Media Wars and the New Left: Governability and Media Democratisation in Argentina and Brazil

Abstract

This article examines the determinants of government strategies towards dominant media actors in the light of increased questioning of the media's role in democratic politics. It compares the first two Kirchnerist presidencies in Argentina with the first two PT-led governments in Brazil. While these governments initially adopted accommodation strategies, political crises subsequently disturbed the coexistence of media and government, triggering divergent responses. The study offers an account of media policy options and shifts based upon (i) constraints stemming from political-structural variables, (ii) governmental perceptions of oppositional media power and of civil society allies' strength and (iii) learning from critical junctures that reshaped policy preferences. In sum, the article provides insights into the ways the media realm has become an arena of political struggle in Latin America.

Spanish abstract

Este artículo examina las estrategias gubernamentales hacia miembros de los medios de comunicación a la luz del creciente cuestionamiento del papel de la prensa en la política democrática. Compara las dos primeras presidencias kircheneristas en Argentina con los dos primeros gobiernos encabezados por el PT en Brasil. Si bien estos gobiernos adoptaron inicialmente estrategias de acomodo, posteriores crisis políticas perturbaron la coexistencia entre los medios de comunicación y el gobierno, disparando respuestas diversas. El estudio ofrece un recuento de las opciones y movimientos políticos de los medios basados en: (i) las limitaciones originadas en las variables político-estructurales, (ii) las percepciones gubernamentales del poder de los medios de comunicación opositores y la fuerza de sus aliados de la sociedad civil y (iii) el aprendizaje a partir de coyunturas críticas que reconfiguraron las preferencias políticas. En suma, el artículo da pistas sobre las formas en que el universo de los medios se ha vuelto una arena de la lucha política en Latinoamérica.

Portuguese abstract

Este artigo examina os fatores determinantes das estratégias governamentais em relação a agentes dominantes da mídia sob a luz do crescente questionamento do papel da mídia na política democrática. Comparam-se os dois primeiros mandatos Kirchneristas na Argentina com os dois governos liderados pelo PT no Brasil. Enquanto estes governos inicialmente adotaram estratégias de acomodação, crises políticas ulteriormente abalaram a coexistência entre mídia e governo, servindo de gatilho para reações divergentes. O estudo oferece uma apresentação das opções e mudanças da política em relação à mídia baseadas em (i) limitações decorrentes de variáveis político-estruturais, (ii) percepções governamentais em relação ao poder da mídia oposicionista e da força de aliados da sociedade civil, e (iii) aprendizado a partir de momentos críticos que reorganizaram as preferências de política. Em síntese, o artigo oferece insights sobre as maneiras pelas quais o campo da mídia tornou-se uma arena de disputa política na América Latina.

Introduction

A common feature of the recent wave of left-leaning governments in Latin America has been the eruption of intense ‘media wars’, which have raised questions about their causes and the consequences for democracy. Growing mediatisation combined with persistent elitism and ownership concentration in Latin America's media systems have often been advanced as an explanation. The occurrence of open conflicts and their degree of radicalism have in turn been connected to the populist or social-democratic nature of the governments involved.

I focus here on the determinants of governmental strategies towards dominant media actors in contexts of increased questioning of the media's role in democratic politics. I argue that besides prior political identities, ideological beliefs and policy preferences regarding the media, the adoption of confrontational counter-hegemonic or accommodation strategies depends on particular constraints, opportunities and perceptions. The leftist governments had to balance the demands of media democratisation to which they were committed with concerns for governability.

I compare the two first Kirchnerist presidencies in Argentina with the two first Partido dos Trabalhadores (Workers' Party, PT) led governments in Brazil. The cases comprise Néstor Kirchner's presidency (2003–7) and Cristina Fernández de Kirchner's first presidency (2007–11) on the one hand and Luis Inácio ‘Lula’ da Silva's two terms (2003–10) on the other. Despite their differences in terms of identity and path to power, the initial phases of these governments were strikingly similar regarding their media politics. In fact, Lula and Kirchner each inaugurated their presidencies in 2003 by adopting an accommodation strategy towards their country's dominant media organisations. These pragmatic approaches were reflected in political decisions favouring Globo and Clarín, Brazil and Argentina's largest media organisations, and in the preferential treatment given to their journalistic outlets.

In both cases, political crises, the Mensalão scandal in Brazil 2005 and Argentina's agrarian conflict in 2008, shook this initial equilibrium, shifting the way these outlets covered government. However, the two governments responded differently to these new scenarios. Against what might be expected, the government that emerged politically weakened in the aftermath of the crisis opted for a radical confrontation; while in Brazil the PT-led government, bolstered after the 2006 electoral triumph, chose a comparatively moderate course of action. As I will show, while Lula abandoned the former accommodation strategy after 2005, he did not enter into the overt war with leading media players that has characterised Kirchnerism since 2008. These divergent reactions require explanation. Based on primary ¹ and secondary sources, I offer an account of media policy options and shifts. I first seek to determine constraints stemming from political-structural variables such as (i) the existing congressional correlations of forces, (ii) the linkages between political and media elites, (iii) the way electoral systems affect media influence upon legislators and (iv) the level of media sector interest group organisation. I then stress the role of the perceived availability and strength of progressive allies in civil society and media activist organisations, and of the politicians' subjective assessments of media power. These last two factors, I will show, are not static but vary upon learning from critical junctures. Reconstructing the way governments interpreted certain key episodes helps illuminate the rationale behind the strategies adopted.

Government and Media

Variation in governmental media strategies in the context of Latin America's ‘left turn’ has often been described but seldom theorised. Most accounts establish a connection between radical confrontational strategies towards the established media and the populist variant of leftist governments. ² Waisbord argues that Latin American populists share a view of the media that is rooted in a mixture of Marxism and nationalism: the media are the powerful instruments of domination of anti-popular minorities. Such a view supports a call for radical reforms. While not denying other intervening factors, this perspective suggests that ideas and beliefs play a prominent role in policy formation.

This view, however, does not account for the cases. Lula, his inner circle and the PT in general shared critical beliefs about the media as elite or class actors, and advocated media democratisation reforms. ³ The Kirchners, in contrast, had no previous public commitments on the subject. Still, the latter eventually pushed for an aggressive reform, whereas the former did not translate their ideological commitments into policy.

Furthermore, programmatic beliefs about the media do not explain why both governments initially accommodated media interests and why, in the aftermath of equivalent crises, both recast their strategies in the ways they did. If ideas are the primary factors shaping governmental choices, then political actors should be much less sensitive to external environments than appears to be the case. ⁴ This is not to say that ideas play no role at all, but that they do so only under favourable conditions. Ideas do affect the ways in which actors make sense of the world they face; political-institutional settings and vested interests condition their viability. ⁵

Turning away from voluntaristic accounts of policy choices as simple expressions of ideological preferences, scholarly research on the variations in the way the Latin American Left actually governs has focused instead on the context. Economic and social policy variation, especially the degree of radicalism and elite defiance, has been explained by politico-institutional variables such as path to power, party system and political competition patterns. ⁶ Several studies on Brazil have reconstructed the political constraints and veto points that curtailed the ability of the PT-led government to translate its preferences into policy. ⁷ These ‘hard’ institutional factors influence governmental perceptions of the political viability of different policy options vis-à-vis the media.

