Democratization and Intelligence and Internal Security Agencies: A Comparative Analysis of the Cases of Brazil and Portugal (1974-2014) ^*

Abstract

The development of intelligence and internal security institutions in Portugal and Brazil were directly influenced by the ways in which both countries transitioned to democracy and by the nature of their political regimes prior to the mid-1970s. This article compares these processes, highlighting how Brazilian and Portuguese intelligence and public security institutions and bodies developed over the course of the 20th century, and in particular following democratization of their political regimes. Our analysis examines the main factors explaining the arrangements adopted, and the obstacles and dilemmas encountered in the institutionalization of these services and of the new democratic regimes. The results suggest that the type of transition to democracy is the main factor explaining differences between the two cases. However, more comparative research is needed to deepen empirical and theoretical understandings and to allow for greater generalization about the relationship between changes in political regimes and reforms of coercive and intelligence institutions.

Brazil; Portugal; intelligence services; public security; democratization; process tracing

This article offers a comparative analysis of the development of intelligence and public security institutions in Portugal and Brazil, from the start of their processes of democratization in the mid-1970s¹ . Our interest in this comparison is based on a similarity that is fundamental to the analysis: regime transitions in both countries are part of three dozen similar cases that were part of the "third wave of democratizations" ( HUNTINGTON, 1993 ). The fact that they initiated regime changes at the same historical moment facilitates comparisons of the ways intelligence and internal security organs developed following democratization. The comparison also makes it possible to examine hypotheses that relate the type of the previous regime and the mode of transition as variables that account for obstacles encountered in the consolidation of new democratic regimes, and in particular dilemmas relating to the institutionalization of intelligence and public security services under conditions of legitimacy and democratic control.

We share the view, expressed by several researchers ( ANTUNES, 2002 ; BRANDÃO 2010 ; BRUNEAU, 2010 ; BRUNEAU and DOMBROSKI, 2006 ; CEPIK, 2001 , 2003 , 2005 , 2009 ; CEPIK and AMBROS, 2014 ; CEPIK and ANTUNES, 2003 ; MATEI and BRUNEAU, 2011 ; MENDONÇA, 2010 ; NUMERIANO, 2010 ; PINTO, 2010 ), that institutionalized, legitimate and effective intelligence and internal security services are essential if democratic regimes are to fulfil their functions of maintaining internal order, sovereignty and effectiveness in the formulation of public security policies. The crucial problem surrounding this question is the tension, inherent to all democratic regimes, between the coercive functions of public security organs and the civil liberties and rights of citizens, and between the needs of intelligence services for secrecy and the imperative for democratic controls of their activities ( CEPIK, 2005 , pp. 68-69). These dilemmas are particularly important for countries emerging from long periods of dictatorship, such as Portugal and Brazil.

In essence, in the newly democratized countries, intelligence and internal security organs are often heirs, or even direct continuations, of those that had operated during dictatorships ( ANDREGG and GILL, 2014 ), which were closely linked to the imperatives of political repression and the contingencies of Cold War. By contrast, intelligence services in more established democracies typically developed primarily under the strong influence of diplomacy and war ( CEPIK, 2003 ). As such, the amalgamation between public security and national security, as well as between internal and external enemies, permeated the initial steps taken towards institutionalizing intelligence services and maintaining internal order in most newly democratized countries, often with deleterious effects on civil liberties.

One of the most influential theories of path dependency in the literature on transitions to democracy ( O'DONNELL, 1988 ) argues that the more gradual and controlled these processes were, the greater the difficulties of consolidating democracy, as a result of 'authoritarian residue' under new regimes. For Share and Mainwaring (1986) , while the way in which democracy is initially established does not fully determine its future development, the initial conditions of the new regime do establish the rules of the game, modes of interaction and the limits imposed on actors and political change. Path dependency approaches have also been used, explicitly or implicitly, in valuable comparative studies looking specifically at the development of intelligence and public security services after the 'third wave' of democratization ( BRANDÃO, 2010 ; CEPIK, 2005 ; MENDONÇA, 2010 ; NUMERIANO, 2010 ; PINTO, 2010) . However, as we will argue here, it remains difficult to assess the real influence of the various explanatory variables included within such models.

The main similarities between the Portuguese and Brazilian cases are that intelligence and public security organizations had focused primarily on repressing domestic political opposition, thus making their legitimization under new democratic regimes especially problematic. However, there are a greater number of divergences between the two countries. In Portugal, during the Salazar period, there were colonial wars and a revolution, the Revolução dos Cravos (Carnation Revolution), in 1974, which produced a rupture between the old and the new regimes. This was followed by decolonization and the country's accession to the European Union after 1986. This process of regional integration was followed by reforms to internal security organizations that removed military oversight and, subsequently, were incorporated into political responses to terrorist attacks carried out in Europe in the years following the September 11, 2001 attacks in the United States. These reforms

aimed to promote greater interstate cooperation for regional security. Therefore, the institutionalization of the Portuguese intelligence services was firmly embedded in the rules of the democratic regime, subordinated to civilian coordination and parliamentary control.

By contrast, Brazil underwent a negotiated transition under the control of authoritarian leaders, which resulted in a democratic regime with a high degree of political continuity ( ANDREGG and GILL, 2014 ), significant military prerogatives and weak civilian ( CEPIK and AMBROS, 2014 ) and legislative control ( MATEI and BRUNEAU, 2011 ) of intelligence and security organizations. These characteristics, associated with incipient state capacity to guarantee citizens' rights, the stigma of political policing earned by the intelligence services in the past, the difficulty of transferring powers from the military to civilians, and a lack of clarity about intelligence activities made the legitimization and institutionalization of the country's intelligence and security systems under the new democratic regime extremely difficult, in spite of important reforms that have been carried out in recent years ( BRUNEAU, 2010 , BRUNEAU and DOMBROSKI, 2006 ; CEPIK and AMBROS, 2014 ; MATEI and BRUNEAU, 2011 ; MENDONÇA, 2010) . Moreover, according to Bruneau (2015) , the absence of external threats, combined with low state capacity, constituted the main factors explaining the low level of effectiveness of the Brazilian intelligence services.

We therefore seek to answer two fundamental questions: 01. How are these intelligence and internal security systems constituted and how do they operate? 02. What degree of democratic political control is exercised over intelligence and security organizations?

Our main hypothesis is that the core explanatory factors for the configuration and recent evolution of intelligence and security systems in the two countries are: 01. the characteristics of the dictatorship; 02. the mode of political transition; 03. the changes occurring during the consolidation of democracy; 04. the initial institutional design of these bodies; 05. the interaction between the various actors involved, and their strategic choices; and 06. civil-military relations. These variables have different weights and differ in the configuration of the current intelligence services, according to the political history of each country, as we will argue throughout the article. In order to identify these variables, we trace the processes that mark the transitions to democracy in both cases and the trajectories of institutional changes ( AMORIM NETO and RODRIGUEZ, 2016 ; COLLIER, 2011 ).

This work deals mainly with national-level external intelligence and internal security services (increasingly focused on police and law enforcement activities, 'new threats' and counter-terrorism), and, to a lesser extent, military intelligence, regional and international cooperation structures, and state public security departments which command their own judicial and military police, though the state level will not be prioritized in the analysis. Military defence policies will only be addressed where they focus on internal security. The sources of data used in the research come from official documents, press coverage and the academic literature.

