. 2010. “The Cingranelli-Richards (CIRI) Human Rights Dataset”
http://www.humanrightsdata.org
Clark, Tom S. and Drew A. Linzer. 2012. “Should I Use Fixed or Random Effects?” The Society for Political
Methodology Working Paper 1315.
Collings, Anthony. 2001. Words of Fire: Independent Journalists who Challenge Dictators, Drug Lords, and Other
Enemies of a Free Press. New York: New York University Press.
Coppedge, Michael. 1997. A Classification of Latin American Political Parties. Working Paper 244 Kellogg
Institute, Univeristy of Notre Dame.
Corrales, Javier and Michael Penfold. 2007. “Venezuela: Crowding out the Opposition.” Journal of Democracy
18(2): 99-113.
28
Cox, Gary and Scott Morgenstern. 2001. “Latin America’s Reactive Assemblies and Proactive Presidents.”
Comparative Politics 33(2): 171-189.
Dahl, Robert A. 1988. On Democracy. New Haven: Yale University Press.
De Boef, Suzanna and Luke Keele. 2008. “Taking Time Seriously.” American Journal of Political Science 52(1):
184-200.
Djankov, Simeon, Carlee McLiesh, Tatiana Nenova and Andrei Shleifer. 2001. “Who Owns the Media?”
Working Paper 8288, National bureau of Economic Research, http://www.nber. org/papers/w8288
Drukker, David M. 2003. “Testing for serial correlation in linear panel-data models.” The Stata Journal 3(2):
168-177.
Egorov, Georgy, Sergei Guriev and Konstantin Sonin. 2009. “Why Resource-poor Dictators Allow Freer
Media: A Theory and Evidence from Panel Data.” American Political Science Review 103 (4): 645-668.
Flores-Macías, Gustavo A. 2010. “Statist vs. Pro-Market: Explaining Leftist Governments’ Economic Policies
in Latin America.” Comparative Politics 42 (4): 413-433.
Fox, Elizabeth and Silvio Waisbord, (eds.). 2002. Latin Politics, Global Media. Austin, Tex.: University of Texas
Press.
Freedom House 2011. “Freedom of the Press 2011: Ecuador.” www.freedomhouse.org/report/freedompress/2011/ecuador.
Graber, Doris. 2003. “The Media and Democracy: Beyond Myths and Stereotypes.” Annual Review of Political
Science 6: 139-60.
Gunther, Richard and Anthony Mughan. 2000. “The Media in Democratic and Nondemocratic Regimes: A
Mutlilevel Perspective.” In Democracy and the Media: a Comparative Perspective (Eds.) Richard Gunther and
Anthony Mughan. New York: Cambridge University Press, pp. 1-27.
Hallin, Daniel and Stylianos Papathanassopoulos. 2002. “Political Clientelism and the Media: Southern
Europe and Latin America in Comparative Perspective.” Media, Culture & Society 24: 175-195.
Hughes, Sallie and Chappell Lawson. 2004. “Propaganda and Crony Capitalism: Partisan Bias in Mexican
Television News” Latin American Research Review 39(3): 81-105.
_____. 2005. “The Barriers to Media Opening in Latin America.” Political Communication 22(1): 9-25.
Human Rights Watch (HRW). 2011. “World Report 2011: Ecuador.” http://www.hrw.org/world-report2011/ecuador, accessed August 17, 2012.
Iyengar, Shanto and Donald R. Kinder. 1987. News That Matters: Television and American Opinion. Chicago: The
University of Chicago Press.
Kitzberger, Philip. 2010. “The Media Activism of Latin America’s Leftist Governments: Does Ideology
Matter?” GIGA Working Paper No. 151.
Kucinski, Bernardo. 1991. Jornalistas e Revolucionarios: Nos Tempos da Imprensa Alternativa. [Journalists and
Revolutionaries: During the time of the Alternative Press.] São Paulo: Edusp.
Lauría, Carlos. 2011a. “In Latin America, A Return of Censorship.” Attacks on the Press 2010: Americas
Analysis, February 15, 2011. http://cpj.org/2011/02/attacks-on-the-press-2010-americas-analysis.php,
accessed August 24, 2012
_____. 2011b. “Confrontation, repression in Correa’s Ecuador.” Committee to Protect Journalists (CPJ),
September 1, 2011. http://cpj.org/reports/2011/09/ confrontation-repression-correa-ecuador.php,
accessed August 17, 2012.