Outside formal political institutions, media institutions can themselves function as constraints or even veto players. In Latin America's instrumentalist media cultures, large media conglomerates constitute strategic actors for governability. Since they concentrate control over key resources (ideas, information, public agendas, legitimacy and reputations) and have the ability to act en bloc to pursue their interests, they potentially influence effective policy-making and, eventually, the ability of a government to stay in power. ⁸ In the media actors' case, the subjective dimension of constraints, the fact that they are perceived rather than objective, becomes clear. Media power is, ultimately, reputational power: ‘As long as politicians … believe that media has great effects, they will usually act in consequence. At the regional level, Latin American politicians perceive huge media influences on policy making.’ ⁹

In contrast to understandings of the media originating from ideologies, these perceptions are usually forged contextually along paths to power. However, the ways in which media power reputations coexist or compete with ideological beliefs, and how they affect strategic shifts, cannot be settled merely through static descriptions of institutional constraints or governmental beliefs about the media. While important, pre-existing beliefs must instead be examined in their interplay with short-term experiences and with challenges faced during the governing process. Therefore timing-sensitive narratives illuminate how such developments reinforce or alter pre-held beliefs, and how they influence the way in which governmental decision-makers evaluate correlations of forces and their own chances of survival, adjusting their strategies accordingly.

I will hypothesise two such learning processes affecting governmental perceptions and strategic choices. The first consists of episodes that impact on the government's evaluations regarding the availability and effectiveness of allies in civil society for certain courses of action. When considering choices, governments estimate the likely availability of support for an eventual switch towards confronting established actors. ¹⁰ Defying the dominant media assumes that one has the capacity to mobilise credible voices and to resist hostile narratives in the media-dominated public sphere.

The second learning process concerns the dynamics of media power perceptions. The real influence that media content has on public attitudes and behaviour is constitutively opaque. Governments fill this cognitive gap with judgements of causality instead. Sequences of events influence how judgements about likely media effects are constructed. In the present experiment, both governments were placed in a setting of sudden media coverage that was perceived by them as hostile, followed later by elections. The electoral results gave the governments a measure of media power. The causal force assigned to media coverage, in turn, informed these governments' assessments of their prospects for survival, and is central in reconstructing the rationale underlying their strategic choices and shifts.

Parallel Developments

By 2000, Argentina's and Brazil's media systems were, like most of Latin America's, exhibiting high levels of audience and ownership concentration and a persistently elitist character. Both media systems were characterised by the presence of dominant, family-controlled media conglomerates.

Through its alliance with the military regime, by the 1980s Roberto Marinho's Rede Globo had achieved absolute dominance in Brazil's television market, averaging a 75 per cent share of the national audience. The first direct presidential race in 1989, in which Lula was defeated by Fernando Collor, to whom the network was committed, represented the apex of TV Globo's political influence. Throughout the 1990s, opportunities for competitors opened up. However, until 2008 TV Globo managed to keep its audience share above 50 per cent, far above the 16 per cent of its nearest rival. By 2002, with 223 stations or affiliates, the network was reaching 55 million homes. In 2005 its total revenue was triple that of the country's second and third networks, and it received over half of Brazil's advertising budget. The conglomerate additionally controlled newspapers, newsweeklies, radio stations, pay-TV networks, production companies and other interests in cultural industries. However, TV Globo's hegemonic position rested not only on its overall size but also on its communicative-symbolic power. Its newscast, the Jornal Nacional (JN), the first to be aired simultaneously on all affiliate stations in 1969, played a central role in consolidating a sense of national identity, and since then it has been by far the most influential national news provider and agenda setter. In addition to its TV journalism, Globo's telenovelas have had an unprecedented influence on Brazilian popular culture. ¹¹

Argentina's Clarín started as a modest newspaper in 1945. It was only during the 1970s that it began to expand. Clarín's pragmatic ties to the military regime led to an association with the state in the production of newsprint. By the 1980s it had become the country's largest and most influential daily. Its expansion and the parallel crisis of traditional political actors increased Clarín's autonomy from the political field and enhanced its capacity to organise the public agenda. The liberalisation and privatisation process initiated in 1989 under Carlos Menem began its transformation into a multimedia conglomerate. The removal of cross-media ownership restrictions enabled the newspaper to move into the broadcasting sector, where it acquired leading television and radio stations.

Throughout the 1990s the Clarín Group expanded from publishing and broadcasting to cable TV and internet provision, news agencies, audio-visual production and soccer transmission, among other interests. Controlled by Ernestina Herrera de Noble, the founder's widow, and Hector Magnetto, its CEO, the group became one of the country's leading economic conglomerates. Since 1999 it has been financed through international capital markets. Its undisputed ascendancy in public agenda and opinion formation has been made possible by its multiple popular outlets and by its prestigious news media's capacity to act in a coordinated fashion, headed by its newspaper and its 24-hour news channel. ¹²

The politics of accommodation

The electoral triumphs of Lula in 2002 and Kirchner in 2003 formed part of the initial wave of leftist victories that occurred in the context of the regional economic downturn from 1998 to 2002. Despite their different paths, Kirchner, a Peronist from a remote province, emerged in the aftermath of the 2001 crisis as a political outsider; Lula, a union and leftist party leader, followed a gradual party-institutional route. Both rose to government promising an alternative to neoliberalism. ¹³ However, these promises coexisted with concerns about governability, creating important incentives to privilege key strategic actors like Globo and Clarín.

Kirchner took on the presidency in the aftermath of the 2001 mass protest mobilisations and political representation crisis. Due to Menem's defection from the runoff, Kirchner eventually won the presidency with only 22 per cent of the vote. The newly elected president tackled his first challenge, to reconstruct political authority, by promising a ‘renewal of politics’ aimed at building support among the sceptical urban middle classes. In this context, his relationship with the public, the media and especially the Clarín Group assumed strategic importance for his political future. ¹⁴ As discussed below, Kirchner shared and overstated the ‘common sense’ of political leaders regarding Clarín's role in governability. Kirchner cared about Clarín not only because of its perceived power but also because he believed that its outlets' audiences overlapped with his own potential constituency. ¹⁵

In 2002 Lula competed for the presidency for the fourth time. A long de-radicalisation process had made his 1989 leftist platform seem far away. While he had once denounced the media as part of the establishment, Lula's professionalised campaign hired marketing gurus to communicate a moderate message. To appease the establishment, Lula's inner circle decided to signal continuity with the macro-economic orthodoxy in a ‘Letter to the Brazilian People’. To ensure a victory in the runoff, the PT allied with non-leftist parties and leaders who were, in turn, allies of the Globo network. Predictably, demands for media democratisation, part of the historical agenda of the PT and its civil society allies, disappeared from the electoral programme. Shortly before the election Lula held private meetings with the Marinho family. The night he was elected he appeared on a popular Sunday show broadcast by TV Globo. The next day he sat beside the JN anchor during the entire newscast, commenting on the elections. ¹⁶

For their part, the media conglomerates also had incentives for seeking accommodation at the outset of Lula's and Kirchner's presidencies. In the context of the 1990s inflation control policies, which pegged the respective local currencies to the US dollar, both groups made investments and thereby amassed significant debt in US dollars. When currency devaluations and market contraction resulting from the 1998 crisis occurred in 1999 in Brazil and in 2002 in Argentina the two conglomerates were left in a vulnerable position. In 1999, Globo's debt amounted to US$ 2 billion. By 2002, Clarín faced the possibility of being taken over by its creditors.

