The communication of Courts of Accounts and Prosecution Services on social media: the challenges of accountability in the digital democracy

Abstract

This article deals with the communication of Brazilian Courts of Accounts and Prosecution Services on social media platforms Facebook, Twitter, and Instagram, since the opening of their accounts. The study innovates in the discussion about democratic controls over public administration in the so-called “digital democracy,” focusing on institutional communication of Courts of Accounts and Prosecution Services, showing that controllers also need to be concerned about their transparency and accountability. Having verified that they maintain professional communication departments that operate intensively on social media, with emphasis on the Federal Court of Accounts (TCU) and the Federal Prosecution Service (MPF) on Twitter, the article specifically analyzes the content of the two institutions’ communication in this platform by using the Iramuteq software. Both TCU and MPF kept posting a variety of content on Twitter since first opening their accounts and at important moments in recent political history. The research question addressed was: Does expanding communication channels necessarily mean greater accountability? The study found that these institutions, when communicating with society on social media, did not necessarily respond to the public accountability requirements that democratic theory implies.

Keywords:
communication of public institutions; democratic theory; control; transparency; content analysis

Resumo

Este artigo trata da comunicação que os Tribunais de Contas (TCs) e os Ministérios Públicos (MPs) brasileiros promovem nas redes sociais Facebook, Twitter e Instagram, desde a abertura de suas contas. Inova na discussão a respeito do controle democrático sobre a administração pública com foco na chamada “democracia digital”, mostrando que os controladores também precisam se preocupar com a transparência e responsabilização nesses espaços. Tendo verificado que os TCs e os MPs mantêm departamentos profissionais de comunicação que operam intensamente nas redes sociais, com destaque para o TC da União (TCU) e o MP Federal (MPF) no Twitter, este trabalho analisa, de forma específica, o conteúdo da comunicação das duas instituições nessa rede, com a ajuda do software Iramuteq. Tanto o TCU como o MPF mantiveram intensa atividade de postagem de conteúdos diversos no Twitter, desde a abertura de suas contas e em momentos importantes da história política recente. A pergunta que orienta o trabalho é: ampliar os canais de comunicação significa necessariamente maior accountability? O que se constatou é que, ao se apresentar mais à sociedade pelos meios digitais, esses órgãos de controle não necessariamente responderam aos requisitos de responsabilização pública presentes na teoria democrática.

Palavras-chave:
comunicação de instituições públicas; teoria democrática; controle; transparência; análise de conteúdo

Resumen

Este artículo trata de la comunicación que los Tribunales de Cuentas (TC) y los Ministerios Públicos (MP) brasileños promueven en las redes sociales Facebook, Twitter e Instagram, desde la apertura de sus cuentas. Innova en la discusión acerca de los controles democráticos sobre la administración pública en la llamada “democracia digital”, mostrando que los controladores también deben preocuparse por su transparencia y rendición de cuentas. Habiendo comprobado que mantienen departamentos de comunicación profesional que operan intensamente en las redes sociales, con énfasis en el Tribunal de Cuentas de la Unión (TCU) y el Ministerio Público Federal (MPF) en Twitter, el artículo analiza específicamente el contenido de la comunicación entre las dos instituciones en esta red con la ayuda del software Iramuteq. Tanto el TCU como el MPF mantuvieron sus intensas actividades de publicación de contenidos varios en Twitter desde la apertura de sus cuentas y en momentos importantes de la historia política reciente. La pregunta que guía el trabajo es: ¿expandir los canales de comunicación significa necesariamente una mayor accountability? Lo que se encontró es que, al presentarse más a la sociedad a través de los medios digitales, estos organismos de control no respondían necesariamente a los requerimientos de rendición pública de cuentas presentes en la teoría democrática.

Palabras clave:
comunicación de instituciones públicas; teoría democrática; control; transparencia; análisis de contenido

1. INTRODUCTION

The Federal Constitution of 1988 (CF-88) established that the external control of the Brazilian public administration is to be exercised by the Court of Accounts (Tribunal de Contas, or TC) and by the Prosecution Service (Ministério Público, or MP). The TCs have the role of assisting the Legislative Branch in its oversight function, and the MPs have an even broader role of defending diffuse and collective rights, including administrative probity.

Both institutions enjoy formal autonomy from the government so as to control the correct application of public money, provide transparency to administrative activities, and also empower civil society, since both of them may take action at the request of citizens, thus enhancing the democratic control of public administration (Arantes, 2002; Arantes, Loureiro, Couto, & Teixeira, 2010; Kerche, 2009; Olivieri, Loureiro, Teixeira, & Abrucio, 2015). In other words, these institutions were designed to make sure public agents that are part of the direct and indirect administration are held accountable, ensuring the visibility of all government and administrative activities. Thus, TCs and MPs, at different levels of the federation, are instruments for achieving democracy as the government of visible power (Bobbio, 2000, p. 87), making democratic accountability effective.¹

However, the literature on recent changes in the accountability network of the Brazilian public administration has pointed out that the members of these institutions - especially the MP, with its criminal attributions - have acquired a prominent political role, directly influencing the political and democratic dynamics of the country (Arantes & Moreira, 2019; Avritzer & Marona, 2017; Kerche & Marona, 2018; Silva, 2020).

In addition, and more importantly for the present work, the studies also draw attention to a classic problem of contemporary democratic theory, concerning who controls the controllers: while designed to exert accountability toward other institutions, they are hardly subject to any formal external control, except in the appointment of their top officials and with regard to their budgets. In fact, as we shall try and establish in this article, TCs and MPs have used social media rather intensively to communicate in direct contact with other institutions and with citizens, especially in episodes of great political controversy, such as the Car Wash operation.

Thus, we propose to examine here, for the first time, the communication that TCs and MPs have been promoting on social media, analyzing how it is structured on Facebook, Twitter, and Instagram, since their respective accounts were launched. In this sense, the article innovates in the discussion about the democratic control over public administration in the so-called “digital democracy” (Gomes, 2011, 2016), focusing on the institutional communication of Brazilian TCs and MPs, not in their role as agents of control, but as a way of finding out whether such bodies, by conveying more information to the public, effectively become more accountable. In essence, the question driving this investigation is: does the expansion of communication channels necessarily result in greater accountability?

Having established that the Brazilian TCs and MPs maintain professional communication departments that are heavily active on social networks, with special emphasis on the Twitter presence of the Court of Accounts of the Union (Tribunal de Contas da União, or TCU) and the Federal Prosecution Service (Ministério Público Federal, or MPF), we also made a specific analysis of the content of both institutions on that network, relying on the Iramuteq software (Ratinaud & Marchand, 2012). Both the TCU and the MPF have been constantly posting on Twitter since their accounts were initially set up and at important junctures in recent political history, regarding this vehicle as a means of communication to ensure their accountability to the general public.

In what follows, the text is divided in four sections: first, we outline the theoretical aspects on which the analysis is based. After describing the methodology applied in the study, we present the data concerning the presence of TCs and MPs in social networks. In the following section, we analyze in detail the content of the TCU’s and MPF’s posts on Twitter. Finally, we discuss the main inputs of this study to the field under scrutiny and we offer some concluding remarks.