_____. 2013. “Challenges to Democracy in the Western Hemisphere.” Testimony before the Subcommittee
on the Western Hemisphere Committee on Foreign Affairs of the United States House of
Representatives. Downloaded on October 23, 2013. http://www.cpj.org/blog/HHRG-113-FA07Wstate-LauriaC-20130910.pdf
29
Lawson, Chappell, and Sallie Hughes. 2005. “Latin America’s Post-Authoritarian Media.” In Rachel A. May
and Andrew K. Milton, eds., (Un)civil Societies: Human Rights and Democratic Transitions in Eastern Europe and
Latin America. Lanham, MD: Lexington Books.
Levitsky, Steven and Kenneth M. Roberts, eds. 2011. The Resurgence of the Latin American Left. Baltimore, Md.:
The Johns Hopkins University Press.
Marshall, Monty G., Ted Robert Gurr and Keith Jaggers. 2014. “Polity IV Project: Political Regime
Characteristics and Transitions, 1800-2013, Dataset Users’ Manual.” Center for Systemic Peace.
Matos, Carolina. 2008. Journalism and Political Democracy in Brazil. Lanham, Md.: Lexington Books.
____. 2011. “Media and Democracy in Brazil.” Westminster Papers in Communication and Culture 8(1): 178196.
McMillan, John and Pablo Zoido. 2004. “How to Subvert Democracy: Montesinos in Peru.” CEPR Discussion
Paper No. 4361. http://papers.ssrn.com/sol3/papers.cfm?abstract_id=548741
Morgan, Jana. 2011. Bankrupt Representation and Party System Collapse. Pennsylvania
State University Press.
Morgenstern, Scott and Aníbal Pérez-Liñán and Juan Javier Negri. 2008. “Parliamentary Opposition in NonParliamentary Regimes: Latin America,” Journal of Legislative Studies 14:1-2: 160-89.
Murillo, María Victoria, Virginia Oliveros, and Milan Vaishnav. 2010. Dataset on Political Ideology of
Presidents
and
Parties
in
Latin
America.
Columbia
University.
Available
at:
http://mariavictoriamurillo.com/data/ or http://www.virginiaoliveros.com.
Norris, Pippa. 2000. A Virtuous Circle? Political Communications in Post-Industrial Democracies. Cambridge:
Cambridge University Press.
Norris, Pippa, ed. 2010. Public Sentinel: News Media and Governance Reform. Washington DC: The World Bank.
Negretto, Gabriel L. 2006. “Minority Presidents and Democratic Performance in Latin America.” Latin
American Politics and Society 48(3): 63-92.
Otis, John. 2014. “Halftime for the Brazilian Press: Censorship via the courts.” Committee to Protect
Journalists Brazil Report, May 6, 2014. http://www.cpj.org/reports/2014/05/halftime-for-brazilianpress-censorship-violence-censorship-via-courts.php, accessed October 10, 2014.
Payne, J. Mark. 2007. “Balancing Executive and Legislative Prerogatives: The Role of Constitutional and
Party-Based Factors.” In Democracies in Development: Politics and Reform in Latin America (Eds.) J. Mark
Payne, Daniel Zovatto G., and Mercedes Mateo Díaz. Washington D.C. Inter-American Development
Bank.
Pérez-Liñán, Aníbal. 2007. Presidential Impeachment and the New Political Instability in Latin America. Cambridge:
Cambridge University Press.
Pérez-Liñán, Aníbal and Andrea Castagnola. 2009. “Presidential Control of High Courts in Latin America: A
Long-Term View (1904-2006).” Journal of Politics in Latin America 1(2): 87-114.
Pinto, Juliet. 2008. “Muzzling the watchdog: The case of disappearing watchdog journalism from Argentine
mainstream news.” Journalism 9(6): 750-774.
Porto, Mauro. 2007. “TV news and political change in Brazil: The impact of democratization on TV Globo’s
journalism. Journalism 8(4): 363-384.
Rafsky, Sara. 2012. “Both sides cry ‘press freedom’ in Argentine newsprint battle.” CPJ Blog: Press Freedom
News and Views, http://cpj.org/blog/2012/01/both-sides-cry-press-freedom-in-argentine-newsprin.php,
accessed August 21, 2012.
Ross, Michael L. 2001. “Does Oil Hinder Democracy?” World Politics 53(2): 325-61.
Samples, Tim R. 2010. Of Silence and Defiance: A Case Study of the Argentine Press during the Proceso of 1976-1983.