Even before Lula and Kirchner took office they signalled goodwill regarding the groups' problems. In mid-2002, prior to the upcoming elections, the PT supported a constitutional amendment that would allow up to 30 per cent foreign capital in communications enterprises. During its first days, Lula's government showed a willingness to extend credits to Globo through the National Development Bank. ¹⁷

During Duhalde's interim government just prior to Kirchner's first term, and after intense lobbying by Clarín, Congress had approved a law establishing a cap on foreign ownership of ‘cultural industries’, thereby preventing the group's holders from losing control of their assets to foreign creditors. This initiative, dubbed by critics the ‘Clarín Law’, was sanctioned by the executive a few days after Kirchner took office.

It is plausible that Clarín's financial weakness reinforced the conglomerate's initial benevolence towards the Kirchner administration. Observers also consider it to be an example of a broader pattern wherein Clarín agrees with governments in their initial phase, only to become a tenacious critic once it has achieved its goals. ¹⁸

Certainly, these considerations do not exhaust the reasons for the initial benevolence. For Globo, as for any media institution, opposing Lula's rising popularity would have been costly. Kirchner, meanwhile, took office as an unknown figure. His popularity only increased later. Given the context of political and economic instability in Argentina, however, Clarín's CEO and the conglomerate's journalistic leadership stood behind the initial government decisions, which were seen as conducive to national recovery. ¹⁹

The initial coverage reflected these incentives. Globo seemed to celebrate the election of the political leader it had traditionally opposed. In fact, in contrast to the paulista print media, which was critical from the outset, the group's ‘[network] news coverage of Lula's first months in office was characterised by a very positive tone’. ²⁰ Clarín, for its part, praised the main policies, exalted the president's virtues and took a mostly gentle tone during Kirchner's first three years in office. ²¹

Both administrations reciprocated. During his presidency, Kirchner invested in a personal relationship with Magnetto. He also appointed an officer sympathetic to sectoral requests to the broadcasting regulatory authority. Clarín was favoured by several decisions. In 2004, the main national broadcasters had their licences renewed for ten years. In 2005, a decree granted an additional ten-year grace period for all television licensees. Three days before handing over the presidency to his wife, Kirchner approved the merger of the country's two leading cable providers, granting Clarín a dominant position in cable TV and internet provision markets, a sector that, by 2007, represented 60 per cent of the conglomerate's revenues. Additionally, Clarín retained a significant share of the official advertising; it benefited from the blocking of new cable distribution permit requests while its journalistic outlets obtained privileged access to and scoops from government sources. Until 2008, Kirchner tried to favour the group's expansion ambitions regarding the telecommunications sector. Simultaneously, however, he also sought to counter-balance the group's dominant position by trying to tempt foreign business organisations, Telmex and Prisa, to enter the Argentine media sector and by strengthening other local entrepreneurs. ²²

While Kirchner was pragmatic at the level of owners and editors, his rapport with the press was tense from the outset. Apart from choosing the conservative La Nación as an ideological opponent right from the beginning, Kirchner tended to present journalism as dependent on non-elected powers and intellectually subordinate to neoliberal-technocratic common sense. The executive viewed itself as leading the recovery of politics' autonomy in the face of the media's fake neutrality. This viewpoint, sporadically present in discourse from 2003 on, would become commonplace after 2008. At the level of governmental communications practices, Kirchner routinely resorted to controlled events, exerted strict vertical control over sources, and avoided conventional interactions such as interviews and press conferences. These tactics contributed to tension with journalists. ²³

The early accommodation impulses of Lula's administration can also be detected in a series of government decisions and non-decisions in addition to the above-mentioned privileges accorded to TV Globo. The PT had historically been aligned with media democratisation agendas, and the civil society organisations that had come together during the 1990s in the Fórum Nacional pela Democratização da Comunicação (National Forum for the Democratisation of communications, FNDC) were organically linked to the party, with overlapping memberships. While these democratisation demands had been silenced during the 2002 campaign, there was uncertainty about what would happen afterwards. This contrasts with Argentina's experience. As shown below, the Brazilian government's initial strategy of accommodating Globo also entailed containing or deflecting pressures from inside the party and from its core allies.

With Lula in office, the party's radical factions and other leftist allies in Congress renewed their enthusiasm for media democratisation. As Table 1 shows, the number of legislative initiatives to regulate broadcasting jumped abruptly in 2003. This activism was not, however, echoed in Planalto Palace.

Table 1. Legislative Initiatives Affecting the Broadcasting Sector in both Federal Chambers

Source: Author's adaptation of data from ABERT, Relatório de gestão. Bienios 2006/2008, 2008/2010.

The key agency appointments made the accommodation impulses and containment strategies within the government even more apparent. The Ministry of Communications (MINCOM) had historically been ceded by the executives to close allies of the big broadcasters. In 2003, despite initial expectations, the MINCOM did not go to the PT. Miro Teixeira, Lula's initial choice, was an ambiguous one. While his party was heir to Varguista trabahlismo, he represented a pragmatic position and cultivated good relations with the Marinhos and other big broadcasters. Despite his appointment of some FNDC cadres in lower tiers, no significant threat to Globo emerged from Teixeira's tenure. ²⁴

However, in the context of the transition to digital television, a decree sanctioned under Teixeira stimulated the creation of a Brazilian standard and created an advisory board that included representatives from civil society, thus creating some institutional space for voices that viewed this transition as an opportunity to subordinate commercial interests to civic-political goals. This initiative represented a challenge to Globo's interests, which, in association with the electronic equipment supplier NEC do Brasil, had invested in adopting the Japanese standard. By January 2004, Teixeira had been replaced by a politician from the conservative Partido do Movimento Democrático Brasileiro (Party of the Brazilian Democratic Movement, PMDB), yet the progressive groups committed to developing the Brazilian norm kept working autonomously. With the outbreak of the Mensalão, Lula was forced to redistribute cabinet posts. As part of this reshuffle, Hélio Costa, another conservative PMDB member and himself a former TV Globo correspondent and media entrepreneur, was designated head of MINCOM in July 2005. As many observers agree, his main agenda was to guide the digital transition in Globo's interest. In 2006, the Japanese–Brazilian standard was adopted by decree. ²⁵

These pragmatic moves towards Globo coexisted with tensions between government and press institutions. After a brief honeymoon, aggressive press coverage revived old petista critical views of the media and divided Lula's inner circle over the attitude to be assumed. Although torn himself, Lula's prudential avoidance of confrontation mostly prevailed. ²⁶ Additionally, a trend that would fully emerge after the 2005/06 crises was already visible: Lula embodied the man of the povo (common people) who had reached the top political position, previously the exclusive domain of social elites. This brought latent class prejudice in reporting to the fore, while his direct-communication appeals were denounced as ‘populist’. Journalists were further alienated by the adoption of the governmental communications practices including regular presidential broadcasts (‘Café com o Presidente’), refusal of journalistic conventions, and staged events used as agenda-setting devices. ²⁷

Amidst this initial tension, two brief but disruptive episodes would have important consequences. Both were sparked by allies' initiatives, not by the executive itself. First, in the aftermath of the scandal around a New York Times column that claimed Lula had a drinking problem, the journalists' union promoted the creation of a Federal Journalism Council, a non-governmental body intended to set ethical standards and regulate professional conduct. Second, a group of progressive filmmakers working at the Ministry of Culture proposed the creation of a National Agency for Cinema and Audio-visual Activities with regulatory authority over the audio-visual sector, including television. Both initiatives triggered fierce media reactions, with the projects presented as threats to press freedom and expressions of the government's authoritarian-statism. These reactions led government to rapidly shelve both proposals. While the first initiative reached the legislative proposal stage, the second was leaked to the press as a first draft.