The article is structured in two main parts: the first, divided into two sections, maps processes of the development of intelligence and internal security organizations, from their origins to their consolidation under Portugal's and Brazil's respective dictatorships. The second part, also with two sections, addresses current configurations of intelligence systems and organizations that developed during the process of democratization in both countries, up to the end of the 2000s. Finally, we offer some final considerations on the themes discussed, from a comparative perspective.

The evolution of the intelligence and security services in Portugal

The first organizations fitting the conceptual definition of these services were created during the period of the Estado Novo Salazarista. The Portuguese Estado Novo (New State) emerged with the military coup of 1926, and the insurrections of 1927, 1928 and 1931, which saw the rise of the then Finance Minister, Antonio Salazar, to power, allowing him to institutionalize the dictatorship and transform it into the Estado Novo ( CEREZALES, 2008 , 2010 ; PINTO, 2000 ; NUMERIANO, 2010 ). In 1933, Salazar took command of the Council of Ministers and, although there was a president, General Carmona, it was Salazar himself who led the regime until 1968.

The political police in Portugal during Salazarism

The beginning of the Salazar period saw the 'production' of the state's repressive apparatus, through its reorganization and expansion ( CEREZALES, 2008 , 2010 ). The police forces, the Polícia de Segurança Pública (Public Security Police, PSP) and the Guarda Nacional Republicana (Republican National Guard, GNR), were created, reinforcing their loyalty to and functions for the Estado Novo. The Polícia de Vigilância e Defesa do Estado (State Surveillance and Defence Police, PVDE) was also established in August 1933. It would later, in the context of the Second World War and the reorganization of the state's repressive apparatus to combat communism, become the notorious Polícia Internacional de Defesa do Estado (International Police for Defence of the State, PIDE) ( GOUCHA, 2011) . During the government of Marcelo Caetano (1969-1974), PIDE was transformed into the Direção-Geral de Segurança (Directorate–General of Security, DGS), via a process of institutional reorganization designed to defend the totalitarian regime ( PIMENTEL, 2009) .

Following the Spanish Civil War, the 'red threat' moved to the center of the PVDE's concerns. As Pimentel (2009) points out, elsewhere, after the end of World War II, fascist and national-socialist political police forces were dismantled; in Portugal, however, with the survival of the dictatorial regime, the political police acquired new powers and a new name. Decree-Law no. 35.046, dated 10/22/1945, created the PIDE to provide legal support for renewed repression, now under its command.

During the Colonial War (1960-1962), sub-delegations and border and surveillance posts were created in Angola and Mozambique, placing the PIDE "in relation to overseas territories, as the armed forces were" ( PIMENTEL, 2009 , p. 40). In 1962, the number of PIDE employees overseas exceeded the number on the Mainland and Islands. Furthermore, Pimentel (2009) and Numeriano (2010) emphasize the PIDE's role in monitoring left- and right-leaning members of the military, within the regime. The institution's activities were wide-ranging, extending reached all spheres of power. Therefore, despite its civilian activity in the metropole, the PIDE at this stage was also active alongside the armed forces overseas. According to Cerezales (2008) and Pimentel (2009) , the PIDE did not become militarized, as the Republican National Guard (GNR) and the Public Security Police (PSP) had in its actions in the Portuguese colonies. However, as Numeriano (2010) points out, the PIDE, and later the DGS, continued as a military intelligence body even after 1974. As such, we can infer a partial militarization of the PIDE/DGS due to the breadth of its role in political policing, as a military intelligence agency in the provinces, and in the monitoring of military personnel under the Salazar regime. This body then, clearly had a broad scope, spanning across all spheres of power.

Marcelo Caetano, who replaced Salazar in 1968 due to the old dictator's deteriorating health, dismantled the PIDE and created the Direção-Geral de Segurança (Directorate General of Security, DGS) in November 1969. The new institution continued its practices under the tutelage of the Interior Ministry, with the exception of its activities in the colonies, which came under the command of the Ministry for Overseas Territories. Despite rhetoric of political 'opening', there was a hardening of the regime. This led to growing protest and, in response, intensified repression, now led by of the DGS. In this context, discussions began that would lead to the coup of 25 April 1974, putting an end to the Salazar regime and beginning Portugal's transition to democracy. During the first six decades of the twentieth century, the Portuguese dictatorship had created a series of institutions of political intelligence and repression (the PVDE, PIDE and DGS). After its downfall, these became the main targets of the leaders of the new political regime seeking to achieve effective transitional justice ( PINTO, 2006) .

The 25th of April 1974 and the transition to democracy in Portugal

The military coup of April 25, 1974 (hereafter the Carnation Revolution), which ended the Salazar regime, initiated a troubled period of transition to democracy, which only became institutionalized with the constitutional revision of 1982. According to Lobo, Magalhães and Pinto (2009) , the context that preceded the Carnation Revolution can be characterized by a lack of international support for the regime in the context of the Cold War, by an atmosphere of crisis in the state and by the intensification of decolonization struggles in Portugal's African colonies. The Portuguese case stands out for the singular role of the military, whose coup led the country towards democracy and decolonization. Marcello Caetano's inability to end the Colonial War is recognized as one of the key factors leading to 25 April.

Between 1974 and 1976, a dynamic transition period unfolded, characterized by differences and conflicts between the leaders and groups that had supported the coup, mainly between the Movimento das Forças Armadas (Armed Forces Movement, MFA) and some conservative generals who resisted the end of Portugal's colonial empire. This factor opened up a space for political and social mobilization, and a concomitant crisis of the state, which may explain the inability of moderate elites to dominate the subsequent rapid institutionalization of representative democracy ( PINTO, 2006 , p. 39).

The transition via rupture with the Salazarist regime became especially visible with the so-called 'cleansing' of the public administration and companies, a purge of sixty generals from the armed forces and also the dismantling of the PIDE/DGS ( PINTO, 2010 ). The adaptation of the apparatuses of policing and repression from the dictatorial period to the new context of the transition was turbulent ( CEREZALES, 2008, 2010) , 2010 ). It was the MFA’s intention to expand the purges of the police forces through the creation of the Comando Operativo do Continente (Operational Command of the Continent, COPCON), an organization that expressed the revolutionaries' distrust of the police forces. The rupture, however, occurred more abruptly and perceptibly with organizations linked to political policing, ie. PIDE/DGS, whose members that had been involved in illegal acts were expelled from the government during the early stages of the transition. With regard to intelligence information or services in the period prior to the constitutional revision of 1982, two key bodies should be highlighted, the Serviço Diretor e Coordenador de Informação (Information Management and Coordination Service, SDCI), linked to the Conselho da Revolução (Revolutionary Council) and the Divisão de Informações (Information Division, DINFO), linked to the Estado-Maior-General das Forças Armadas (General Staff of the Armed Forces, EMGFA).