Master’s Thesis. Austin, Tex.: University of Texas Digital Repository.
30
Samuels, David J. and Matthew S. Shugart. 2010. Presidents, Parties, and Prime Ministers: How the Separation of
Powers Affects Party Organization and Behavior. Cambridge: Cambridge University Press.
Searight, Jason. 2012. Party System Collapse: The Roots of Crisis in Peru and Venezuela. Stanford: Stanford
University Press.
Sen, Amartya. 1999. Development as Freedom. New York: Alfred A. Knopf, Inc.
Shugart, Matthew S. and Carey, John M. 1992. Presidents and Assemblies: Constitutional Design and Electoral
Dynamics. Cambridge: Cambridge University Press.
Skidmore, Thomas E. 1988. The Politics of Military Rule in Brazil, 1964-1985. Oxford, U.K.: Oxford University
Press.
Smith, Anne-Marie. 1997. A Forced Agreement: Press Acquiescence to Censorship in Brazil. Pittsburgh: University of
Pittsburgh Press.
Stein, Elizabeth A. 2008. Leading the Way: The Media and the Struggle against Authoritarian Rule. Doctoral
Dissertation: University of California, Los Angeles.
Stein, Elizabeth A. and Marisa Kellam. 2014. “Programming Presidential Agendas: Partisan and Media
Environments that Lead Presidents to Fight Crime and Corruption.” Political Communication 31(1): 2552.
Stockmann, Daniela and Mary E. Gallagher. 2011. “Remote Control: How the Media Sustain Authoritarian
Rule in China.” Comparative Political Studies 44(4) 436-467.
UNDP. 2004. Democracy in Latin America: Towards a Citizens’ Democracy. New York: The United Nations
Development Programme.
Voltmer, Katrin. 2013. The Media in Transitional Democracies. Cambridge: Polity.
Von Doepp, Peter and Daniel J. Young. 2013. “Assaults on the Fourth Estate: Explaining Media Harassment
in Africa.” Journal of Politics: 75(1): 36-51.
Waisbord, Silvio R. 2000. Watchdog Journalism in South America: News, Accountability, and Democracy. New York:
Columbia University Press.
_____. 2002. “Antipress Violence and the Crisis of the State.” The Harvard International Journal of Press/Politics
7(3): 90-109.
_____. 2004. “Scandals, Media, and Citizenship in Contemporary Argentina.” American Behavioral Scientist
47(8): 1072-1098.
_____. 2006. “In journalism we trust? Credibility and fragmented journalism in Latin America.” In ed. Katrin
Voltmer Mass Media and Political Communication in New Democracies. New York: Routledge, pp. 64-77.
Weyland, Kurt, Raúl L. Madrid and Wendy Hunter. 2010. Leftist Governments in Latin America: Successes and
Shortcomings. Cambridge: Cambridge University Press.
Whitten-Woodring, Jenifer. 2009. “Watchdog or lapdog? Media freedom, regime type, and government
respect for human rights. International Studies Quarterly 53(3): 595-625.
Wiesehomeier, Nina and Kenneth Benoit. 2009. “Presidents, Parties, and Policy Competition.” The Journal of
Politics 71(4): 1435-1447.
Wooldridge, Jeffrey M. 2002. Econometric Analysis of Cross Section and Panel Data. Cambridge: MIT Press.
31
Appendix
32
Table 3. Summary Statistics
Online Appendix
Silencing Critics: Why and How Presidents Restrict Media Freedom in Democracies
Online Appendix
Table A1. The Determinants of Annual Press Freedom in Latin America, 1993-2012
(Excluding Extreme Observations)
1
Online Appendix
Table A2. Included Country-Year Observations in Annual Dataset
2
Online Appendix
Table A3. Included Presidents (* Leftist Coding)
3
Online Appendix
4
Online Appendix
5
Online Appendix
Table A4. Codebook
Variable
Coding Explanation and Source
Press Freedom Index
(PFI)
Reversal of the Press Freedom Index (PFI), an index of 3
component environments of media freedom including
economic, political and legal media freedom, scaled from 1 to
100, with 100 indicating a freer media environment. Component
values have been reweighted for 1993 to 1995 to match weights
of later years.
Source: Freedom House
Average PFI of
Democracies
Excluding Latin
America
Leftist President
Average reversed PFI of all countries for which Freedom House
ratings data are available and that score 5 or higher on Polity IV
Projects “Polity 2” Score, which measures political regimes from
autocracy (-10) to democracy (10).