Political crisis and media coverage

In both cases, it was a political crisis that disrupted the equilibrium. In Brazil the crisis was initiated by a corruption scandal, while in Argentina it was triggered by an agrarian conflict. The coverage of these episodes by Globo's and Clarín's news outlets, aligned to most of the mainstream media, signalled a shift. In both cases, the government and their closest allies interpreted this change as an unmistakable sign of political activism by the media elites, ultimately aimed at ousting them from the presidency.

The fragile harmony between the government and Clarín began to come under strain during the last year of Kirchner's presidency. Divergences over policy and business interests increased mutual distrust and were reflected in news coverage. ²⁸ But the turning point came a few months after Cristina Fernández's appointment, in the context of the agrarian conflict that occurred between March and June 2008. After the government passed a resolution that adjusted export taxes for agricultural products to fluctuations in world market prices, farmers launched a lockout with massive roadblocks, thereby paralysing the country.

As the agrarian conflict unfolded and farmers progressively gained support from significant segments of the urban middle classes, the Kirchners blamed the growing hostility towards the government on Clarín's coverage of events. The Clarín outlets' framing of the protest as a ‘historical strike’ and the extensive live coverage of the protesters convinced the government that Clarín was taking sides with those who wanted to oust the president, a persuasion it started to voice publicly. ²⁹

The government was not alone in harbouring this perception. A wider segment of society, especially middle-class progressives, started to criticise the biases of Clarín and the other mainstream media organisations. These voices constituted the support base for the Kirchners' nascent counter-hegemonic media policy. One central aspect of the coverage that mobilised these voices was the prevalence of a binary narrative based on class and race stereotypes. The media presented the anti-government protests as legitimate spontaneous actions by honest, productive, autonomous white citizens in opposition to the spurious political-machine-based mobilisation of poor, non-white clienteles. ³⁰

Yet the framing contest was a difficult challenge for the government. To counter the frame that opposed a freely mobilised people to a clientelist machine, the government utilised the classic populist divide between the people (pueblo) and the powerful minorities (oligarquías). Eventually, though, the opposition became so massive that the government lost the Senate vote over the export tax resolution in June 2008. This outcome would contribute decisively to the subsequent dynamic.

From June 2005 until the 2006 elections, a series of scandals shook Brazilian politics. The Mensalão scandal began with revelations by a federal deputy of a scheme consisting of monthly allowances to congressmen in return for their support for government legislation. The scheme involved PT leaders and top government officials. This was followed by the Sanguessugas scandal, which uncovered paybacks to legislators (from both the government coalition and the opposition) for the approval of budget amendments for the acquisition of ambulances. On the eve of the presidential election, the police detained two members of the PT with a large sum of money in a hotel. They were supposedly buying photographs that proved the involvement of Partido da Social Democracia Brasileira (Brazilian Social Democracy Party, PSDB) oppositional leaders in the aforementioned ambulance purchases. The media's use of the images of the arrest triggered the so-called ‘dossier scandal’. While Lula was ultimately re-elected in a runoff, the exposés had a high political cost: many of the president's close advisors had to resign, and the PT's image was severely tarnished.

With the outbreak of the Mensalão, TV Globo abandoned its previously benevolent attitude and joined the main newspapers in their increased hostility towards both the president and his party. Corruption became the almost exclusive topic of political coverage. As different observers noticed, a denunciation frenzy (denuncismo) possessed journalists, frequently at the expense of professionalism. Moral categories dominated the frames used in political reporting, and other relevant dimensions of politics and policy processes were omitted and distorted. ³¹

These tendencies intensified as the 2006 elections approached. According to Porto, while TV Globo maintained relatively balanced airtime, ‘Lula's coverage was predominantly negative and focused heavily on denunciations against him and his party’. JN devoted unprecedented attention to the elections, two-thirds of its airtime in the final two weeks. Most of the coverage ‘dealt with the dossier scandal that came to light in the last two weeks of the first round’. Two days before the first round, JN repeatedly broadcast pictures of the confiscated money, thus amplifying the scandal. Subsequent research has suggested that this media frenzy had an impact on voting behaviour and prevented Lula from winning the first round. ³²

As a result of the reporting on these scandals, the perception of a media bias against the PT government, deliberate or not, aimed at interrupting its mandate or at least impeding its re-election, gained ground among government officials, party members, leftist militants, progressive intellectuals, journalists and sectors of civil society. ³³ At the peak of the crisis and with impeachment a pending possibility, Lula mobilised support from among his social base. In June 2005, over 40 organisations published a statement in which they accused the elites and the mass media of ‘launching a campaign to demoralise both the government and the president in order to undermine his administration or to overthrow him’. ³⁴

The overt bias also drew criticism from scholars, journalists, activists and media observers. Media behaviour became a matter of debate. The 2005–6 crisis coverage seemed to many to be a regression to Globo's manipulative practices of 1989. Others criticised coordinated behaviour to control the public agenda on the part of ‘big media’, taken to include Globo, Editora Abril, Folha Group and Estado Group. Besides corporate interests, critics detected ideological and class factors as also being behind antilulismo and antipetismo. ³⁵ A documentary comparing Marinho to Citizen Kane was widely circulated in leftist circles. Militants frequently referred to the golpe midiático of April 2002 in Venezuela as an antecedent. Others pointed to lacerdismo and the role of the press in the 1964 coup. The expression ‘PIG’ (Partido da Imprensa Golpista) became widely used among critical journalists and leftist bloggers.

The airtime devoted by JN to the photos of the dossier scandal generated special controversy. The government-leaning newsweekly Carta Capital claimed that the photos had been leaked deliberately to damage Lula's candidacy. It accused TV Globo and several newspapers of actively plotting against Lula by omitting important facts about the leak itself and by failing to report the involvement of PSDB politicians in the scandal. This context resulted in resignations, firings and protests by journalists and media professionals working for TV Globo and for other media. ³⁶

The case of Franklin Martins, a former guerrilla fighter in the 1960s and now appointed columnist for JN, would have a major impact on subsequent developments. In May 2006, Globo cancelled his contract, alleging reasons based on ‘audience research’. Martins attributed his dismissal to his ‘refusal to join the news media's tendency to present facile denunciations that lack sufficient evidence’. Martins' ‘commentaries were usually more cautious about the allegations involving the president, when compared to the general tone of the news media’ and, according to him, ‘TV Globo's executives and editors grew increasingly uncomfortable’ with it. ³⁷

Divergent Paths

Despite the striking similarities in the initial settings and subsequent development of the relationships between governments and the dominant media actors, the critical junctures of 2005–6 and 2008 drew divergent responses from the Lula and the Kirchnerist administrations. However, the differences in the strategies deployed by both governments should not obscure a commonality: both responses signalled a departure from what Gómez Bruera calls ‘elite-centred strategies’. ³⁸ In other words, they shared the underlying diagnosis that relying exclusively on accommodating strategic actors and accepting given power relations in the media sphere was no longer a viable, or, at least, reasonable, strategy. This insight opened up some space or opportunities for reformist agendas that had previously been ignored.