In 1982, a revised constitution, which closed down Revolutionary Council, introduced the new Law of National Defense and the Armed Forces (Law nº 29/82) and initiated a reorganization of the intelligence system. This process finally led to the subordination of the military to party-political power ( LOBO, MAGALHÃES and PINTO, 2009 ). In order to distinguish it from the old political policing system, the Sistema de Informações da República Portuguesa (Information System of the Portuguese Republic, SIRP) was also created in 1984 through Law nº 30/84. Included within the SIRP were the: Conselho Superior de Informações (High Council of Information, CSI), Serviço de Informações de Segurança (Security Information Service, SIS), Serviço de Informações Estratégicas de Defesa (Strategic Information for Defense Service, SIED) and the Serviço de Informações Militares (Military Information Service, SMI). During the first ten years following the coup, the civilian side of the intelligence community remained suspended in an institutional limbo. Roberto Numeriano (2010) states that in the period between 1974 and 1984 'fear' and 'acephaly' dominated Portugal's intelligence services. This was due to the legacy of abuses committed by the political police during the Estado Novo period ( PINTO, 2012 ).

It was only in 1984, with the creation of the SIRP, that new intelligence structures within the democratic Portuguese State truly began to emerge. This was directly linked to two terrorist incidents on Portuguese territory in the early eighties, which showed the importance of intelligence organizations to the country's new political elites ( PINTO, 2012 , p. 164). In the 1990s, another legislative amendment modified the SIRP Framework Law (Organic Law Nº 04/2004), and, in 2004, SIED was created ( GOUVEIA, 2013) . Thus, during the period of the transition to democracy, between 1975 and 1982, constitutional changes and reforms substantially altered the internal and external security structures of the Portuguese State. We emphasize, therefore, that new political forces promoted a rupture with former institutional arrangements, which will emerge as the central point of difference when compared with the Brazilian case. This finding supports the hypothesis that the kind of transition to democracy is fundamental for understanding the dilemmas and constraints surrounding the institutionalization of intelligence services under the future democratic regime.

The security and intelligence services in Brazil until redemocratization

In embryonic form, institutionalized intelligence activity in Brazil began in 1927 with the creation of the Conselho de Defesa Nacional (National Defence Council, CDN) – the first federal body to be established with the sole purpose of processing intelligence for the Presidency of the Republic ( FIGUEIREDO, 2005 ). As the Old Republic (1889-1930) came to an end, the country witnessed an intensification of political rivalries and the strengthening of workers' movements and opposition to the ruling elites. When Getúlio Vargas assumed the Presidency of the Republic with the Revolution of 1930, he reorganized the National Defence Council and created Sections of National Defence, for each ministry.

With the establishment of the Estado Novo dictatorship (1937-1945), intelligence and policing services became clearly political, geared towards repressing opponents and enemies of the regime. The political police par excellence, which led the repression, as well as the 'Estado Novo's' propaganda machine and censorship system, was the Departamento de Investigações e Propaganda (Department of Investigation and Propaganda), the infamous DIP.

The institutionalization of Brazilian intelligence effectively began after World War II. The Departamento Federal de Segurança Pública (Federal Department of Public Security, DFSP) was created under the Ministry of Justice and Internal Affairs in 1944, the embryo of would become the current Federal Police ( ROCHA, 2004 , p. 66). Intelligence activity directly linked to the state came about in 1946, with the creation of the Serviço Federal de Informações e Contrainformações (Federal Information and Counter-Information Service, SFICI), by president-elect, General Eurico Dutra, who had been Minister of War during the 'Estado Novo' and succeeded Vargas. The SFICI was born under the National Security Council (the new name for the National Defence Council) and was headed by senior officers of the Armed Forces. The militarization of civilian intelligence services in Brazil has already been noted.

In fact, the Brazilian intelligence activity acquired unprecedented proportions following the coup of March 31, 1964. Just over two months after the overthrow of then-President João Goulart (1961-1964), the Serviço Nacional de Informações (National Information Service, SNI), on June 13 of that year. The main justification for the creation of the body that would become the symbol and coordination center for the 'intelligence community' of Brazil's authoritarian regime, which survived for five years even after the dictatorship ended, was that the new government needed to establish an intelligence service that was aligned with the Doutrina de Segurança Nacional (National Security Doctrine, DSN), idealized within the framework of the Escola Superior de Guerra (Higher School of War, ESG), since 1949 ( FICO, 2001, p , pp. 39-42).

Initially, SNI was led by General Golbery de Couto e Silva, until General Costa e Silva assumed the presidency in 1967 and appointed General Emílio Médici as its new head. With the demise of the SFICI and the appearance of the SNI, the DFSP itself almost disappeared. However, the body was nationalized on November 16, 1964 ( ROCHA, 2004 ), with the creation of the Polícia Federal (Federal Police, PF). The protagonism of the military in public security organs was guaranteed by the implementation of an integrated model: municipal-based police, such as civil guards, were dissolved, whereas all state Polícias Militares (Military Police, PMs) were unified under the command of state governors.

In 1968, when armed leftist groups began to operate and state repression intensified, the government proclaimed Institutional Act Nº 05 (AI-5) on December 13, 1968. AI-5 was the regime's most repressive legislation; the 'years of lead' (1968-1976) of the dictatorship began. Against this backdrop, in 1970, SNI became part of a larger structure of intelligence activities, the Sistema Nacional de Inteligência (National Intelligence System, SISNI). Formally coordinated by SNI, SISNI also included the intelligence centers of the military ministries, the Divisões de Segurança e Informações (Security and Information Divisions, ISD) of the civil ministries, and the Assessorias de Segurança e Informações (Security and Information Offices, ASI) located in state enterprises and public institutions (universities, etc). As for the military ministries, the navy had the Centro de Informações da Marinha (Navy Information Center, CENIMAR), founded in the 1950s; the army, the Centro de Informações do Exército (Army Information Center, CIE), the strongest and most active of the military intelligence organs, created in 1967 specifically to combat armed left-wing organizations; and as similar information structure within the airforce, with the Centro de Informações de Segurança da Aeronáutica (Airforce Center for Information and Security, CISA), established in the late 1960s.

The SNI leadership gained increasing influence over the decisions of the presidency, to the point that two of the five Generals who headed it – Emílio Médici and João Figueiredo – became Presidents of the Republic.

The strengthening of the National Security Council in 1968 and the establishment of the Sistema de Segurança Interna (Internal Security System, SISSEGIN), accompanying the military regime's escalating repression, sought to institutionalize these repressive functions. They included SISSEGIN, the Destacamentos de Operações de Informações (Detachments of Information Operations) (DOI) of the Centros de Operações de Defesa Interna (Centers of Internal Defence Operations, CODI), bodies formed after 1970, organized around military and regions and commanded by army officers specifically for purposes of political repression and combating the armed left. The DOI-CODIs were composed of officers and sub-officers of the three armed forces as well as members of the military police forces and state judiciaries. The CIE also had regional offices with intense anti-subversive activities.

It is worth noting that the functions of political policing and fighting the armed opposition were also carried out by Departamentos Estaduais de Ordem Política e Social (State Departments of Political and Social Order, DEOPS), organs of the public security ministries of Brazilian states. Here, the state of São Paulo in particular stood out for the significance and brutality of its repressiveness. These bodies comprised the regime's 'security community', which was responsible for nearly four hundred political deaths and disappearances and thousands of people tortured, mostly in the first half of the 1970s.

Political liberalization and divisions within the 'security community'

It is important to emphasize a number of aspects that distinguish Brazil's dictatorship from similar regimes in the region. It lasted the longest, was the most successful economically and was less repressive than its counterparts. The Brazilian democratization process is also distinguished by the fact that it developed through negotiations that remained under the strong control of authoritarian leaders.