2. THEORETICAL ASPECTS OF THE ANALYSIS

The discussion on public communication using the internet can be classified into two major analytical axes. On one side, we have the literature that examines the growing use of internet tools by the public administration to provide public services and promote transparency in the virtual space (Amaral, 2016; Cunha & Miranda, 2013; Pinho, 2008, 2011; Pinho, Iglesias, & Souza, 2005; Raupp & Pinho, 2013). On the other, the debate is framed by a broader approach to the transformations produced by technological advances in the way the State operates and in its relations with society, particularly with regard to the impact of new digital communication tools on political representation and democratic participation (Castells, 1999; Esteves, 2011; Miguel & Biroli, 2010; Weber 2011).

In normative terms, public communication - by making transparent the acts of the State branches and the institutions linked to them - is a condition for democratic accountability, which is basically effected through institutional mechanisms that seek to ensure accountability and continuous oversight of both elected representatives and the top public bureaucracy with decision-making authority (Mainwaring & Welma 2003; O’Donnell 1998; Pelizzo & Stapenhurst 2014). In other words, communication by State institutions ensures that the actions of government leaders are visible to citizens (Rocha, 2012; Weber, 2017), because the sovereign people must have the ability to exercise control over the rulers and, in order to do so, it is necessary to know and have access to the acts of public management, not merely exercising control at the electoral stage, but following the decision-making process of elaboration and implementation of public policies (Abrucio & Loureiro, 2005; Arantes et al., 2010).

The concept of accountability is central to modern democratic theory (Bobbio, 1997; Dahl, 1997; Gruber, 1986; Held, 2006). It implies that rulers and public office holders have a duty to be accountable to the public based on democratic criteria that define the scope and purpose of each institution (Abrucio & Loureiro, 2005). Thus, what should be answered to society is not simply defined by the incumbents of public offices, but by a broader process that involves the creation of legal frameworks by politicians elected by the people, the interplay of mutual controls between the branches of government - to avoid the concentration of decision making in one of them - and the definition of republican standards to avert the exercise of power in a personalistic way or on behalf of any vested interest group (O’Donnell, 1998).

Following this argumentative line of democratic theory, Judith Gruber (1986) analyzed how State power, especially its unelected components, should respond to society. In the author’s view, the control of the rulers by the ruled is circumscribed by democratic-republican rules, to prevent the values and interests of bureaucrats from overriding accountability, turning into a self-defense of those who are in office (Gruber, 1986). This is the classic question posed by Max Weber: how to ensure that bureaucrats are controlled politically, and not according to their designs? (Abrucio & Loureiro, 2018; Weber, 1993).

In this sense, not all communication made by public institutions and/or by bureaucrats is an effective instrument of accountability. It can only be so if it complies with the terms of the democratic rules set forth in the legislation, based on the idea that there can be no power without control by others and by society, as well as on a republican dynamic of putting forward institutional information. This differentiation has often been overlooked by the literature on public administration and political communication, and the goal of this article is to analyze whether the communicative profusion of the Brazilian MPs and TCs indicates a greater predominance of democratic accountability rather than a way of presenting information to society that serves primarily the interests and values of the members of these institutions themselves, or that fails to address the most important aspects for the public control of the controllers

The logic of transparency advanced by the open government doctrine, for example, may produce a fallacy with regard to accountability. By open government we mean that public processes become transparent, and the citizen is encouraged to interact with public administration through new forms of participation (Schmidthuber, Ingrams, & Hilgers, 2021). However, the idea that open government promotes the strengthening of social control of State power is not clear-cut, among other factors, because the positive effects of openness seem to depend on how democratic accountability is carried out (Matheus & Janssen, 2019; Schmidthuber et al., 2021).

Furthermore, it has been pointed out that analyses on transparency policies need to employ more appropriate strategies for assessing the effects of data openness, such as counterfactual reasoning (Michener, 2019), i.e., if accountability effects are to be discussed, it is necessary to have mechanisms to gauge the impact of data openness.

Among the current range of possibilities for institutional communication required for democratic accountability, there are those that operate within social media, such as Twitter, Facebook, and Instagram, as part of the “digital democracy” process (Gomes, 2011, 2016), which are often viewed as capable of “amplifying the State official discourse as well as transparency and accountability procedures” (Weber, 2017, p. 47). However, the communication conveyed on social media may occasionally be characterized either as promoting corporate interests or neglecting the issues legally defined as those that the public should be informed about in relation to both the government (Fontes & Marques, 2022) and, in the case of the object of this study, the organs of oversight.

Thus, although the focus of the analysis revolves around the originator of the communication-in this case, two State institutions, namely the Brazilian TCs and MPs-this does not mean that the analysis is unrelated to a broader discussion on public communication and political culture. We recognize the importance of communication initiatives that seek to empower society, strengthen the struggles of citizens for collectively shaping the interest of the State and society (Gomes, 2011), among which we could mention regulatory reforms such as the Access to Information Law (Lei nº 12.527, de 18 de novembro de 2011).

Nevertheless, even though modifications in formal rules stipulate that access to information and transparency of administrative acts must be guaranteed, including through the internet, in practice things are still substantially different (Gomes, 2005). Obstacles are often created to comply with the legislation, which is why it is necessary to verify if this is also replicated “[...] in computer-mediated interactions between public authorities, politicians and those governed” (Amaral, 2016, p. 4).

In view of these points, the approach adopted here is justified because it is informed by the results of previous research on the use of social media. As Goldbeck, Grimes, and Rogers (2010) point out, social networks can help increase public communication and transparency. Among them, Twitter stands out as the one with the greatest potential to disseminate information and the most widely used by public administration and professional politicians as a form of citizen contact (Hong, 2013; Vergeer, Hermans, & Sams, 2011).

The problem is that, in the same way that the use of Twitter by professional politicians reflects a high volume of activity when compared to the other social networks analyzed, it paradoxically does not always translate into increased transparency or accountability for the elected office or for the public administration. On the contrary, according to a study by those authors in relation to U.S. congressmen, it does not necessarily lead to accountability if professional politicians maintain an account on the platform with high traffic and interaction (Goldbeck et al., 2010, p. 15). In this regard, the authors’ distinction between transparency and outreach is all the more significant.

The notion of transparency applied to cyber communication is related to free access to information for the general public, made available on the global computer network to ensure accountability, while outreach is the diffusion or dissemination of information and ideas through the same network to a specific or general audience, but that is geared toward self-promotion (Goldbeck et al., 2010). The distinction is important because, in outreach situations, government or politicians’ communication may be “telling people only what they want them to know, and not necessarily what they need to know in order for them to understand the actions and consequences of the activities of public officials and elected representatives” (Amaral, 2016, p. 46).

In a study on the institutional communication of the Public Defender’s Office (DP) of the State of Rio Grande do Sul, which encompassed the use of its social networks, Carnielli (2017) found that the topics focus on citizen’s rights, but nevertheless tend to channel a discourse that is contradictorily self-referential or self-centered. In other words, communication revolves around a project of institutionalization on behalf of the DP itself and the corporate interests of its members. And it is precisely an institution designed for the defense of collective and individual rights, especially of people exposed to a situation of vulnerability and hyposufficiency, that disregards the citizen in its public communication, while advertising its own activities and those related to its management, with an approach similar to what is observed with regard to outreach.