The Kirchnerist response

From the end of March 2008, a few weeks after the agrarian lockout had begun, the steady escalation between the Argentine government and the ‘dominant media’ (mainly Clarín) developed into all-out confrontation. This ‘mother of all battles’ was fought on every front. The government deployed all its political resources, with the ultimate aim of radically altering power relations in the media sphere, in other words, of crushing Clarín's dominant position.

In April 2008, beginning with the replacement of the head of the broadcasting regulatory agency, the Kirchners approached civil society sectors committed to achieving legal reforms to democratise the media. After a long process, this alliance concluded with the sanctioning in October 2009 of a comprehensive media law, the Ley de Servicios de Comunicación Audiovisual (Audio-visual Communication Services Law, LSCA). The legislation mainly addressed plurality and diversity through structural regulations aimed at reversing media ownership and audience concentration via bans on cross-ownership, limits on broadcasting licence numbers, and subscriber caps for pay-TV services, among other policy mechanisms. Under such rules the Clarín Group and other organisations, albeit to a lesser degree, would be forced to divest a number of assets. ³⁹

The agrarian conflict also triggered the Kirchners' going public with a discourse that defined the ‘dominant media’ as the real and unelected opposition. This depiction of the media gradually developed into a cultural war, fought on screens, papers and in new media; and an expanding circle of allies popularised academic, media-critical discourses that deconstructed the ideological, corporate and journalistic biases in dominant media narratives on a daily basis. These Gramscian tactics were central to mobilising support during the legislative debate on the LSCA. ⁴⁰

Simultaneously, myriad political, judicial and administrative measures were undertaken to undermine Clarín's interests and credibility. The public questioning of Clarín's role during the dictatorship, especially the alleged appropriation by Noble's widow of children of the desaparecidos and the acquisition, together with La Nación, of shares in the newsprint factory Papel Prensa, supposedly through extortion and in complicity with the military, led to the risk of criminal prosecution for Clarín's and La Nación's owners. ⁴¹ In 2009, the government bought the transmission rights for first-division soccer from the national football association to broadcast games on free-to-air television. In doing so, the government overturned the long-standing, exclusive possession, since 1991, of these rights by a pay-per-view channel co-owned by Clarín, a key resource in the expansion of the group's business operations. Additionally, the government unblocked the distribution of cable licences, decided to adopt the Japanese–Brazilian standard for digital television (although Clarín had an interest in the US standard), reduced official advertising in Clarín's outlets, and fuelled new pro-government media.

The Lula government's response

The Brazilian government's response to the media hostility unleashed by the Mensalão contrasts with the Kirchnerist confrontation strategy. However, its approach cannot be accurately described as continued accommodation. In his second term Lula in fact exhibited a hybrid strategy, with elements of both accommodation and defiance.

Lula's government did not embark on legislative reforms of media regulation. Nor did it challenge the broadcasting status quo by appointing uncomfortable figures linked to reformist agendas to head regulatory authorities. Hélio Costa, mentioned above, remained communications minister for the whole of Lula's second term.

Nevertheless, significant changes gradually became visible. Despite his avoidance of all-out public confrontation, Lula's critiques of media bias and distortion became more audible and were increasingly framed in terms of ‘them’ against ‘us’ so as to emphasise the Brazilian media's elitism. ⁴² The government ended TV Globo's privileged relationship, increased its own use of direct-communication devices, and increasingly circumvented journalists. In addition, during Lula's second term new spaces opened up inside the state apparatus where certain reforms were promoted and civil society was mobilised around establishing a communications reform agenda by the end of the presidency.

In March 2007, Lula appointed Franklin Martins, the fired JN columnist, as head of the Presidential Secretariat for Social Communication, thus sending an unmistakable signal that the government was distancing itself from Globo. From the Secretariat, Martins deepened the redirection of government advertising away from traditional to alternative and regional media outlets. Authorised by the president, he also went beyond the Secretariat's established tasks and intruded on some of MINCOM's responsibilities. In 2008, in a move driven by Martins and justified with the need to counter-balance market-based logic and promote debates that commercial television was not interested in, the government created a national public media system that united existing public radio stations and created a national public television station.

This promotion of debates and institutional opening remobilised civil society, which in turn helped invigorate human rights and civic-oriented policies in the media realm. This feedback effect was clearly visible in the long struggle to regulate children's rights regarding television through the establishment of broadcasting hours regulations and a content rating system for children, which, given Brazil's multiple time zones, affected the scale economies of national broadcasters. These initiatives had existed since Fernando Henrique Cardoso's presidency. However, until 2007, big media had successfully vetoed each regulatory attempt through PR campaigns and judicial action. In the new context, however, the government finally succeeded in upholding a Ministry of Justice decree imposing certain obligations such as programme classification rules and children's viewing hours over unfettered profit maximisation.

The climax of this government-sponsored remobilisation of civil society was reached in 2009 at the World Social Forum, when Lula announced a Federal Communication Conference (CONFECOM). This national public policy conference, a participatory constitutional mechanism, entailed extensive mobilisation at the municipal, state and federal levels. In December 2009, over 1,600 elected national delegates voted on over 570 proposals that, according to Lula, would constitute the input to a future regulatory law. The participants, who were demonised or ignored by big media, were not limited to NGOs or social movements. While the Brazilian Broadcasters Association (ABERT), dominated by Globo, and the National Newspaper Association (ANJ), representing the big newspapers, retired from the debates, other business organisations representing telecommunications companies and dissident broadcasters remained. Lula's ability to manage intra-business conflicts of interest undermined Globo's capacity to keep regulatory debates behind closed doors. By the end of his second term Lula had established an unprecedented level of public debate on the need for democratic media regulations in Brazil. This new climate, as Table 1 has shown, sparked a new peak of legislative activism between 2009 and 2010. Six months before handing over power to Dilma Rousseff, Lula created a commission, led by Martins, intended to generate proposals for a new regulatory framework for broadcasting. ⁴³

As a former FNDC member put it to me, the governmental strategy since Lula's second term has not consisted of open confrontation but rather of ‘eating from the borders’: Lula expanded state activity in those areas with less resistance, enabled critical voices in civil society, promoted some public debate on the need for media regulation, and isolated the traditional media.

Explaining variations in strategy

This section explores first the opportunities and constraints in the institutional, media and social arenas affecting the perceived viability of different strategic choices. Second, it addresses strategic shifts focusing on dynamic processes that affected governmental beliefs.