The long and extremely gradual process of political liberalization, as well as the use of electoral competition as a key institutional resource in the transition process, 'normalized' the control and unilateral imposition of rules by the existing rulers within the Brazilian political world ( ARTURI, 2001 ). Former authoritarian leaders were assured that there would be no 'revenge' against state agents who had committed crimes in the repression of the opposition, another pressing issue in the consolidation of the new democratic regimes over the last two decades.

From the beginning of the 'slow, gradual and secure' process of political liberalization, led by General Ernesto Geisel shortly after he assumed the presidency of the Republic in 1974, the SNI paradoxically expanded its activities, enjoying great autonomy during the late 1970s. The SNI frequently clashed with the DOI-CODIs and the DEOPS, defending the political opening of the Geisel and Figueiredo presidencies against the 'hardline' position within the military, embedded in the latter bodies, which opposed the regime's liberalization. This conflict led to the dramatic dismissal of Minister of the Army General Sylvio Frota by Geisel on October 12, 1977.

General João Figueiredo took office on March 15, 1979, but gained power without relying on the regime's most repressive legislation (AI-5), which had been abolished by Geisel in December 1978. His accession coincided with the outbreak of demonstrations by students and workers calling for redemocratization. Opposed to President Figueiredo's continuation of of political opening, the far right, a large part of which was based in the agencies of intelligence and repression, especially in the DOI-CODIs and the CIE regional offices, carried out a series of attacks following the amnesty of 1979 and the return of exiles to the country. But the ambitions of these extreme groups within the regime were dashed by the Riocentro incident in Rio de Janeiro on April 30, 1981. Indeed, the fatal bombing of the car of a DOI-CODI sergeant in the parking lot of a convention center where thousands of people were attending a concert, had important consequences for the country's political future ( ARTURI, 2000 ; ANTUNES, 2002) . The episode put an end to the attacks that had aimed to interrupt the transition process.

The results of the elections for state governor and state and federal deputies in 1982 gave a significant political victory to the opposition and were directly responsible for the regime losing control of the transition process. It was in this context that the mobilization for direct presidential elections surprisingly erupted in 1984, known as the 'Diretas-já' (Direct elections now) movement, the strongest and most concrete demand for democratization since 1964. The movement for direct elections for the Presidency of the Republic mobilized millions of people across the country, but failed to prevent the rejection of the draft constitutional amendment that would have implemented it in April 1984.

Eventually, the democratic opposition, which had succeeded in provoking dissidence within the regime's political and military forces, indirectly elected the moderate opposition leader Tancredo Neves to the presidency via electoral college in January 1985. The subsequent events are well known: the election of the Tancredo Neves and José Sarney ticket in January 1985; Tancredo's hospitalization, after falling seriously ill on the eve of his inauguration on March 15; and his death on April 21. Thus fate and key decisions ensured that Jose Sarney, former president of the PDS, the party of the dictatorship, became Brazil's first civilian president since 1964.

The Sarney presidency and the National Information Service (1985-1990)

The stage of genuine democratisation developed under José Sarney's presidency, known as the 'New Republic'. In May 1985, the ban on communist parties was lifted, the franchise was expanded to its full extent, including illiterates. A National Constituent Assembly, elected in 1986, was convened to draft and approve the new democratic constitution in 1988, and direct elections to the Presidency of the Republic were held in 1989.

With regard to the intelligence services, the SNI continued to operate during José Sarney's government, a situation that was unprecedented during the 'third wave of democratizations'. The SNI continued to focus on opposition groups, producing monthly reports on the country's internal security and monitoring strikes which were multiplying during the period ( CEPIK and ANTUNES, 2003 ). This situation ensured that a military ethos – which continues even today – remained hegemonic within the civil intelligence activities, maintaining practices and legacies that had been instituted during the dictatorship ( NUMERIANO, 2010) .

The current configuration of Portugal's intelligence system

The current configuration of the Sistema de Informações da República Portuguesa (Information System of the Portuguese Republic, SIRP) is consistent the fourth and fifth periods of institutional evolution of these services as identified by Gouveia (2013 , p. 70). Since 2004, the SIS and SIED have moved closer, falling under the same supervisory body, the Secretary-General of the SIRP, and the SIRP has undergone administrative reforms with Law Nº 09 of February 2007. This law is considered to be the Organic Law of the SIRP, which brought together different legislation and organized the different units according to the Framework-Law of the SIRP ( GOUVEIA, 2013 , p. 71).

In general, the main policing structures in Portugal are linked directly to the Ministry of Internal Administration, while the Judicial Police, given its remit of conducting criminal investigations, is linked to the Ministry of Justice. It should also be noted that the Security Information System (SIS) reports directly to the Prime Minister, who is also linked to the Secretary-General of the SIRP. Control over intelligence activities is external and parliamentary. Thus, control over intelligence activities in Portugal has, since 2004, become highly institutionalized. This is thanks to a clear break with the legacy of authoritarianism, and in particular the Political Police, presenting a consistent pattern of demilitarization, with the separation of civil and military intelligence functions ( GOUCHA, 2011 ).

On August 29, 2008, the Internal Security Law was passed, outlining the country’s new Internal Security Policy, Sistema de Segurança Interna (Internal Security System, SSI), the security forces and services, and police measures. The new law aimed to allow greater coordination and cooperation between the bodies responsible for internal security. The central body of the new Internal Security System is the Conselho Superior de Segurança Interna (Higher Council for Internal Security, SSC), which is composed of representatives from the National Assembly, the Secretary-General of the SIRP, the Chief of Staff of the Armed Forces, the Representative of the Integrated System for Protection Operations and the Director-General of Prison Services. In this way the Secretary-General of the Internal Security System (SGSSI) was established, reporting directly to the Prime Minister, who in turn must relay the system's main activities to the President.

These changes within the SSI are related to regional integration and the new demands of regional security organization and cooperation within the European Union. The fight against terrorism is central in this respect, since police cooperation between the Portuguese security forces and the European Police Service (EUROPOL), which was established in 2010, plays a key role in regional security. In addition to police cooperation with this European agency, there have been increased demands for information exchange and border control, especially through SSI's collaboration with both FRONTEX and EUROSUR, concerning the control of border crossings within in the European Union. Following the new Internal Security Law, new organic laws reforming the main internal security forces, the Republican National Guard (GNR) and the Public Security Police (PSP) were also passed.

In the Portuguese case then, as provided for in the SIRP Framework-Law of 2004 and implemented by the Organic Law of 2007, control over intelligence activities executed by SIS or SIED is the responsibility of both: 01. the Secretary-General of the SIRP, who refers directly to the Prime Minister, who in turn links with the Higher Council for Information, which is under the supervision of the Attorney General of the Portuguese Republic; and 02. by the SIRP Supervisory Board, which reports to the National Assembly ( GOUCHA, 2011 , pp. 44-46).

Finally, it is worth highlighting that the reforms and changes to Portugal's internal security respond to various issues and embody an adaptation to the post-11 September context of strengthening regional integration. It should also be noted that military control has weakened as democratic consolidation has advanced. The institutionalization of these services and their legitimacy among the population have also grown, with the creation of more effective control structures to regulate the security and intelligence forces. As we shall see below, the structure and functioning of the current Brazilian intelligence system is still a long way off from the level of democratic institutionalization achieved in Portugal.