In light of these considerations, this paper aims to show how institutions formally designed to perform accountability, such as the Brazilian TCs and MPs, communicate on social media. The controllers have a central role in democracy and are financed by public funds, thus their actions must comply with the democratic principle according to which public institutions - including the top level of bureaucratic decision-making - must be subject to the control of the governed.

Therefore, it is important to know how TCs and MPs are currently positioned in social media and what kind of communication they produce, as well as whether such channels help broaden or, on the contrary, reduce public discussion on issues related to transparency and accountability of these institutions (Carnielli, 2017).

3. METHODOLOGY

The research is qualitative in nature. It maps the institutions’ social media presence in terms of communication structure and content, encompassing 33 portals of the Courts of Accounts of Brazil’s federal units - the Federal Court of Accounts (TCU), the five Municipal Courts (TCMs), and the 27 State Courts (TCEs) - as well as the portals of 28 Prosecution Services (MPs) - the Federal Prosecution Service (MPF) and the Prosecution Service of the Federal District (MPDF), which are part of the Prosecution Service of the Union (MPU), and the 26 Prosecutions Services (MPs) at the state level, organized in each of the 26 states of the federation. The survey was implemented in two stages.

The first stage of data collection was done by means of guided navigation in the official portals of the TCs and MPs. These visits took place between December 1^st, 2020 and February 1^st, 2021, a period in which none of the portals were under maintenance. In each portal, we searched directly for data from the communication department on the following topics: the employee’s position (whether permanent or commissioned); the employee’s expertise; whether or not the expertise was related to professional communication; the institution’s presence on Twitter, Facebook, and Instagram; the year in which the institution joined the channel; the number of followers and posts in the channel.

The second stage used content analysis with the help of the Iramuteq software (Ratinaud & Marchand, 2012), following the criteria indicated by Sampaio and Lycarião (2021) regarding the reliability, validity and replicability of the study that makes use of this type of methodology. Thus, the content analysis is related to the Twitter posts of the TCU and the MPF since the year of entry, in 2009, until February 2020 (unit of analysis). This material is publicly available to users of that social network. It is worth saying that, once this interval was defined, the data were collected automatically through Twitter’s public Application Programming Interface (API), using collector scripts in R. They were then saved in Excel spreadsheets with profile, location, date and time of the posts, as well as the text content of the tweets and retweets (RT).

Guided by the research questions stated in the previous section, the analysis was methodologically based on the reduction and structuring procedures (Mayring, 2004; Viegas & Borali, in press). This analysis was divided in two phases². In the first one, reduction was necessary because the software requires that the corpus to be inserted follow a specific pattern, especially when faced with corpuses composed of multiple texts, as is the case here. In order to ensure greater processing speed and optimize computational resources, a preliminary trimming of the corpus is crucial, which ends up reducing the material for analysis, but it preserves the essential contents in a condensed form (Mayring, 2004; Viegas & Borali, Forthcoming).

As indicated by Viegas and Xavier (2021), a script in Python language was used for data “cleanup” (which consists in the removal of special characters, punctuation, links, codes and skipped lines), to reduce the essential content. At this stage, after the Excel spreadsheet with tweets and metadata was loaded, it was possible to convert it into a dataframe for manipulation, on which the data cleanup was performed. Once this was done, the scritp exported the “clean” (reduced) texts into a plain text (.txt) file, using the format required by Iramuteq. Thus, it was possible to load the corpus into Iramuteq in a faster and more practical way, reducing processing time, memory usage, and other potential bugs and glitches of the software. Then, a new reduction involved the removal of pronouns, conjunctions, articles, and other words considered stopwords.

In the second phase, in line with Viegas and Xavier (2021), once the data had been reduced to its essential content, the following tests based on Reinert’s method (1990) were used: word clouds, based on word frequency calculation; similarity analysis, which provides a graphical visualization (clusters) of interconnection; and content relationship among the most used words. At this stage, after the tests applied, structuring made it possible to accurately analyze definitions and typical text passages (“key examples”).

Structuring is aimed at filtering out particular aspects of the data and making a cross-section of the material on the basis of criteria established in advance or in order to appraise the material according to specific criteria (Mayring, 2004; Viegas & Borali, in press). Based on the results of the two tests applied (word cloud and similarity analysis) and on the theoretical aspects of the analysis, typical text passages (“key examples”) were extracted from the database (reduced unit of analysis). Once these passages were extracted, conventional content analysis was performed (Sampaio & Lycarião, 2021) with a view to establishing rules that yielded a coding guide, which rendered the task of structuring and analyzing the content more precise, as described, and discussed in the final two sections of this article.

4. THE BRAZILIAN COURTS OF ACCOUNTS AND PROSECUTION SERVICES ON SOCIAL MEDIA

All Brazilian TCs and MPs have in their organizational charts a department devoted to institutional communication. These departments are in charge of official publications, media management, and publicity and/or journalism, which is hardly a novelty given the importance of public communication for State institutions (Weber, 2017). Nevertheless, there is still insufficient information available about the activity. The data collected in this survey offer some key insights. According to Table 1, all the Prosecution Services in Brazil and 97% of the Courts of Accounts have a specific department for communication, with hired professionals, and virtually all of them have official pages on the three mentioned platforms.

TABLE 1
FREQUENCY OF COMMUNICATION CHANNELS FOR TCS AND MPS IN 2021

COMMUNICATION ProxIEs	Prosecution Service		Court of Accounts
NUMber of Institutions	28	-	33	-
Has a communication department (CD)	28	100%	33	100%
Has a communications professional	28	100%	32	97%
Has an official page on Facebook	27	99%	30	90%
Has an official page on Twitter	28	100%	33	100%
Has an official page on Instagram	28	100%	31	93%

• Source: Survey data.

Moreover, activity in the online environment has become one of the main vehicles for public communication by TCs and MPs, not only on their official webpages, but especially on social media. The three networks Facebook, Twitter and Instagram are used by more than 90% of the TCs and MPs (Table 1), with an emphasis on Twitter, mirroring for the Brazilian case what has already been mentioned in the international literature.

In the case of TCs, their respective accession to social networks took place in a period spanning from 2009 to 2020 for Instagram and Facebook. All TCs have a presence in at least one social network, the most frequent being Twitter, followed by Instagram and Facebook. Similar to the TCs, for the MPs the most used network is Twitter, followed by Instagram and Facebook. Unlike the TCs, however, the MPs have opened accounts on social networks in the same period, at the beginning of the 2010s. It is worth noting that most TCs and MPs opened their Twitter account between 2009 and 2010, which may indicate that the initiative was not a mere coincidence.

Among the TCs, only the TCU has a significant number of followers on social networks. Still, if we add all the followers of all the TCs, the result is equivalent to a quarter, or 25%, of the MPs’ followers. Among the MPs, we highlight the MPF, the institution responsible for oversight of the federal public administration. On its own, the MPF concentrates more than half of the Twitter followers of all Brazilian MPs. In aggregate (sum of the three social networks), this institution concentrates approximately 44% of all the MPs’ followers in the country (Table 2).