An examination of the correlation of forces in Congress shows why legal reform was never an option during Lula's presidencies. Brazil's institutional design imposes severe political constraints on governments. Its combination of presidentialism, federalism and proportionality in the national legislature has strong fragmenting effects on party representation, forcing strategies of ‘coalition presidentialism’. As a vast literature attests, to achieve political governability, presidents need to distribute resources and jobs, especially cabinet posts, so as to form broad, pragmatic legislative coalitions. The PT government was no exception to the rule. In fact, the Mensalão was a consequence of the executive's initial difficulty coping with coalition building. In 2003, the PT, the biggest legislative bloc, obtained 91 of 513 (17.7 per cent) seats in the lower chamber. In the Senate it was third in size, with only 14 out of 81 seats (17.3 per cent). Adding leftist allied parties, the bloc reached an estimate of 30 per cent of lower chamber seats. Damaged by the scandals, the PT was in an even weaker position after the 2006 elections. In the lower chamber it secured only 83 seats, while in the Senate it formed the fourth-largest bloc with 11 members. From a comparative perspective, ‘the PT had the lowest number of congressional representatives of any other progressive party in government in Latin America at the time’. ⁴⁴ As a consequence, during Lula's presidencies the PT-led government included between eight and 12 parties, many of them ideologically heterogeneous, in its legislative coalitions.

This feature, a barrier to reformist agendas in general, looked even worse in the realm of media regulation given the linkages between political elites and media powers. These links had begun to take shape under President José Sarney (1985–1990). After appointing Antônio Carlos Magalhães, a conservative politician and close friend to Marinho, as communications minister, Sarney distributed 1,028 broadcasting licences as a way to obtain support, especially on the eve of the 1988 constituent assembly, at which time 91 legislators obtained licences. ⁴⁵ Known as coronelismo eletrônico (‘electronic clientelism’), a new structure that articulated state oligarchies and media empires was consolidated. Sarney and Magalhães best exemplified this structure as they built media groups in their respective states of Maranhão and Bahia that, by becoming Rede Globo affiliates, secured audience domination. In the 1990s, the most important stations in 12 of 21 federal states functioned as Globo affiliates and all were owned by local oligarchs. ⁴⁶

Since 1988, licence concessions require a legislative decree besides executive approval. Consequently, the 1988 constitution banned legislators from holding broadcasting licenses. Nonetheless, a significant number of legislators still directly or indirectly own broadcasting stations. In 2005, 51 federal deputies held such licences. In 2007, 23 senators were identified as direct owners. Observers consider a so-called ‘bancada da mídia’ (media bloc) to exist, which is estimated to total about 30 per cent of congressmen. ⁴⁷

These entanglements have also been evident in the Science, Technology, Communications and Informatics Committee (CCTCI) of the lower chamber. In 2003, 16 of its 51 members (including the president) held 37 licences. In 2004, 15 of 33 members held 29. In 2007, five out of 17 senators in the upper chamber's equivalent committee were licensees. Most of these legislators regularly voted on licence requests and renewals, sometimes even in cases concerning themselves. ⁴⁸

PT legislators and their leftist allies did not participate in electronic clientelism. However, it is common practice among the PT's non-leftist congressional allies. In the asset declarations registered in 2011, 59 of 513 deputies (11.5 per cent) admitted to holding licences. Of the 140 representatives from the three biggest allied parties (PMDB, Partido da República, Partido Democrático Trabalhista), 21(15 per cent) are licensees, a figure greater than the chamber average. ⁴⁹

The correlation of legislative forces looked very different in Argentina under Kirchnerism than it did in Brazil. In the 2007 elections the government's Frente para la Victoria obtained 130 seats out of 257 in the chamber of deputies. After the agrarian conflict, many legislators deserted and the government bloc decreased to about 110 seats. However, the media law's reformist nature, its recognition by international organisations' freedom of expression rapporteurs, and some concessions for amendments helped the legislative coalition expand to include opposition parties, mostly from the Left. ⁵⁰ The law obtained 147 votes, with 105 of them from the official bloc and 42 from opposition representatives. In the Peronist Party-dominated Senate, the bill passed with 44 affirmative versus 24 negative votes. These disciplined majorities were also eased by the fact that the government pushed the bill immediately after the mid-term elections, further reducing the media's capacity to pressure legislators.

In Argentina, local politicians frequently control local media. However, these outlets are not linked to big media as in Brazil's affiliate system. Therefore, Argentina has no equivalent trend to Brazilian coronelismo eletrônico. Moreover, because Congress has no jurisdiction over broadcasting concessions, which remain an executive prerogative, nothing comparable to the ‘bancada da mídia’ has developed.

Media organisations' impact on political careers should also be considered a congressional constraint. Brazilian federal representatives are elected via an open-list proportional representation system that prompts voters to choose individual candidates based on personal qualities and activities. This makes media access, alongside pork-barrelling, a central campaign resource. ⁵¹ In contrast, Argentina's closed blocked-list electoral system increases the role of parties and discipline over individualism, distancing career success from media performance. This makes Argentine legislators less sensitive to media interests. ⁵²

Moreover, in Argentina, the broadcasting sector lacks strong interest groups capable of successful lobbying, as a consequence of the fact that broadcast television was under state control between 1973 and 1989. ⁵³ In sharp contrast, ABERT, the Brazilian sectoral organisation representing commercial broadcasters was born in 1962, when it successfully repealed most of João Goulart's presidential vetoes to a telecommunications regulatory bill and initiated a long tradition of legislative lobbying against reformist attempts. ⁵⁴

Still, the constraining power of media organisations is not predominantly reliant on their capacity to lobby through formal institutions. As already suggested, media power is a function of its reputation among political elites for being capable of affecting careers and governability. Surveys show that all over Latin America political elites increasingly perceive media actors as leading ‘de facto powers’ (poderes fácticos). ⁵⁵ The differences in the intensity of these perceptions between Argentina and Brazil are difficult to assess. One proxy could be general media credibility, assuming that credibility at least partially influences public opinion. Survey data reveals persistently higher media credibility in Brazil. In a 2003 poll by Latinobarómetro, 74.1 per cent of Brazilian respondents expressed ‘satisfaction with media objectivity’ while only 45.9 per cent of Argentines did so. LAPOP data from 2010 assessed media confidence on a 0–100 scale. While Brazil obtained 69.9 points, Argentina only received 53.6. This suggests that higher media credibility could correlate with accommodation strategies, because confrontation is perceived as having a higher cost.

Media power might, in turn, depend on market concentration and journalistic field dominance, since these increase owners' capacity for state capture through such coordinated actions as media shutdowns or reputational damage campaigns to would-be reformers. It is difficult to assess dominance differentials for the respective cases of Clarín and Globo. As demonstrated above, both enjoy dominant positions in their respective contexts and have used their leverage to maintain or expand their positions.