The trajectory of Brazil's intelligence system after democratization

The Serviço Nacional de Informações (National Information Service, SNI) was officially disbanded on March 15, 1990. This fact, along with the congress' general neglect of intelligence, security and defence issues, and the political stigma attached to the sector among civil society ( ANTUNES, 2002 ), led to the absence of any strategic and civil intelligence service in Brazil for a decade. Military institutions, by contrast, remained intact and regulated only by themselves. In fact, the true successor institution of the SNI was the Agência Brasileira de Inteligência (Brazilian Intelligence Agency, ABIN), only created in 1999.

The structure of the Presidency of the Republic was once again modified by president Itamar Franco (1992-1994), following the impeachment of Collor. In a way, with the disbanding of the SNI and other measures, Collor had dismantled the military's guardianship of the presidency, although he did not touch the military intelligence services. In November 1992, the Department for Strategic Affairs was reorganised and incorporated into the newly created Subsecretaria de Inteligência (Sub-Department for Intelligence, SSI). According to Brandão (2010) , the Itamar Franco administration saw backsliding in civilian-military relations, with the reintegration of military personnel removed by Collor into the SSI, and a lack of oversight of the sector. The appointment of an Admiral from the reserve forces to Chief Minister of the Secretaria de Assuntos Estratégicos (Ministry for Strategic Affairs, SAE) again placed civilian intelligence under the command of a member of the armed forces.

During the two mandates of President Fernando Henrique Cardoso (1995-2002), there was an concerted effort to redefine civil-military relations through the creation of a Ministry of Defence, the drafting of a national Política de Defesa Nacional (National Defense Policy, PDN) and the establishment of the ABIN ( BRANDÃO, 2010 , p. 143). The creation of the ABIN was delayed due to political resistance, eventually coming into being with the passing of Law Nº 9,883 on December 07, 1999. This law also instituted the Sistema Brasileiro de Inteligência (Brazilian Intelligence System, SISBIN), with responsibility to 'integrate the planning and execution of the intelligence activities in the country'. The ABIN was designated the 'central body of the SISBIN, with a mission to 'plan, execute, coordinate, supervise and control intelligence activities in the country', directly linked to the President of the Republic and, preferably, under civilian leadership.

Two significant and quite positive aspects of the Law deserve attention: 01. that regulating the participation of the Legislative Power in the analysis of the National Intelligence Policy, established by the President of the Republic; and 02. that which decreed congress' responsibility for supervising the activities of the ABIN. Also during 1999, important changes to the intelligence sector led to the watering down of some advances made in the legislation that had created SISBIN, such as the creation of the Gabinete de Segurança Institucional (Office for Institutional Security, GSI) of the Presidency of the Republic in September, the responsibilities of which were linked to the defunct Casa Militar (Military Head Office).

The GSI gained increased powers during the presidencies of both Fernando Henrique Cardoso and Luís Inácio Lula da Silva (2003-2010), formally subordinating the ABIN, although the latter's Director-General still required congressional approval. As the GSI holder is legally required to be a high-ranking officer nominated by the President of the Republic, this institutional arrangement ensured the formal and legal militarization of the agency, leaving ABIN without direct contact with the Presidency of the Republic.

This situation was altered by President Dilma Rousseff, who, for a few months in early 2016, detached the ABIN from the GSI and subordinated it to the Secretary-General of the Presidency of the Republic. However, a few days after her suspension during her impeachment process, on May 12, 2016, interim president, Michel Temer, elevated the head of the GSI to the rank of chief minister, again subordinating the ABIN. Another issue identified by Priscila Brandão is that the law does not cover the intelligence and counterintelligence activities of state police, major commanders or even the Federal Police, which are "of fundamental importance for the functioning of our institutional foundations" (BRANDÃO, 2010, p. 178).

Subsequently, the sector has undergone important structural changes, some positive for the consolidation of democracy and guaranteeing democratic control over intelligence activities, others not. The GSI, for example, "has become the main aggregator of intelligence flows from various federal agencies and the locus of crisis management in the areas of internal and external security" ( CEPIK, 2005 , p. 84). The creation of the Comissão Mista de Controle das Atividades de Inteligência (Joint Commission for the Control of Intelligence Activities, CCAI), in 2001, implemented the already-planned External Control and Oversight Body. In 2002, the Ministry of Defence created the Subsistema de Inteligência de Defesa (Defence Intelligence Subsystem, SINDE) with the purpose, alongside the Ministry of Defense, of articulating the intelligence centers of the Navy, Army, Aeronautics and Defense General Staff. Finally, the last change we highlight here is the creation of the Subsistema de Inteligência de Segurança Pública (Public Security Intelligence Subsystem, SISP), on December 21, 2000. However, Cepik (2005) points out that, until 2005, "its intended integrator of information flows in the areas of criminal intelligence, security intelligence (internal), counterintelligence and counterterrorism was poorly developed" ( CEPIK, 2005, p , p. 89).

The expansion of the SISBIN should have increased cooperation between agencies. However, problems of overlapping remits and weak levels of specialization persist, hindering the integration of information flows. The SISP is coordinated by the Secretaria Nacional de Segurança Pública (Department for National Public Security, SENASP) of the Ministry of Justice, which includes: the Department of Federal Police (DPF) and the Department of Federal Highway Police (DPRF), within the Ministry of Justice; the Conselho de Controle de Atividades Financeiras (Council for the Control of Financial Activities, COAF), Coordenação Geral de Pesquisa e Investigação (General Coordination Center for Research and Investigation, COPEI) and Secretaria da Receita Federal (Department of Federal Revenue, SRF), from the Ministry of Finance; as well as the Ministry of Regional Integration, Ministério da Defesa (Ministry of Defence, SPEAI), Gabinete de Segurança Institucional da Presidência da República (Institutional Security Office of the Presidency of the Republic, ABIN and SENAD) and the civil and military police forces of the 26 states and the Federal District. This incomplete list of affiliated organizations gives an idea of the scope of this subsystem that, in practice, is only partially integrated with the SISBIN ( CEPIK, 2009 ).

This is the period in which the agencies of the Brazilian 'intelligence community' begin to substitute the term 'information', used during the dictatorship, for 'intelligence', possibly as a way removing the stigma of the past, as well as modernizing the sector's nomenclature ( ANTUNES, 2002 ). The air force was the first to begin a reorganization process, creating its Secretaria de Inteligência (Department of Intelligence, SECINT) in 1991, at the same time that the navy created the Centro de Inteligência da Marinha (Navy Intelligence Center, CIM), replacing CENIMAR. Subsequently, in 2004, SECINT became the Centro de Inteligência da Aeronáutica (Air Force Intelligence Center, CIAer), its current denomination. In 1992, CIE became the Army Intelligence Center (retaining its original acronym). Later, the Escola de Inteligência Militar do Exército (Military Intelligence School of the Army) was created, with the purpose of training new analysts ( MENDONÇA, 2010) .