TABLE 2
NUMBER OF FOLLOWERS OF TCS AND MPS ON TWITTER, FACEBOOK, AND INSTAGRAM IN 2021*

Organs	Court of Accounts				Prosecution Service
	Twitter	Facebook	Instagram	Total TC	Twitter	Facebook	Instagram	Total MP
State	69,824	181,830	97,989	349,643	475,874	642,771	511,665	1,630,310
AC	821	2,430	1,050	4,301	2,710	8,682	12,600	23,992
AL	162	2,239	5,092	7,493	6,161	36,882	16,500	59,543
AM	3,697	10,751	11,700	26,148	1,649	1,662	4,152	7,463
AP	128	1,485	1,181	2,794	11,400	3,386	4,593	19,379
BA	257	956	1,619	2,832	5,923	47,239	42,800	95.962
CE	3,604	13,874	6,930	24,408	1,657	20,644	43,400	65,701
DF	2,613	2,488	-	5,101	13,000	15,697	15,500	44,197
ES	236	12,189	7,517	19,942	2,490	15,471	1,816	19,777
GO	4,323	6,051	5,278	15,652	172	20,955	35,500	56,627
MA	106	311	1,870	2,287	1,405	4,804	32,500	38,709
MG	4,256	14,256	1,411	19,923	16,600	48,697	33,500	98,797
MS	357	41,714	1,398	43,469	997	11,076	7,763	19,836
MT	3,621	-	7,888	11,509	528	13,561	11,900	25,989
PA	545	3,177	2,834	6,556	16,100	12,225	15,700	44,025
PB	5,836	5,550	5,449	16,835	7,290	7,953	13,300	28,543
PE	2,779	5,371	1,442	9,592	15,100	20,915	30,000	66,015
PI	2,405	3,963	4,929	11,297	778	4,759	17,700	23.,237
PR	2,973	12,964	2,146	18,083	2,330	19,369	22,500	44,199
RJ	218	-	3,273	3,491	146,500	67,395	29,000	242,895
RN	2,570	1,505	1,908	5,983	12,300	1,112	12,900	25,200
RO	405	4,604	2,689	7,698	3,056	10,471	5,747	19,274
RR	2,947	829	1,952	5.728	7,062	1,345	1,694	8,756
RS	8,573	11,091	2,119	21,783	60,900	72,116	13,200	146,216
SE	2,838	-	5,122	7,960	40,700	45,274	37,200	123,174
SC	3,603	3,229	4,788	11,620	54,600	5,679	15,900	76,179
SP	4,781	17,689	2,847	25,317	35,400	123,593	24,200	183,193
TO	5,170	3,114	3,557	11,841	9,066	4,266	10,100	23,432
Municipal	3,891	10,906	11,556	26,353	-	-	-	-
TCMBA	1,057	818	-	1,875	-	-	-	-
TCMGO	1,075	1,382	4,482	6,939	-	-	-	-
TCMPA	1,146	4,871	3,812	9,829	-	-	-	-
TCMRJ	215	929	508	1,652	-	-	-	-
TCMSP	398	2,906	2,754	6,058	-	-	-	-
Union	118,300	71,821	30,000	220,121	688,200	530,242	51,700	1,270,142
Overall Total	192.015	267,409	139,545	596,117	1,164,074	1,173,013	563,365	2,900,452

• *The Prosecution Service is not organized at the municipal level. A cell with - indicates that no data was found within the search interval. Source: Survey data.

With regard to social network activities, posts by the MPs are twice as high as those by the TCs, which again shows that the MPs are more active in using social media as institutional communication channels. Twitter is the social network with the highest number of posts among TCs and MPs, especially for the MPF (Table 3). Twitter is not only the most widely used platform by these institutions; for the TCU and the MPF, it is the platform that is prioritized, which confirms its importance for the public communication of State agencies, as already noted in the pertinent literature, both abroad and in Brazil.

TABLE 3
NUMBER OF POSTS BY TCS AND MPS ON TWITTER, FACEBOOK, AND INSTAGRAM IN 2021*

Organs	Court of Accounts				Prosecution Service
	Twitter	Facebook	Instagram	Total TC	Twitter	Facebook	Instagram	Total MP
State	85,162	-	15,803	100,965	155,072	-	41,522	196,594
AC	341	-	234	575	6,899	-	1,707	8,606
AL	1,102	-	788	1,890	1,636	-	1,223	2,859
AM	7,901	-	875	8,776	893	-	395	1,288
AP	22	-	370	392	8,187	-	2,409	10,596
BA	197	-	93	290	3,798	-	992	4,790
CE	3,775	-	726	4,501	4,450	-	2,863	7,313
DF	335	-	-	335	5,962	-	1,542	7,504
ES	404	-	830	1,234	3,644	-	693	4,337
GO	5,146	-	1,105	6,251	7,277	-	3,233	10,510
MA	33	-	218	251	4,252	-	4,817	9,069
MG	4,485	-	93	4,578	9,634	-	984	10,618
MS	958	-	93	1,051	4,631	-	2,348	6,979
MT	9,586	-	1,067	10,653	3,029	-	2,211	5,240
PA	956	-	906	1,862	9,409	-	915	10,324
PB	7,355	-	2,444	9,799	9,499	-	731	10,230
PE	4,468	-	93	4,561	8,333	-	2,690	11,023
PI	1,435	-	585	2,020	2,482	-	1,592	4,074
PR	6,369	-	277	6,646	2,678	-	622	3,300
RJ	175	-	483	658	12,500	-	1,005	13,505
RN	2,877	-	169	3,046	6,010	-	309	6,319
RO	1,018	-	381	1,399	3,599	-	1,224	4,823
RR	1,197	-	759	1,956	1,160	-	299	1,459
RS	6,467	-	231	6,698	6,265	-	156	6,421
SE	3,205	-	348	3,553	7,493	-	1,166	8,659
SC	3,443	-	1,406	4,849	4,136	-	1,463	5,599
SP	7,819	-	397	8,216	10,100	-	1,194	11.294
TO	4,093	-	832	4,925	7,116	-	2,739	9,855
Municipal	9,963	-	3,084	13,047	-	-	-	-
TCMBA	578	-	-	578	-	-	-	-
TCMGO	5,982	-	1,485	7,467	-	-	-	-
TCMPA	1,781	-	925	2,706	-	-	-	-
TCMRJ	123	-	116	239	-	-	-	-
TCMSP	1,499	-	558	2,057	-	-	-	-
Union	5,042	-	204	5,246	36,000	-	288	36,288
Overall total	100,167	-	19,091	119,258	191,072	-	41,810	232,882

• * Facebook data is not available. A cell with - indicates that no data was found within the search interval. Source: Survey data.

It is worth noting that the use of social networks does not entail only public communication. It is also a channel of political communication, not only in the interaction with other State agencies and with citizens, but also in electoral campaigns and in moments of political crisis. In addition to the professionalization of the communication process, through the creation of specific departments and the hiring of skilled professionals in the area, the literature draws attention to the political dimension of communication efforts through social media (Fontes & Marques, 2022; Goldbeck et al., 2010; Hong, 2013; Vergeer et al., 2011), further stressing that attention should be focused on the content and the timing for disseminating messages (Damgaard, 2018; Feres & Sassara, 2016). According to the literature, postings may occur with more intensity precisely in moments of political and social upheaval, moments when a greater protagonism by members of these oversight agencies in the political scene cannot be overlooked.