The existence of widespread beliefs, anecdotes and myths shared especially among elites attest to their reputation. Concerning Globo, Porto reports the following statement attributed to president-elect Tancredo Neves: ‘I can fight with the Pope, with the Catholic Church, with PMDB, with anyone, but I will not fight with Doctor [sic] Roberto [Marinho]’. ⁵⁶ Anecdotes that serve as examples of shutdowns and media lynchings circulate in Brazil. One example is the case of Saturnino Braga, chair of the Parliamentary Investigative Committee in charge of investigating illegal partnerships between Globo and Time Life in the 1960s. Another is Orlando Fantazzini, a PT representative who started a campaign against violence and lowbrow television programmes. While the former was excluded from coverage by any of Globo's outlets, the latter suffered a smear campaign. ⁵⁷

In Clarín's case, phrases such as ‘no government resists more than five negative Clarín front pages’ represent a sort of ‘common sense’. ⁵⁸ As a member of the Kirchners’ inner circle stated, ‘Like every democratic president before him, Kirchner believed that treating Clarín well would guarantee reciprocal good treatment. He saw in Clarín much more than a newspaper, a TV station, a cable provider, a radio, etc.; he considered it one of the greatest power factors in Argentina, an “absolutely assembled system of news control”.’ ⁵⁹ This quotation points to possible differences with Brazil. In Argentina, the memories of the 2001 political collapse may have heightened the perception of the media's impact on government stability. This overstated perception informed Kirchner's communication style. During the 2005 mid-term campaign he declared that governments that do not exercise a ‘permanent campaign … are taken away with the helicopter’. ⁶⁰ He was referring to the image of De la Rúa's fall. According to Sivak, every government since redemocratisation had been interested in influencing the headlines, especially Clarín's. The novelty of Kirchnerismo lay in its daily, obsessive effort to ‘influence, debate and even dispute the headlines’. ⁶¹

On the opposite side, the relationships to potential allies for confrontation contrast sharply. As stated, the PT had been historically aligned with media democratisation agendas and possessed organic links and overlapping memberships with the media reform movement. While these social allies were somewhat disappointed with Lula's moderation strategy of 2002, the PT did not alienate them after it had come to power. As with social allies in general, the PT government counted on different appeasement resources to avoid disaffection. ⁶² For media democratisation activists, some material rewards existed but were somewhat irrelevant, consisting mainly of certain second-tier ministerial appointments and the increased distribution of state advertising to alternative media. Two other resources were far more important: interpersonal linkages created complicity; mutual understanding and an acceptance of the logic under which the government had to operate to survive. The president's symbolic capital and legitimacy enabled him to ask for ‘patience’, to control mobilisation, to set the pace, to appease social movements, and to buy time. Media democratisation activists mostly internalised the existing obstacles constituted by ‘the Brazilian political system’ and the power of the big media to halt reforms, while many trusted Lula's gradualism and capacity to seize opportunities. ⁶³

The Kirchners' relationship with social movements looks, in important ways, like an inverted mirror of the situation in Brazil. The Kirchners, as an ascending leftist movement with no prior ties to organised popular actors and without progressive credentials, had opposite incentives. While Lula counted on prior support, and had assumed government in a context of declining mobilisation, he was not pressed to court movements advocating radical policies. In contrast, the Kirchners were compelled by their prior isolation and the high level of social mobilisation upon their arrival to send strong signals to popular organisations to gain their support. ⁶⁴

While important, these static descriptions of different constraints in the institutional, media and social arenas do not illuminate how governmental strategies were shaped and shifted, in differing directions, after the respective crises. A different sort of account is thus required to describe how concrete event sequences were experienced and affected governmental actors' perceptions of potential allied support and opponents' power in an eventual confrontation.

A first sequential narrative focuses on events influencing evaluations of the likely availability of support in civil society to pursue counter-hegemonic strategies. The early regulatory proposals initiated by close allies of the PT government illustrate such experiences. The proposals to create the journalism council and the draft aimed at audio-visual sector regulation aroused hostile media reactions and resulted in a strong informal veto. ⁶⁵ In both cases a war over meaning broke out. Drawing on pre-existing fears about the PT, big media unanimously framed the initiatives as a threat to freedom of expression by outdated statist interventionism. The fierce application of these narratives to the regulatory attempts raised the perceived transaction costs of any proposal for media regulation. The showdown had a significant demonstration effect that made it ‘highly unlikely that anything similar [would in future] be put forward again’. ⁶⁶ There were two aspects to this demonstration effect. Not only were the episodes an indication of how the media reacted to regulatory initiatives, but they also made clear that the same social allies that promoted such initiatives had only a limited capacity to mobilise, to unite civil society around them, and to counter hostile portrayals. Indeed, these incidents exposed the fact that groups that had successfully mobilised around the FNDC in the early 1990s had faded and fissured after the Cardoso years. This perception matched wider doubts among Lula's inner circle members about the social movements' ability to sustain counter-hegemonic mobilisation during his first years in government. ⁶⁷

As outlined above, Kirchner did not initially take up the issue of media democratisation. However in 2004, sensing that the political climate was changing, formerly separate community radio organisations, media workers unions, communications scholars and NGOs started forming the Coalition for Democratic Broadcasting (CRD). Kirchner's early outreach on human rights issues was an incentive to frame media reform as a matter of human rights and democracy, and thus overcome divisions within the coalition. Additionally, despite disillusionment with governmental concessions to big media, the CRD attained the support of prestigious human rights organisations, which the government had also approached. In fact, the leaders of Madres and Abuelas de Plaza de Mayo helped members of the CRD to gradually access government offices. By the end of Kirchner's term, the coalition's leaders had met the president and obtained some vague promises. Simultaneously, some rather unnoticed government measures regarding legalisation and public recognition of non-commercial broadcasters reinforced the bridges. ⁶⁸

As the agrarian conflict erupted, these bridges eased the approach. But the government promise to sanction a new media law in the midst of the confrontation with Clarín aroused a wave of social support that transcended the CRD activists:

The Kirchners obtained support from intellectuals, personalities from culture, journalism, education, cinema, arts, literature … [who] had perceived the resurgence of a classist anti-Peronist Argentina during the crisis, and … condemned how certain media, among them the Clarín Group, had covered the conflict. … The government had won, additionally, a critical mass, a minority though relevant group, that backed its discourse against the excessive power of Clarín and agreed to curtail it. ⁶⁹

This support was clearly noticed by the government. However, mobilisation around legal reform was not a linear process. While mobilisation was intense during the agrarian conflict, a period of retreat followed the government's June 2008 defeat in parliament. Weakened, and facing mid-term elections, the Kirchners reconsidered their strategy and even contemplated reaching a settlement with Clarín. ⁷⁰ These doubts were reflected in the loss of impetus vis-à-vis legal reform, and consequently generated worry and discomfort among allies. In November 2008, the CRD sent a letter to the president reminding her to take on the commitment and seize the opportunity for media reform, reaffirming their readiness to mobilise and support her efforts. ⁷¹ While Clarín remained intransigently hostile, the government continued to be ambivalent until the elections of June 2009. It was only after these elections, where the government fared badly, that the Kirchners, (re)adopted a confrontational strategy, reminded of its potential for mobilisation. Feeling threatened after the political defeat, radicalisation appeared as a defensive tactic to retain the social support gained during the agrarian conflict, which would otherwise dissipate. In fact, the parliamentary media law debate that occurred between August and October 2009 (re)mobilised massive support through street rallies, forums, Facebook groups and a successful broadcasting show wherein public personalities explicitly revealed their alignment. ⁷²

The media law was the main component of a wider set of initiatives addressing progressive demands that helped consolidate Kirchnerism as a distinct and highly mobilised political identity. ⁷³ While between 2008 and 2009 the government had lost wider support, the reformist agenda it defensively adopted led a core of young, middle-class and intellectual sectors, heirs to the 2001 political crisis, to identify with Kirchnerism. ⁷⁴

Additionally, the mobilisation around the media law in Argentina was framed as a conflict between democracy and dictatorship, based on the fact that the existing media regulations stemmed from the widely repudiated authoritarian regime. Clarín's dealings with the military were, therefore, continually played up. In contrast, Brazil's regulatory framework antedated the 1964–85 military regime, which was also not unanimously condemned. Mobilising around Globo's past record did not, therefore, seem to offer such clear pay-offs. ⁷⁵