Finally, the Ministry of Defence was created on June 9, 1999, and the Ministries of the Navy, Army and Air Force transformed into Estados-Maiores (Commands); the greatest repercussion of this reform to ​​military intelligence was the subordination of the commanders to the Ministry of Defence. The Ministry of Defence also has an Intelligence Sub-Office, which is responsible for proposing military intelligence doctrine, as well as coordinating the Departamento de Inteligência Estratégica (Department of Strategic Intelligence, DIE). The Ministry of Defence, in addition to rationalizing costs and better integrating the separate armed forces, meant an important step towards more democratic civil-military relations, and, until recently, reduced military presence at the center of political power.

The presidency of Luís Inácio Lula da Silva (2003-2010) left an ambiguous legacy with regard to the institutionalization of intelligence and internal security organizations and activities. During his first term, Lula altered practically nothing in the sector, although, through Law 10.683 of 05 May, 2003, he formally brought the ABIN under the coordination of the GSI, commanded by a general. Likewise, he did not make changes to military intelligence, on the contrary, he gradually increased the intelligence assignments of the Ministry of Defense ( MENDONÇA, 2010 ). It was only in July 2007, with the inauguration of Minister Nelson Jobim, former president of the STF and one of the main drafters of the 1988 Constitution, that the Ministry of Defence began to gain political weight, influencing defence matters and subordinating the respective commands ( PASSOS, 2014 ). During the Lula presidency, the National Defence Policy was elaborated in June 2005 and the National Defence Strategy in December 2008, which contributed to the gradual democratic institutionalization of the sector. In April 2018, under the Temer government, a reserve army general became Minister of Defence; the first member of the armed forces to occupy the role, a worrying sign for the maintenance of civilian political control over ​​defence.

These important legal modifications to civil-military relations in Brazil were not accompanied by similar measures in the area of ​​intelligence. On the contrary, the scope of military intelligence has been expanded, while civilian intelligence, whose highest body is ABIN, which legally coordinates SISBIN, was weakened after its involvement in a political scandal, Operation Satiagraha in 2008, which involved the illegal use of its agents in actions by the Federal Police.

Mendonça (2010) , like Priscila Brandão ( ANTUNES, 2002 ; BRANDÃO, 2010 ), uses the notion of path-dependency ( COLLIER and COLLIER, 1991 ) – though, in his case, only at the organizational and institutional level – to demonstrate how the 'Satiagraha case' was a critical moment for the ABIN/SISBIN. The researcher, who is also an ABIN intelligence officer, draws on Wolfgang Krieger's (2009) argument that the most powerful form of control of intelligence activities is legislative control and its individual commissions. However, according to the author, the legislature is generally ineffective and even negligent, until a political scandal or a serious 'intelligence' failure erupts. At that moment, the legislature reacts, either by establishing commissions of inquiry or by forming commissions for reforming of the intelligence system, as happened after the both Satiagraha case in Brazil and the September 11, 2001 attacks in the United States, cases that Mendonça (2010) compares in his work.

The main problem with the institutionalization of civilian intelligence is not, therefore, the stigma that ABIN faces due to its being the heir of SNI ( ANTUNES, 2002 ; BRANDÃO, 2010 ) or its internal militarization – which, according to Mendonça (2010) , is declining – but rather that it is being prevented from becoming the leading civilian institution for the coordination of state intelligence. In this regard, the military seeks to expand its 'territory' and in turn diplomats choose to align themselves with either the military or ABIN agents, at their own discretion.

Besides the problems surrounding the institutionalization of the SISBIN under civilian and democratic control, and continuing military prerogatives resulting from the transition agreement, the ineffectiveness of the democratic state for a large part of the population, evidenced by widespread violence and the failure of public security, have become, in our view, the most important factor impeding the deepening of the democracy in Brazil. Such ineffectiveness and violence primarily affect the poorest. State agencies, however, are reluctant to respond equitably and recognize the rights of the poorest and social minorities. This is particularly the case for public security agencies, who persist in the use of arbitrary and violent methods against these sectors of the population.

A sign of the times, in 2010, was when the army and navy, along with the Rio de Janeiro state police and the Federal Police, participated in a major joint police operation in a favela in Rio de Janeiro, showing that militarization is also affecting public security ( PASSOS, 2018 ). This phenomenon was reinforced by the growth of Garantia da Lei e da Ordem (Law and Order Guarantee, GLO) operations by the military and its central role in coordinating the security of the 2014 Football World Cup and the 2016 Olympic Games.

In the Table 01 below we synthesize the main explanatory variables revealed in our process mapping of the two cases, allowing for both their comparison and differentiation. In particular, it should be noted that both the type of regime change and the duration of the transition affect the relationship between the old and new institutions and affect the form of interaction between the sectors and civil-military relations.

Table 01
Comparative framework with explanatory variables

	Brazil	Portugal
Previous regime	Authoritarianism	Authoritarianism
Type of regime change	Negotiated	Rupture
Previous institutions	Preserved	Modified
New institutions	Product of previous institutions	Little legacy of previous institutions
Duration of transition	Long	Short-to-Medium
Interaction between sectors	Low	High
Civil-Military relations	Low	High

• Source: Elaborated by the authors.

Final considerations

The main objective of the study has been to compare the institutional arrangements of the intelligence and internal security services in Portugal and Brazil, allowing a better understanding of the development of these institutions and their current challenges, based on democratization processes in both countries. In contrast to the Brazil, Portugal has become a kind of exemplary control case for the democratization of intelligence services and the consolidation of democracy in recent years ( NUMERIANO, 2010 ). As our mapping of these processes shows, path dependency is very evident and the key explanatory variable accounting for the democratic institutionalization of these organizations is the type of transition rather than the characteristics of the previous regime, ie. in the "nature of the rupture" ( PINTO, 2010 , p. 407). Indeed, we believe that where transition occurs via a rupture with the previous regime, the variable 'type of transition' is decisive, since the new leaders consciously seek to break with the institutions and practices of the past. As a result, the previous regime lacks importance when new political and coercive institutions are being designed. In those countries where dictatorships ended in collapse, as in Argentina ( ESTÉVEZ, 2014 ), or by coup, as in Portugal, new democratic regimes face fewer constraints and authoritarian legacies in their transitions. This allows them to carry out deep political and institutional ruptures in relation to the previous regime, especially with respect to the coercive and intelligence agencies. In these types of transition, stronger ruptures accelerate the demilitarization process, and therefore also the consolidation of democracy, which depends upon the transfer of control of intelligence from military to civilian authorities ( BRUNEAU, 2010) . Transitions of this kind are also more likely to avoid problems of confusion about the role of intelligence under democracy, a problem identified by Bruneau and Dombroski (2006) . In the Portuguese case, this had specific characteristics, especially in relation to the ways in which the bodies supporting the Salazarist regime (PIDE/DGS) were demilitarized and disbanded, with transitional justice playing a decisive role ( PINTO, 2010) .

Even so, following the Revolution of April 1974 the new Portuguese intelligence system took almost a decade to emerge, preceded by the end of military tutelage and the beginning of democratic consolidation with the constitutional reform of 1982. A similar phenomenon occurred in Brazil, between the disbanding of the SNI of the dictatorship at the beginning of the Collor presidency in 1990 and the actual creation of the ABIN at the end of 1999, at the beginning of Fernando Henrique Cardoso's second term. This coincidence may have several causes, but it certainly demonstrates the difficulty of new democracies – even those that break definitively with the previous regime, as in Portugal – to overcome the stigma that became associated with the intelligence services during the periods of political repression.