GRAPH 1
FREQUENCY OF TWEETS, REPOSTS, AND RETWEETS BY TCU AND MPF (2009-2020)

As can be seen in Graph 1, the TCU’s activity starts in 2009, has its activity reduced in the following months and, compared to the MP between 2011 and 2014, had low activity on Twitter. The TCU’s and MPF’s activity on Twitter was intensified in terms of frequency and total number of posts between 2014 and 2018, a turbulent period in Brazilian politics, with a marked increase in 2015 for the TCU, year in which it had its peak activity on the social network, when the pressure for the impeachment of President Dilma Rousseff intensified. The graph also shows that, after the 2018 elections and with the decline of Operation Car Wash, for example, there is a sharp drop in the frequency of tweets, reposts, and retweets from both institutions.

It is worth noting that the MPF had already been significantly active on Twitter in an earlier period, on the eve of the 2014 elections (Viegas & Xavier, 2021). The TCU was most active in the period preceding the publication of its preliminary opinion rejecting the accounts of President Dilma Rousseff in 2015, which provided the basis for the impeachment process (Graph 1). These findings most certainly do not establish causality between the actions of oversight agencies and political events, which would require a much broader statistical and contextual analysis.

What we want to highlight here is that digital communication through the social networks of the TCU and, above all, the MPs, may be consistent with a rationale more closely linked to the preferences of their members than to the issues and concerns that the controlling authorities were supposed to address, in accordance with the legal framework of those institutions. Based on the concept borrowed from the American experience, there may be more outreach than transparency in this model of communication with the public (Goldbeck et al., 2010).

Given the political importance of the communication process of State bureaucratic institutions in social media, we shall now proceed with the content analysis, bearing in mind that it has not been the object of study so far.

5. TCU AND MPF TWITTER CONTENT ANALYSIS

In the content analysis of the TCU and MPF posts on Twitter, with the help of Iramuteq, we observed mainly similarities, but also differences in the activities of both institutions. In the case of the TCU, the text corpus was constituted by 5,160 text strings (TSs). There were 71,964 occurrences, of which 6,752 were distinct words and 3,626 had a single occurrence. In the content analysis of the MPF’s Twitter feed, the text corpus consisted of 36,878 TSs. A total of 65,5139 occurrences emerged, with 25,148 distinct words and 13,814 with a single occurrence. Just with this information regarding the text corpus and comparing the two, it is clear that the MPF’s posting activity on the network is much higher than the TCU’s, approximately six-fold.

In the case of the TCU, similarity analysis shows that some clusters are formed, revealing that the profile performs RTs of its own posts. When subjecting this finding to specific content analysis, @brasildigital comes up frequently and is associated at the source with “ao vivo” [live] (Figure 1). This point is interesting because it suggests that the TCU uses its Twitter account mainly to report events it organizes, such as seminars, forums, and programs it broadcasts live on its YouTube channel. But there are other important issues regarding the content of the information that the TCU tries to associate with its audiovisual communication channels.

FIGURE 1
SIMILARITY ANALYSIS - TCU TWITTER

The word cloud generated by Iramuteq reveals the disproportionate use of terms in the corpus, especially in the case of the TCU (Figure 2), converging with the similarity analysis. In the active forms, “TCU” features the highest frequency in the corpus, while the second ranked term is “público” [public]. Aggregating the findings from both the similarity analysis and the word cloud and subjecting this material to conventional content analysis (reduced unit of analysis), we conclude that the content developed by the TCU on Twitter has a predominantly self-referential content, in a process similar to what was already described for the use of digital democracy by US congressmen, as mentioned above (Goldbeck et al., 2010).

FIGURE 2
WORD CLOUD - TCU TWITTER

The occurrences of other words and the connection between them represent another important evidence of the posts’ content. The words “playlist”, “youtube”, “lista” [list], and “adicionar” [add], show that the TCU YouTube channel is well publicized by the profile; and “plenário” [plenary] and “decisão” [decision], the kind of news it typically sought to broadcast on the network between 2009 and 2020. By submitting this reduced content (TCU on YouTube) to conventional theme analysis, corroborating it with the tests applied, we found that the TCU generates content on Twitter aimed at disseminating its oversight activities with regard to other institutions, especially those that are part of the Executive Branch, doing so through Twitter, in addition to replicating it to its other social media and YouTube, thus leveraging the reach of the information it produces. In any case, there are indications that this activity on social media occurs to the detriment of transparency and publicity of management acts performed by the TCU itself, such as data on budget, financial and asset execution, contracts signed and management or cost-saving acts, which corroborates previous research showing that there has been little change in the TCs’ practices in the country (Arantes et al., 2010).

In the case of the MPF, similarity analysis was less trivial, since the number of occurrences is very high, considering the institution’s engagement on Twitter (Figure 3).

FIGURE 3
SIMILARITY ANALYSIS - MPF TWITTER

Because of the limitations posed by Iramuteq’s incomplete dictionary, some specific terms, jargons, and proper names end up “tainting” certain tests, especially in the similarity analysis. In the case of Figure 3, even limiting the occurrences to words with frequency greater than or equal to 80 in order to clean up the visualization, it was necessary to exclude pairs of outlier terms, such as: “deborah” and “duprat”; “habeas” and “corpus”; “eduardo” and “cunha”; “sustentação” and “oral” [oral argument]; “roberto” and “gurgel”; “falsidade” and “ideológico” [fraudulent misrepresentation]; “raquel” and “dodge”; “lava” and “jato” [car wash]; “improbidade” and “administrativa” [administrative misconduct]; “procurador” and “república” [federal prosecutor]; “mp” and “eleitoral” [electoral prosecution service]; “lavagem” and “dinheiro” [money laundering]. In general, these are proper names of politicians under investigation and, above all, of heads of the MPF (Prosecutors General), which suggests personalism in the communication, but also technical terms or terms with figurative meaning, warranting their removal by Iramuteq, since their inclusion would generate disconnected and meaningless clusters.

Just as for the TCU, the words “youtube”, “radio” and “ao vivo” [live] form a cluster, indicating that the MPF uses its Twitter profile to disseminate its activities in real time, thus maximizing the reach of communication about events, seminars and trial sessions. In this respect, we noticed an important difference when compared to the TCU. By subjecting this finding (reduced content) to specific (conventional) content analysis, we found that the MPF, in tandem with the National Association of Federal Prosecutors (Associação Nacional dos Procuradores da República, or ANPR), operates as a “spokesperson” for routine matters, but also for those linked to the interests of MPF members, especially functional prerogatives. The ANPR is a private entity advocating the corporate interests of MPF members (Viegas, 2020).

One of the most important aspects was found, based on the clusters generated, in the content of news with strong media repercussion that were published in the MPF’s Twitter account. Typically, news derive from the activities of members of the institution, often identified by their names, which again suggests the personalism in communications, as in the cases of anti-corruption cases and the Car Wash Operation (Lava Jato), and involve topics about the performance of the national political system, which are replicated by the ANPR and by other MPF offices, thus corroborating a previous exploratory study (Viegas & Xavier, 2021). As an example, Rodrigo Janot and Raquel Dodge, heads of the MPF for four (2013-2017) and two years (2017-2019), respectively, are mentioned more than 600 times each in the unreduced database. Only the coordinator of Car Wash in Curitiba, Deltan Dallagnol, was mentioned 62 times in the MPF’s Twitter communication related to this operation.