A second time-sensitive narrative illuminates relevant variations in the perceived consequences of the media actors' hostile behaviour for political survival. As mentioned, the Kirchners shared and overstated a general belief in Clarín's strategic importance for governability. The latter's hostile coverage during the agrarian conflict convinced them that the group's CEO had taken sides with those determined to topple the government. They ascribed Clarín responsibility for the growing urban support for landowners. This motivated a first stage of confrontation that coincided with the agrarian conflict. Yet after their defeat in Congress, the Kirchners doubted the political viability of confrontation and thus sought to re-approach Clarín. The Clarín side, however, (wrongfully) interpreted the congressional defeat as a definitive political defeat and thus ignored the government's calls, instead engaging in further discrediting the government so as to accelerate the ‘succession’ and ‘the birth of a post-Kirchnerist order’. ⁷⁶

During the 2009 mid-term elections campaign, the Kirchners found themselves heavily opposed by Clarín and most of the other mainstream media. The electoral defeat reinforced their beliefs about media power. Therefore, intransigent media opposition led them to all-out confrontation as the only choice to ensure political survival. It was only in the post-electoral context that the final mobilisation required to pass the media law, the reversal of soccer transmission rights, and some of the other radical measures would take place. The losses in popularity and the electoral defeat, which the government ascribed to media hostility, appear, from a wider perspective, to be related to the economic cycle. In 2009, the Argentine economy was still feeling the effects of the 2008 financial crisis. From 2010 on, economic recovery revamped the government's popularity beyond its core constituency.

In Brazil, the timing of the political, economic and media cycles was different. In 2005, when the political crisis first erupted, the country was expanding and leaving the 1998–2002 economic crisis behind. Social policies, mainly Bolsa Família, and minimum wages were improving the living conditions of numerous formerly excluded groups, particularly in the Nordeste region. Meanwhile, the mainstream media, caught up in a moral frenzy, fuelled and reflected a climate of outrage among the urban middle and upper classes. While Lula and especially the PT were losing favour among these strata, undetected support was growing elsewhere. While government allies initially perceived and contested a media campaign to undermine the administration, as the 2006 general elections approached some individuals close to Lula started sensing that the media frenzy did not reflect the real political situation. Not by chance, Franklin Martins was, before Lula's appointment, the first to make this perception public. In an interview given before the elections, Martins announced the presence of a new phenomenon for Brazilian politics:

We had a behaviour pattern from the end of the struggle against the dictatorship: … the middle class formed an opinion about it and that opinion extended to the periphery. Like a stone in the lake: the stone fell in the middle class, forming concentric waves in all directions. The middle class was the class of the so-called opinion leaders: you conquered them and you had solved it all. What did we see with the Mensalão crisis? The middle class formed the belief that the government was taken by a gang, by banditry, etc. … It formed that belief, threw the stone in the lake, the waves began … they hit somewhere, there was a dike, and the waves returned. Where did they hit? They hit the C class. It is the guys who earn from two to five minimum wages. … ‘Now that my life improved you want to overthrow this government?’ And so the C class also says what it has to say, that means, now we start having more than one centre of information and opinion. From now on it is the C class that forms opinion. … The process of political majority formation in the country has become much more complex than five years ago. … The newspapers directed at the A and B classes have not grown for a long time; those directed at the C class are proliferating all over the country. This indicates that many more people who in principle have different aspirations, life stories, concerns, habits, tics … are being incorporated in the market and in citizenship. … In a certain way, Lula's election meant that these people are feeling like part of the game. ⁷⁷

After the 2006 elections, this perception spread among government members, allies and PT intellectuals. ⁷⁸ In a 2008 interview, Lula himself, at the peak of his popularity, declared his indifference to the historical ideology and behaviour of Brazil's media. The biased nature of the media, he reasoned, had lost importance since audiences had become ‘intelligent’ and able to discern for themselves what went beyond fact. ⁷⁹

If after the Mensalão the perception gained ground that Globo and other big media sought to topple Lula, on the eve of the 2006 elections the government nucleus learnt that the power of these media was relative. The media reasserted their elitism and partisanship (as the old PT beliefs held), but they lost their perceived centrality for political survival. Therefore, it was reasoned that even when attacked, the best strategy might be simply to ignore or circumvent instead of engaging in energy-consuming wars that could not be won either in Congress or on the screens. To advance media change, it would be better to make gradual and more subtle moves. The actions taken during Lula's second term were consistent with this diagnosis.

Conclusion

In contemporary Latin America the realm of the media has become one of the central arenas of political struggle. The presence of powerful media conglomerates that function as strategic actors following their own corporate and ideological agendas presents important dilemmas for democratic politics. While the growing power of media organisations in the region has increased demands for democratic regulation, their perceived veto capacity and state capture has led political players to see accommodation as the only viable course of action.

I have focused on the strategic choices vis-à-vis dominant media actors of two governments that, in addition to experiencing the pressure to accommodate big media, held or established some form of commitment regarding media democratisation demands. Instead of relying on assigned prior political traditions or ideological identities (such as the populist or non-populist Left), I have established the importance of environmental factors and critical junctures as determinants of governments' strategic choices.

Drawing on the literature on the impact of institutional factors on policy preferences, I have found that significant differences in institutional configurations and the political articulation of media interests create contrasting political opportunities. In Brazil, the executive's room for manoeuvre is comparatively more constrained by the congressional correlation of forces, the electoral system, the linkages between political and media elites, the features of interest group organisation, the overall prestige of media institutions and the features of civil society. However, I have also shown that such ‘hard’ constraints, while important, do no tell the whole story. In contrast to other policy areas, the power of the strategic actors in the media sphere is mainly reputational, emphasising the subjective nature of opportunities and constraints that underpin political action. Consequently, I analysed a second layer of factors based on timing and learning processes. This analysis explored how certain junctures and sequences of events affected government perceptions, which, in turn, decisively informed the two governments' respective strategic stances.

This analysis of the perception of political opportunities and constraints in the realm of media politics contributes to the broader literature on Latin America's left turn by investigating variation in policy and elite defiance within the governments of different countries. I showed how certain learning processes operate as cognitive mechanisms that shape governments' assessments of the correlation of forces, the availability of support, and chances of political survival.

The perceived consequences of the media's hostile behaviour for political survival during the respective crises were crucial, but in the cases studied here, the causal conclusions differed and therefore the governments took opposing trajectories in response to media hostility. Government evaluations of media power and subsequent strategies vis-à-vis dominant media depended on whether hostile media coverage occurred during phases of economic expansion or economic crisis. Different overlaps decisively influenced governmental assessments of the media group's capacity to affect their own political fortunes. These varying assessments, in turn, crucially influenced subsequent governmental strategies.

In sum, while ideas and pre-held beliefs remain important, I have emphasised the centrality of institutional environments and political junctures in shaping political choices. However, as the Argentine case has made clear, under certain circumstances the pressing need to stabilise a political identity can be achieved by going public with ideological beliefs. Certain ideas that resonate among possible constituencies have the potential to create new political identities and facilitate collective action. ⁸⁰ Therefore, while prior identities matter, at certain critical junctures political actors take decisions based on short-term perceived correlations of forces that, a posteriori, define political identities.