The various authors who have worked on this topic have proposed diverse explanatory variables, relating democratization processes and new intelligence and security institutions, to explain the constraints to civilian and democratic control over these organizations following democratization. For those who, like Zaverucha (2000 , 2005 ) and Numeriano (2010) , focus their analysis on civil-military relations, the most important factor is the demilitarization of the external and internal intelligence agencies. They argue this is a necessary condition for the consolidation of the democratic regimes, something they argue has not yet been achieved in Brazil. Cepik (2005 , 2009 ), Brandão (2010) and Mendonça (2010) , meanwhile, are more attentive to institutional factors, interests and interaction among the main actors involved. However, there are nuances between their interpretations. Priscila Brandão (2010) attributes the constraints to the institutionalization of the civilian intelligence sector in Brazil to the legacy of the transition and choices made by key actors during 'critical moments' of these processes. In this interpretation, also based on a path-dependency model, the type of the previous regime has less explanatory value than the type of transition, a process that left authoritarian legacies (militarization, stigma of intelligence agencies, political and social neglect of the sector, etc) that in her view still constrain the legitimate institutionalization of intelligence services and civilian political dominance over the military. According to Brandão (2010) , the current structure of the intelligence services still favours military over civilian interests. She concludes that the form of transition influences the pattern of civil-military relations under the new democratic regime more than the reform of intelligence institutions. Thus, for Brandão (2010) , civilian and legislative interests become "a more relevant variable than institutional constraints in the process of refounding national civilian intelligence agencies in Brazil, Chile and Argentina" ( BRANDÃO, 2010, p , p. 259).

Cepik's argument (2005) is different from those that propose strong links between the type of transition and its effects on the institutionalization of intelligence and public security services, although he recognizes the repercussions of those processes in the current constraints faced by the sector. Cepik (2005) , in a paper on Brazilian intelligence agencies, based on a model developed by Peter Gill (1994) , concludes that in 2004, "Brazilian structural reforms in the area of ​​intelligence were largely successful from the point of view of their adaptation to the context of a consolidated democratic regime" ( CEPIK, 2005, p , p. 97). However, according to the author, "the professionalization of intelligence services in Brazil still has a long way to go with regard to the armed forces and the police" ( CEPIK, 2005, p , p. 96).

Mendonça (2010) , meanwhile, is the only author who does not include processes of democratization as a variable influencing the current configuration of intelligence agencies. The current situation of the sector, according to the author, is characterized by a "hypertrophy of military intelligence" ( MENDONÇA, 2010 , p. 77) and by a deep inter-agency conflict between the military, diplomats and civilians in the ABIN, in which the first two actors seek to prevent the latter agency from effectively coordinating the SISBIN.

The Brazilian intelligence system has evolved through advances and retreats in civilian coordination and democratic control, and is now at another critical moment in its institutionalization. With regard to path dependency models based on democratization processes, we can say that in the case of negotiated and gradual transitions like the Brazilian one, it becomes very difficult to identify the importance of variables relating to the type of previous regime, transition and dynamics of democratic consolidation, given the high degree of political and institutional continuity between these phases. There is a need for more empirical and comparative studies to further develop theoretical approaches to the subject.

Brazil's intelligence system, unlike Portugal's, still faces the challenge of institutionalization at a moment of conflict and political uncertainty. Deep reform of the structure and the institutions of the country's public security (federal and state), as well as the institutional development of the organizations of intelligence and order under civilian coordination and political control – a fundamental aspect for democratic consolidation, as pointed out by Bruneau (2010) – are fundamental priorities for the sector today. Recently, there have been worrying signs of regression in civil-military relations and threats to civilian political control over coercive state agencies, and even the democratic regime itself. These include: military intervention in security in Rio de Janeiro in early 2018, followed by the appointment of former Defence Minister, Raul Jungmann, to the new Ministry of Public Security, the appointment as of a reserve military officer, General Silva e Luna Minister, as Minister of Defence – the first time since the ministry was created in 1999 – and finally the political pressure by army commander General Villas-Bôas publicly exerted on the Federal Supreme Court on the eve of a vote on a request for habeas corpus by former President Lula in April 2018. In summary, in Brazil, unlike Portugal, which advanced its democratic controls with Organic Law Nº 04 of 2004, constraints to the consolidation of the democratic regime persist and have worsened in recent years, in particular those concerning military prerogatives and civilian control over state agencies of intelligence and coercion.