Source: Freedom House and Polity IV Project
Dichotomous variable, where 1 denotes leftists, and 0 denotes
centrist, rightist, and other. We take into account rhetoric,
support base, and policy directions in classifying presidents,
holistically, as being on the left.
Source: Coded by authors, generally following Levitsky and
Roberts (2011), Murillo, Oliveros and Vaishnav (2010), Baker &
Greene (2011), Wiesehomeier and Benoit (2009), and Coppedge
(1997)
Note: Presidents coded as leftists are indicated by an asterisk (*) in Table
A3.
Landslide Win
Dichotomous variable, where 1 denotes presidents who were
elected with a margin of victory of 20 percent or more over the
second-highest vote getter in the sole or both rounds, and 0
denotes otherwise.
Source: Coded by authors, based on electoral data from
national electoral institutes and online databases
Number of General
Strikes
The total number of general strikes: "Any strike of 1,000 or
more industrial or service workers that involves more than one
employer and that is aimed at national government policies or
authority."
Source: Arthur Banks' Cross-National Time-Series Data
Archive
6
Online Appendix
continued from previous page
Variable
Coding Explanation and Source
Number of Anti-Govt
Demonstrations
The total number of anti-government demonstrations: "Any
peaceful public gathering of at least 100 people for the primary
purpose of displaying or voicing their opposition to government
policies or authority, excluding demonstrations of a distinctly
anti-foreign nature.”
Source: Arthur Banks' Cross-National Time-Series Data
Archive
Start of Banking Crisis
Dichotomous variable, where 1 denotes the beginning of a
banking crisis, and 0 indicates that no crisis occurred or that a
crisis continued from a previous year.
Source: Luc Laeven and Fabian Valencia's Systemic Banking
Crises Database, International Monetary Fund
Decree Powers
Indicates the nature and extent of decree authority accorded to
the president on a scale of 0 to 1, where 1 indicates that
presidents have constitutional authority to issue decrees that
become law and 0 indicates that presidents do not have decree
authority. Intermediate values indicate that presidential decree
authority is temporary, restricted, or must be delegated by the
congress.
Source: Table 43 in the Statistical Supplement to the United
Nations Development Programme report Democracy in Latin
America (2004/5)
Constraints on
Executive
Refers to the extent of institutionalized constraints on the
decision-making powers of chief executives, imposed by any
"accountability group," especially legislatures and political
parties. Scored as follows:
3 = slight to moderate limitations on executive authority
4 = intermediate category
5 = substantial limitations on executive authority
6 = intermediate category
7 = executive parity or subordination
Source: Polity IV Project
Judicial Independence
Indicates the extent to which the judiciary is independent of
control from other sources, such as another branch of the
government or the military, where 0 = not independent, 1 =
partially independent, and 2 = independent.
Source: The CIRI Human Rights Data Project.
Note: CIRI data are not available for 2012; we used 2011 values
for 2012. If we excluded 2012 all together, our results do not
substantively change.
7
Online Appendix
continued from previous page
Coding Explanation and Source
Federalism
Dichotomous variable, where 1 denotes political systems that are
functionally politically decentralized either because of a
constitutionally federal political system and/or because of
politically powerful elected governors, 0 otherwise. These
include Argentina, Brazil, Colombia, Mexico and Venezuela
Source: Coded by authors
Number of Journalists
Killed
The total number of journalists killed because of their work
(confirmed) or potentially killed because of their work
(unconfirmed), including retribution, combat, or dangerous
assignment.
Source: Committee to Protect Journalists
Note: Coded as confirmed killed only if CPJ reasonably believed
that a journalist was murdered in direct reprisal for his/her
work; killed in crossfire during combat; or killed carrying out a
dangerous assignment (e.g. covering street protests). Coded as
unconfirmed murder if the motive is unclear, but the journalist
potentially was killed because work (constantly updated).
Journalists killed in accidents are excluded from database.
Human Development
Index
An index of economic and social well-being including measures
of wealth, income equality, education and health. The Index
ranges from 0 to 100, with 100 indicating the most welldeveloped countries by not only their degree of economic
development, but also how well that translates to citizens' wellbeing and human capital.
Source: United Nations Development Programme
Natural Resources
(% of GDP)
"Total natural resources rents are the sum of oil rents, natural
gas rents, coal rents (hard and soft), mineral rents, and forest
rents."
Source: World Bank World Development Indicators
8