Moreover, similarity analysis of the MPF’s Twitter content shows that some clusters emerged, revealing that the profile does RTs of itself, just like the TCU (Figures 1 and 3, above). However, for the MPF, this occurs with more intensity through the interaction of other offices within the institution itself, called Offices of Federal Circuit Prosecution and Offices of Federal Prosecution, spread throughout Brazil, in addition to the ANPR (Viegas & Xavier, 2021). That is, there are MPF offices spread throughout Brazil with their own Twitter accounts, as well as the ANPR, which boost the dissemination of content produced by the MPF’s root account, suggesting that this may be an institutional communication strategy. Not surprisingly, the word “rt” [retweet] is among the most frequent in the text corpus (Figure 4).

FIGURE 4
WORD CLOUD - MPF TWITTE

The word cloud generated from the analysis of the MPF text corpus also shows, similarly to the TCU case, the disproportionate use of terms, especially the use of “mpf”, which also suggests the conveyance of self-referenced information (Figure 4). The visual representation by the word cloud converges toward the analysis of similarity of the MPF content on Twitter. In this aspect, we identified the occurrences among these words and the connectedness between them, such as the words “pgr”, “ação” [lawsuit], “denunciar” [to denounce], which are related to the activities performed by the MPF. Like the TCU, we found evidence that the MPF prioritizes content posted on Twitter in connection with the control it performs over other institutions, to the detriment of transparency and publicity of the institution’s management acts so that other institutions and the general public can learn about them.

5. DISCUSSION AND CONCLUDING REMARKS

The Brazilian TCs and MPs are active in posting on social networks, with a large volume of activity on Twitter when compared to the other social networks reviewed. This communication is still barely studied, but it deserves attention from researchers devoted both to political communication and to the study of the forms of democratic control over public administration, especially considering the centrality of Twitter in generating journalistic agendas for politics, which cannot be underestimated.

As a result of the formal arrangement that outlines their attributions in defending rights in a broad sense, the MPs are perceived as being closer to the public (Arantes, 2002; Arantes et al., 2010; Kerche, 2009). This may help to understand the more active posting of diverse content by the institution on social media when compared to the TCs.

However, for a better understanding of this phenomenon, we cannot disregard the protagonism that the members of the institution have assumed in recent years (Arantes & Moreira, 2019; Avritzer & Marona, 2017; Kerche & Marona, 2018; Silva, 2020). Also as a result of the autonomy conferred to the MP and the discretional power of its members, these bureaucrats are in a position to select priorities for action, which, in recent years, has turned predominantly to issues associated with the fight against corruption at the expense of other areas for which they are equally entitled to act, such as health and education (Kerche & Viegas, 2020).

Before analyzing the research findings, it is worth highlighting an aspect that, while outside the focus of this study, may help explain the communicational action developed by these organizations on social media: MPF members are often in a position to articulate career political strategies when carrying out inquiries and lawsuits with repercussions inside and outside the institution (Viegas, 2020), so it is likely that this is replicated in the area of communication on social media. This heightened level of personalism among MPF members affects the way in which communication might accomplish the accountability of the institution as a whole.

Among the key findings of our survey, we draw attention to the use that the TCU and the MPF make of Twitter. In this context, there is a lower incidence of issues related to transparency and accountability of the institutions themselves, such as budgetary, financial and asset execution; contracts; management or cost-saving acts; personnel data; public policy data; costs; payments; governance policies; goals plan; social engagement actions: conferences, councils, public hearings, public consultations, and social dialog, among others.

In fact, even though the time span is not short (10 years) and the analysis could be done by cleavages (cross-sections to verify trends at specific moments), the set of data analyzed and the comparison of their activities on the web suggest that the TCU and the MPF are mainly producing communication with self-referenced or self-centered content, with a high degree of personalism in the case of the MPF. That is, a communication model geared towards the promotion and legitimization of their oversight activities with regard to other institutions and society, instead of ultimately publicizing information that would generate input for political and social actors to “control the controllers”.

The self-centered model of communication, in varying degrees, is found in other State institutions, as pointed out in the literature. However, unlike other cases, the TCU and the MPF are institutions designed to exercise control over the public administration, which would imply the need to set an example to others by reporting to the public how they work. The point is not to say that the TCU and the MPF are not transparent, but to highlight what is the focus of the transparency they display on social media.

According to the research data, the TCU and the MPF predominantly use communication that promotes transparency about the acts of oversight, and not about how the controller follows the rules and legislation governing its operation, whose prerequisites were established by the democratic decision of politicians. Controllers define the object of their communication and the language used to present it, and these processes affect the quality of accountability.

In this sense, we draw attention to the content disseminated by these two institutions on Twitter, which contains excessively technical expressions (“accounting speak” and “legal speak”, which is even more evident in the case of the MPF). This occurs, for example, with the constant mention of articles of laws and technical concepts such as bidding and tendering, i.e., something abstract and of difficult wording for most of the population, which may have little meaning for the average citizen and even for civil society, which may be engaged but not specialized in the topic, thus hindering the effective transparency of these institutions.

As studies on open government have warned (Gil-Garcia, Gasco-Hernandez, & Pardo 2020; Matheus & Janssen, 2019; Schmidthuber et al., 2021; Schnell & Jo, 2019), this excessively technical form of communication does not contribute to interaction with the citizens capable of generating new models of participation. Explaining what was not right, which investments in education or healthcare were not adequate, in a simple, accessible and not self-centered language, may encourage citizen participation and, in the future, serve as a reference for assessing the impact of open data.

In addition to observing practices of self-promotion and legitimization vis-a-vis other institutions and society, there is also a political component embedded in the communication of these agencies. In the case of Twitter, we found that the activities of the TCU and the MPF in publicizing anti-corruption efforts become more or less intense at specific times, for example, at the height of the Car Wash Operation and, in particular, during the impeachment process of Dilma Rousseff and the 2018 elections. Even though a causal relationship cannot be inferred between the higher or lower frequency of Twitter posts and specific political contexts (Graph 1), due to the limitations of this study, especially in the case of the MPF, the individual action of prosecutors is given considerable relevance in the content that is disseminated, making it difficult to distinguish the communication and transparency of the institution from what is personal to its members.

Certainly, the topic of this study calls for new interdisciplinary research, with additional efforts in the areas of public administration and political communication, in order to develop analyses combining different methods, both qualitative and quantitative. There is scarce research, for example, on how members of the MP chart their own political careers within and outside that bureaucracy (Viegas, 2020). In this sense, it may be interesting to know how prosecutors have used their social networks to promote and legitimize themselves, assisted or not by the internal communication departments of each MP office, and therefore, in some cases, in a professional way and with their political careers in mind.

This line of research might even have its scope widened to include other institutions and their members, which are also part of the justice system and the public administration accountability network, such as the Federal Attorney General’s Office (Advocacia-Geral da União, or AGU), the State Attorney’s Offices, the Police and the Judiciary itself. As this activity on social media may involve the institutional profile, i.e., the official accounts, for example, of the MPs, the AGU, the Police and the Judiciary, which replicate the disseminated content, along with entities representing corporate interests, such as class associations, there is a good chance that the interaction of individual profiles of members of these institutions with the official channels to which they are affiliated ultimately serve to maximize political career strategies through the angle of communication.