References

• AMORIM NETO, Octavio and RODRIGUEZ, Júlio César Cossio (2016), O novo método histórico-comparativo e seus aportes à ciência política e à administração pública. Revista de Administração Pública . Vol. 50, Nº 06, pp. 1003-1027.
• ANDREGG, Michael M. and GILL, Peter (2014), Comparing the democratization of intelligence. Intelligence and National Security . Vol. 29, Nº 04, pp. 487-497.
• ARTURI, Carlos S. (2001), O debate teórico sobre mudança de regime político: o caso brasileiro. Revista de Sociologia e Política . Nº 17, pp. 11-31.
• ARTURI, Carlos S. (2000), Le Brésil: une tentative de démocratisation octroyée (1974-1985) . Villeneuve d'Ascq: Presses Universitaires du Septentrion. 502 pp..
• ANTUNES, Priscila Carlos Brandão (2002), SNI & ABIN: uma leitura da atuação dos serviços ao longo do século XX . Rio de Janeiro: Editora FGV. 220 pp..
• BRANDÃO, Priscila Carlos (2010), Serviços secretos e democracia no Cone Sul: premissas para uma convivência legítima, eficiente e profissional. Niterói: Ímpetos. 302 pp..
• BRUNEAU, Thomas C. (2015), Intelligence reform in Brazil: a long, drawn out process. International Journal of Intelligence and Counter Intelligence . Vol. 28, Nº 03, pp. 502-519.
• BRUNEAU, Thomas C. (2010), Controlling intelligence in new democracies. International Journal of Intelligence and Counterintelligence . Vol. 14, Nº 03, pp. 323-341.
• BRUNEAU, Thomas C. and DOMBROSKI, Kenneth R. (2006), Reforming intelligence: the challenge of control in new democracies. In: Who guards the guardians and how: democratic civil–military relations. Edited by BRUNEAU, Thomas C. and TOLLEFSON, Scott D.. Austin: University of Texas Press. pp. 145-177.
• CEPIK, Marco Aurélio Chaves (ed) (2011), Inteligência governamental: contextos nacionais e desafios contemporâneos. Niterói: Ímpetus. 317 pp..
• CEPIK, Marco Aurélio Chaves (2009), Prólogo. In: Democratización de la función de inteligencia: el nexo de la cultura nacional y la inteligencia estratégica. Edited by SWENSON, Russel G. and LEMOZY, Susana C.. Washington: National Defense Intelligence College. pp. V-IX.
• CEPIK, Marco Aurélio Chaves (2005), Regime político e sistema de inteligência no Brasil: legitimidade e efetividade como desafios institucionais. Dados . Vol. 48, Nº 01, pp. 67-113.
• CEPIK, Marco Aurélio Chaves (2003), Sistemas nacionais de inteligência: origens, lógica de expansão e configuração atual. Dados . Vol. 46, Nº 01, pp. 75-128.
• CEPIK, Marco Aurélio Chaves (2001), Espionagem e democracia: agilidade e transparência como dilemas na institucionalização de serviços de inteligência. Rio de Janeiro: Editora FGV. 232 pp..
• CEPIK, Marco Aurélio Chaves and AMBROS, Christiano (2014), Intelligence, crisis, and democracy: institutional punctuations in Brazil, Colombia, South Africa, and India. Intelligence and National Security . Vol. 29, Nº 04, pp. 523-551.
• CEPIK, Marco Aurélio Chaves and ANTUNES, Priscila Carlos Brandão (2003), The new Brazilian intelligence system: an institutional assessment. International Journal of Intelligence and Counter Intelligence . Vol. 16, Nº 03, pp. 167-194.
• CEREZALES, Diego Palacios (2010), Repressive legacies and the democratisation of Iberian police systems. South European Society and Politics . Vol. 15, Nº 03, pp. 429-448.
• CEREZALES, Diego Palacios (2008), Estado, régimen y orden público en el Portugal contemporáneo (1834–2000). Doctoral Thesis . Complutense University of Madrid. Madrid.
• COLLIER, David (2011), Understanding process tracing. PS: Political Science & Politics . Vol. 44, Nº 04, pp. 823-830.
• COLLIER, David and COLLIER, Ruth Berins (1991), Shaping the political arena . Princeton: Princeton University Press. 877 pp..
• ESTÉVES, Eduardo E. (2014), Comparing intelligence democratization in Latin America: Argentina, Peru, and Ecuador Cases. Intelligence and National Security . Vol. 29, Nº 04, pp. 552-580.
• FICO, Carlos (2001), Como eles agiam . Rio de Janeiro: Record. 269 pp..
• FIGUEIREDO, Lucas (2005), Ministério do silêncio: a história do serviço secreto brasileiro de Washington Luís a Lula (1927-2005). Rio de Janeiro: Editora Record. 590 pp..
• GILL, Peter (1994), Policing politics: security intelligence and the liberal democratic state. London: Frank Cass. 384 pp..
• GOUCHA, Roberto Manuel Henriqueto (2011), O sistema de informações em Portugal. Master's dissertation . Instituto Universitário de Lisboa.
• GOUVEIA, Jorge Bacelar (2013), Os serviços de informações em Portugal: organização e fiscalização. Revista de Direito e Segurança . Nº01, pp. 63-85.
• HUNTINGTON, Samuel P. (1993), The third wave: democratization in the late twentieth century. Norman: University of Oklahoma Press. 366 pp..
• KRIEGER, Wolfang (2009), Oversight of intelligence: a comparative approach. In: National intelligence system: current research and future prospects. Edited by TREVERTON, Gregory F. and AGRELL, Wilheim Cambridge. Cambridge: Cambridge University Press. pp. 210-234.
• LOBO, Marina Costa; PINTO, Antonio Costa and MAGALHÃES, Pedro C. (2009), As instituições políticas da democracia portuguesa. In: Portugal: uma democracia em construção Ensaios de homenagem a David B. Goldey. Edited by CABRAL, Manuel Villaverde; LOBO, Marina Costa, and FEIJÓ, Rui Graça. Lisboa: Imprensa de Ciências Sociais. pp. 141-174.
• MATEI, Florina Cristiana and BRUNEAU, Thomas (2011), Intelligence reform in new democracies: factors supporting or arresting progress. Democratization. Vol. 18, Nº 03, pp. 602-630.
• MENDONÇA, Ariel M. de (2010), A proposta de reformulação do Sistema Brasileiro de Inteligência (SISBIN): similaridades e diferenças com o processo de reforma na Comunidade de Inteligência dos Estados Unidos da América. Dissertação de Conclusão do Curso de Altos Estudos de Política e Estratégia da Escola Superior de Guerra.
• NUMERIANO, Roberto (2010), Serviços secretos: a sobrevivência dos legados autoritários. Recife: Editora da UFPE. 395 pp..
• O'DONNELL, Guillermo (1988), Transições, continuidades e alguns paradoxos. In: A democracia no Brasil: dilemas e perspectivas. Edited by REIS, Fábio Wanderley e O'DONNEL, Guilhermo. São Paulo: Vértice. pp. 41-71.
• PASSOS, Anaís (2018), The military mystique: democracies and the war on crime in Brazil and Mexico. Doctoral Thesis . Department of Political Science. Institut d’Etudes Politiques de Paris.
• PASSOS, Anaís (2014), As reformas institucionais do Ministério da Defesa (2007-2011). Marter's dissertation . Programa de Pós-Graduação em Ciência Política. Universidade Federal do Rio Grande do Sul.
• PIMENTEL, Irene Flunser (2009), A História da PIDE . Lisboa: Círculo de Leitores e Temas e Debates. 608 pp..
• PINTO, António Costa (2010), Coping with the double legacy of authoritarianism and Revolution in Portuguese democracy'. South European Society and Politics . Vol. 15, Nº 03, pp. 395-412.
• PINTO, António Costa (2006), O Legado autoritário e a transição portuguesa para a democracia, 1974-2004. In: Portugal 30 Anos de Democracia (1974 – 2004) . Edited by LOFF, Manuel e PEREIRA, Maria da Conceição Meireles. Porto: Editora da Universidade do Porto. pp. 37-72.
• PINTO, António Costa (ed) (2000), Portugal contemporâneo . Madri: Edições Seguitur. 320 pp..
• PINTO, Maria do Céu (2012), Portugal's intelligence evolution in the Post-9/11 World. International Journal of Intelligence and Counter Intelligence . Vol. 25, Nº 01, pp. 160-177.
• ROCHA, Bruno L. (2004), Polícia Federal após a Constituição de 1988: polícia de governo, segurança de Estado e polícia judiciária. Master's dissertattion . Instituto de Filosofia e Ciências Humanas. Programa de Pós-Graduação em Ciência Política. Universidade Federal do Rio Grande do Sul.
• SHARE, Donald and MAINWARING, Scott (1986), Transições pela transação: a democratização no Brasil e na Espanha. Dados . Vol. 29, Nº 02, pp. 207-236.
• ZAVERUCHA, Jorge (2005), FHC, forças armadas e polícia: entre o autoritarismo e a democracia (1999-2002). Rio de Janeiro: Record. 285 pp..
• ZAVERUCHA, Jorge (2000), Frágil democracia: Collor, Itamar, FHC e os militares (1990-1998). Rio de Janeiro: Civilização Brasileira. 333 pp..

• 1
  This article is partially based on the chapter 'The intelligence and internal security services in Brazil and Portugal', published by the authors in 2011 in a collection edited by Marco Cepik (2011) . The current article addresses a more recent period and approaches the comparison, the control of processes and the verification of the hypothesis with greater rigor.
• *
  The authors would like to thank the three anonymous referees for the comments, and, for their valuable contributions and assistance, the Professor Igor Castellano da Silva and the research assistants of the Study Group on State Capacity, Security and Defense at the Universidade Federal de Santa Maria: Camille Costenaro, Juliany Santos Bernardino, Mariana Garghetti Buss and Taís Chiquitelli.

Publication Dates

• Publication in this collection
  17 Oct 2019
• Date of issue
  2019

History

• Received
  08 June 2018
• Accepted
  10 Feb 2019