We conclude that it is necessary to initiate an analysis of the communication of oversight institutions and of their actions according to an open government framework that uses the democratic theory of accountability as a reference, avoiding a naive approach based on the idea that the more information these institutions publish on social media, the more accountable they are to the public. It is necessary to return to Weber’s classic argument, revisited by Gruber (1986), that democratic control of the unelected must distinguish the pursuit of self-interest by members of the State from what is effectively the public interest, thus devising democratic-republican mechanisms to hold accountable all structures of State power.

ACKNOWLEDGMENTS

The authors wish to thank Lucas Xavier and Amanda Toledo for their support. They also thank the anonymous referees for their suggestions to the current version, as well as FGV EAESP for the financial support provided to this research.

REFERÊNCIAS

• Abrucio, F., & Loureiro, M. R. (2005). Finanças públicas, democracia e accountability. In C. Birderman, & P. R. Arvate, (Orgs.), Economia do setor público no Brasil(pp. 75-102). Rio de Janeiro, RJ: Editora Campus Elsivier.
• Abrucio, F., & Loureiro, M. R. (2018). Burocracia e ordem democrática: desafios contemporâneos e experiência brasileira. In R. Pires, G. Lotta, V. E. Oliveira (Orgs.), Burocracias e políticas públicas no Brasil: interseções analíticas(pp. 23-56). Brasília, DF: Ipea.
• Amaral, M. S. (2016). Congresso no Twitter: parlamentares e partidos políticos em 140 caracteres (Tese de Doutorado). Universidade Federal da Bahia, Salvador, BA.
• Arantes, R. (2002). Ministério Público e política no Brasil São Paulo, SP: Educ-Sumaré.
• Arantes, R., Loureiro, M. R., Couto, C. G., & Teixeira, M. A. C. (2010). Controles democráticos sobre a administração pública no Brasil: Legislativo, Judiciário, Tribunais de Contas e Ministério Público. In M. R. Loureiro, F. Abrucio, & R. Pacheco (Orgs.), Burocracia e Política no Brasil. Desafios para o Estado democrático no século XXI(pp. 109-147). Rio de Janeiro, RJ: Editora FGV.
• Arantes, R., & Moreira, T. (2019). Democracia, instituições de controle e justiça sob a ótica do pluralismo estatal. Opinião Pública, 25(1), 97-135.
• Avritzer, L., & Marona, M. (2017). A tensão entre soberania e instituições de controle na democracia brasileira. Dados, 60(2), 359-393.
• Bobbio, N. (1997). O futuro da democracia: uma defesa das regras do jogo Rio de Janeiro, RJ: Paz e Terra.
• Bobbio, N. (2000). O futuro da democracia Rio de Janeiro, RJ: Paz e Terra .
• Camargo, B., & Justo, A. M. (2013). Iramuteq: um software gratuito para análise de dados textuais. Temas em Psicologia, 2(21), 513-518.
• Carnielli, F. Z. (2017). A instituição da cidadania: a comunicação da Defensoria Pública do RS. In M. H. Weber, M. P. Coelho, C. & Locatelli (Orgs.), Comunicação pública e política: pesquisa e práticas (pp. 439-460). Florianópolis, SC: Editora Insular.
• Castells, M. (1999). A sociedade em rede São Paulo, SP: Paz e Terra.
• Constituição da República Federativa do Brasil de 1988 (1988). Brasília, DF. Recuperado de http://www.planalto.gov.br/ccivil_03/constituicao/constituicao.htm
  » http://www.planalto.gov.br/ccivil_03/constituicao/constituicao.htm
• Cunha, M. A. V. C., & Miranda, P. R. M. (2013). O uso de TIC pelos governos: uma proposta de agenda de pesquisa a partir da produção acadêmica e da prática nacional. Organizações & Sociedade, 20(66), 543-566.
• Dahl, R. (1997). Poliarquia. Participação e oposição São Paulo, SP: EDUSP.
• Damgaard, M. (2018). Cascading corruption news: explaining the bias of media attention to Brazil’s political scandals. Opinião Pública, 24(1), 114-143. Recuperado dehttps://doi.org/10.1590/1807-01912018241114
  » https://doi.org/10.1590/1807-01912018241114
• Esteves, J. P. (2011). Sociologia da comunicação Lisboa, Portugal: Fundação Calouste Gulbenkian.
• Feres J., Jr; & Sassara, L. (2016). Corrupção, escândalos e a cobertura midiática da política. Novos Estudos CEBRAP, 35(2), 205-225. Recuperado de https://doi.org/10.25091/s0101-3300201600020011
  » https://doi.org/10.25091/s0101-3300201600020011
• Fontes, G. S, & Marques, F. P. J. (2022, março). Defending democracy or amplifying populism? Journalistic coverage, Twitter, and users’ engagement in Bolsonaro’s Brazil. Journalism Recuperado de https://doi.org/10.1177/14648849221075429
  » https://doi.org/10.1177/14648849221075429
• Gil-Garcia, J. R., Gasco-Hernandez, M., & Pardo, T. A. (2020). Beyond transparency, participation, and collaboration? A reflection on the dimensions of open government Public Performance & Management Review, 43(3), 483-502.
• Goldbeck, J., Grimes, J., & Rogers, A. (2010). Twitter use by the U.S. Congress. Journal of the American Society for Information Science and Technology, 61(8), 1612-1621.
• Gomes, W. (2005). Internet e participação política em sociedades democráticas. Revista Famecos, 27, 58-78.
• Gomes, W. (2011). Participação política on-line: questões e hipóteses de trabalho. In R. C. M. Maia, W. Gomes, & F. J. A. Marques(Orgs.), Internet e participação política no Brasil(pp. 19-46). Porto Alegre, RS: Sulina.
• Gomes, W. (2016). 20 anos de política, estado e democracia digitais: uma “cartografia” do campo. In S. P. Silva, R. C. Bragatto, & R. C. Sampaio(Orgs.), Democracia digital, comunicação política e redes: teoria e prática (pp. 25-45). Rio de Janeiro, RJ: Letra & Imagem.
• Gruber, J. E. (1986). Controlling bureaucracies: dilemmas in democratic governance Los Angeles, CA: University of California Press.
• Hong, S. (2013). Who benefits from Twitter? Social media and political competition in the U.S. House of Representatives. Government Information Quarterly, 30(4), 464-472.
• Kerche, F. (2009). Virtude e limites: autonomia e atribuições do MP no Brasil São Paulo, SP: Edusp.
• Kerche, F., & Marona, M. (2018). O Ministério Público na Operação lava Jato: como eles chegaram até aqui? In F. Kerche, & J. Feres Jr. (Orgs.), Operação Lava Jato e a democracia brasileira(pp. 69-100). São Paulo, SP: Contracorrente.
• Kerche, F., & Viegas, R. R. (2020). Ministério Público: de defensor de direitos a combatente da corrupção. In Anais do 12º Encontro da ABCP, João Pessoa, PB.
• Lei nº 12.527, de 18 de novembro de 2011 (2011). Regula o acesso a informações previsto no inciso XXXIII do art. 5º, no inciso II do § 3º do art. 37 e no § 2º do art. 216 da Constituição Federal; altera a Lei nº 8.112, de 11 de dezembro de 1990; revoga a Lei nº 11.111, de 5 de maio de 2005, e dispositivos da Lei nº 8.159, de 8 de janeiro de 1991; e dá outras providências. Brasília, DF. Recuperado de http://www.planalto.gov.br/ccivil_03/_ato2011-2014/2011/lei/l12527.htm
  » http://www.planalto.gov.br/ccivil_03/_ato2011-2014/2011/lei/l12527.htm
• Mainwaring, S., & Welma, C. (2003). Democratic accountability in Latin America Oxford, UK: Oxford University Press.
• Matheus, R., & Janssen, M. (2019). A systematic literature study to unravel transparency enabled by open government data: the Window Theory. Public Performance & Management Review, 3(3), 503-534.
• Mayring, P. (2004). Qualitative content analysis. In U. Flick, E. Kardrff, & I. Steinke (Orgs.), A companion to qualitative research (pp. 265-269). London, UK: Sage Publications.
• Melo, C. de F., Cavalcante, A. C. & Façanha, K. Q. (2019). Invisibilidade do adoecimento psíquico do trabalhador: limites da integralidade na rede de atenção à saúde. Trabalho, Educação e Saúde, 17(2), e0020132.
• Michener, G. (2019). Gauging the impact of transparency policies. Public Administration Review, 79(1), 136-139.
• Miguel, L. F., & Biroli, F. (2010). Mídia, representação e democracia no Brasil - estudos sobre comunicação política São Paulo, SP: Hucitec.
• O’Donnell, G. (1998). Accountability horizontal e novas poliarquias. Lua Nova, 44, 22-54.
• Olivieri, C., Loureiro, M. R., Teixeira, M. A. C., & Abrucio, F. (2015). Control and public management performance in Brazil: challenges for coordination. International Business Research, 8, 181-190.
• Pelizzo, R., & Stapenhurst, F. (2014). Government Accountability and Legislative Oversight New York, NY: Routledge.
• Pinho, J. A. G. (2008). Investigando portais de governo eletrônico de Estados no Brasil: muita tecnologia, pouca democracia. Revista de Administração Pública, 42(3), 471-93.
• Pinho, J. A. G. (2011). Sociedade da informação, capitalismo e sociedade civil: reflexões sobre política, internet e democracia na realidade Brasileira. Revista de Administração de Empresas, 51(1), 98-106.
• Pinho, J. A. G., Iglesias, D. M., & Souza, A. C. P. (2005). Governo eletrônico, transparência, accountability e participação: o que portais de governos estaduais no Brasil mostram. In Anais do 29º Encontro da Associação Nacional de Pós-Graduação e Pesquisa em Administração, Brasília, DF.
• Raupp, F. M., & Pinho, J. A. G. (2013). Ranking dos legislativos locais na construção da accountability: um estudo a partir dos portais eletrônicos de municípios de Santa Catarina. Revista de Contabilidade da UFBA, 7(1), 69-83.
• Ratinaud, P., & Marchand, P. (2012). Application de la méthode ALCESTE à de»gros» corpus et stabilité des «mondes lexicaux»: analyse du «Cable-Gate» avec IraMuTeQ. In Actes des 11eme Journées internationales d’Analyse statistique des Données Textuelles, Liège, Belgium.
• Reinert, M. (1990). Alceste une méthodologie d’analyse des données textuelles et une application: Aurelia De Gerard De Nerval. Bulletin of Sociological Methodology/Bulletin de Méthodologie Sociologique, 26(1), 24-54.
• Rocha, A. C., Spaniol, E. L., Schommer, P. C., & Sousa, A. D. de. (2012). A coprodução do controle como bem público essencial à accountability. In Anais do 36º Encontro da Associação Nacional de Pós-Graduação e Pesquisa em Administração, Rio de Janeiro, RJ.
• Sampaio, R. C., & Lycarião, D. (2021). Análise de conteúdo categorial: manual de aplicação Brasília, DF: Enap.
• Schmidthuber, L., Ingrams, A., & Hilgers, D. (2021). Government openness and public trust: the mediating role of democratic capacity. Public Administration Review, 81(1), 91-109.
• Schnell, S., & Jo, S. (2019). Which countries have more open governments? Assessing structural determinants of openness. American Review of Public Administration, 49(8), 944-956.
• Silva, F. S. (2020). From car wash to Bolsonaro: law and lawyers in Brazil’s Illiberal Turn (2014-2018). Journal of Law and Society, 47, S90-S110.
• Vergeer, M., Hermans, L., & Sams, S. (2011). Online social networks and micro- blogging in political campaigning: the exploration of a new tool and a new campaign style London, UK: SAGE Publications.
• Viegas, R. R. (2020). Governabilidade e lógica das designações no Ministério Público Federal: os procuradores políticos profissionais. Revista Brasileira de Ciência Política, 33, e234299. Recuperado dehttps://doi.org/10.1590/0103-3352.2020.33.234299
  » https://doi.org/10.1590/0103-3352.2020.33.234299
• Viegas, R. R. & Borali, N. (no prelo). Análise de conteúdo e o uso do Iramuteq. Revista Latinoamericana de Metodología de la Investigación Social
• Viegas, R. R., & Xavier, L. (2021). The political use of Twitter by the Federal Prosecution Service in Brazil. In Proceedings of the IAPSS Virtual World Congress 2021: Democracy, Identity, and Power, Montréal, Québec.
• Weber, M. H. (1993). Parlamento e governo na Alemanha reordenada: crítica política do funcionalismo e da natureza dos partidos Petrópolis, RH: Vozes.
• Weber, M. H. (2011). Estratégias da comunicação de Estado e a disputa por visibilidade e opinião. In M. Kunsch(Org.), Comunicação pública, sociedade e cidadania (pp. 101-120). São Caetano do Sul, SP: Difusão.
• Weber, M. H. (2017). Nas redes de comunicação pública, as disputas possíveis de poder e visibilidade. In M. H. Weber, M. P. Coelho, & C. Locatelli (Orgs.), Comunicação pública e política: pesquisa e práticas (pp. 23-56). Florianópolis, SC: Editora Insular .

• 1
  In Brazil, there are 33 Courts of Accounts organized at the three levels of the federation: at the federal level (the Federal Court of Accounts, or TCU), at the state level (State Courts of Accounts, or TCEs, in all 26 states, including the Federal District, plus three Courts of Accounts for the municipalities in the states of Bahia, Goiás, and Pará), and at the municipal level (two Municipal Courts of Accounts, or TCMs, in the cities of São Paulo and Rio de Janeiro). The Prosecution Services are organized at the federal level (Prosecution Service of the Union, or MPU) and at the state level (State Prosecution Services, or MPEs).
• 2
  The results of these two phases are presented in a separate section. For the results obtained with the software (report), we suggest taking into account the description provided by Melo, Cavalcante, and Façanha (2019).

• [Translated version] Note: All quotes in English translated by this article’s translator.

Publication Dates

• Publication in this collection
  15 July 2022
• Date of issue
  May-Jun 2022

History

• Received
  22 Sept 2021
• Accepted
  26 Mar 2022