Brazil ’ s Information Ecosystem : What is Transparency ’ s Impact ? 1

Does transparency make a difference in Brazil? The current essay discusses the determinants and health of Brazil’s information ecosystem; it advances arguments on how and why Brazil has tended to underinvest in information and transparency, and discusses postures the government ought to adopt as a means of moving forward. Key-words: Transparency Ecosystem; Transparency Impacts; Transparency and Corruption; Proactive and Reactive Transparency Resumo: A transparência faz a diferença no Brasil? O presente ensaio discute os fatores determinantes e a saúde do ecossistema de acesso à informação no Brasil, desenvolve argumentos que explicam como e por que o Brasil tende a sub-investir em acesso à informação e transparência e discute a postura que o governo precisa adotar como forma de promover avanços. Palavras-chave: Ecossistema da Transparência; Impactos da Transparência; Transparência e Corrupção; Transparência Ativa e Passiva W people ask me what I do, and I tell them that I research transparency, all too frequently the question then becomes, “do you really think transparency makes a difference in a country as corrupt and ignorant as Brazil?” This essay is a response to that question. Questioning the efficacy of transparency – not just in Brazil, but anywhere – is nothing new. Since transparency and good government became fashionable in the 1990s, scholars began to ask about impacts, pointing to the unintended consequences of government transparency, its excesses, and the questionable value of its impact. Yet the subject and object of the question, “what positive impact is transparency having?” can easily mislead. 1 Recebido em 23 de abril e aprovado em 02 de Julho de 2019

Start with 'impact,' the subject of the above question.Positive impacts can be extremely difficult to identify.As well as being colorless and odorless, transparency tends to have diffuse, indirect, and long-term impacts, many of which are preventative and therefore difficult to detect.Moreover, transparency is a property of information (either an input or output), and 'impact' is generally attributed to individuals or institutions, not information proper.For all of these reasons, which are explored in the first section of this essay, transparency's impact can be extremely challenging to gauge.
Proceeding with 'transparency', the object of the above question, the concept is frequently taken to be much more than it is.Transparency has no metaphysical properties, but instead is simply information that is visible and inferable (MICHENER e BERSCH, 2013).In this sense, transparency is not the monopoly of government, but rather is produced and consumed by individuals, collectivities, as well as the news media.The potential impact of transparency therefore depends not just on government, but also on a broader information ecosystem of which the government is one critical element.
For an ecosystem to thrive, the right sort of elements need coalesce.The key and foundational element of a country's transparency ecosystem -an information ecosystem constituted by visible, inferable information -is an educated populace.The other elements consist of a vibrant news media (of which social media can play a part) and a proactive government.Consistent with the metaphor of an 'ecosystem', all elements are self-reinforcing: an educated populace and their respective private and public collectivities (e.g.businesses, associations) represent both the supply and demand sides of information.A proactive government produces and discloses useful, actionable transparent information, which is the stuff that goes into good media reporting, and helps citizens set policy priorities and keep officials accountable.A strong news media, in whatever shape -from citizen reporters, bloggers, to small and big media companies -bridges the information gaps among citizens, government, and private and public collectivities.
All of this is well known, but one important point is often forgotten -it is ultimately government's actions and policies that de-termine the health and transparency of an information ecosystem.Its own production of information plays a big part, but more important is the extent to which government's regulatory and welfare policies cultivate an educated populace and a news media that is free and independent (i.e.vibrant).Good governments invest heavily in the foundations of good information and in institutions that advance and promote transparency.This essay argues that Brazil undervalues information in significant ways, misallocating its finite resources and underinvesting in information.This paper casts light on several misallocations and underinvestments, proposing ways forward.In briefly analyzing the question of transparency's impact, the first part of this essay casts some light on one important reason why countries underinvest in transparency and information more generally.The second part of the essay examines some basic indicators of how Brazilian governments underinvest in information.The conclusion points to specific actions that may help Brazil invest more intelligently.

The Impact of Transparency 2
If one looks at countries such as Sweden and Finland, which have had transparency and freedom of information (FOI) statutes since the mid 18th Century, one would be hard-pressed to identify impacts directly attributable to transparency.One might be tempted to conjecture that, in forcing public servants to reveal their work, transparency made politicians and public servants more disciplined, professional, and public-oriented.The logic makes sense, but it is nearly impossible to establish a direct causal relationship between excellence in governance and transparency.Several rationales help explain why, and in so doing, they also help explain why most leaders around the world and in Brazil tend to underinvest in transparency.
First, transparency is often considered a deterrent and it can be difficult to convincingly identity the impact of deterrents.An official who knows his or her work will be made public should be less inclined towards incompetence, malfeasance, or inefficiency.In this sense, when transparency prevents a negative externality such as corruption from occurring, we will be none the wiser that an impact has occurred.In this example, transparency has accomplished what legislative scholars refer to as "negative agenda-setting".It has denied corruption a way forward.However, the deterrent effects of transparency are difficult to identify, and corresponding changes in behavior should only be detectable over the medium to long term.
In effect, some positive impacts associated with transparency do occur in the short to medium term, but are often misinterpreted as transparency's failures.In December Is it a coincidence that corruption scandals -and perhaps more importantly, their persistence in Brazilian news headlines -have become more salient as the country has become more transparent?While transparency may not have 'triggered' corruption scandals, it may help keep scandals from being summarily buried.Transparency facilitates internal investigations by authorities and guards against claims that information is private or confidential.It also allows officials to publicize and even promote their investigative and prosecutorial efforts -as evinced by the crusading prosecutor, Deltan Dallagnol.The bureaucratic showmanship and activism of Brazil's public prosecutor and federal police, in turn, feeds media coverage.
Evidence for the increasing importance of corruption in Brazil's collective consciousness comes from recent polling.In 2017, one of the most respected polls in the region, Latinobarômetro, found that Brazilians cited corruption as the country's most important problem (FIGUEREIDO, 2017).Most signi-ficantly, it was the first time in Latinobarômetro's 22-year history that corruption had topped Brazilians' preoccupations.Even more surprisingly, corruption beat out the economy -in the middle of Brazil's worst-ever recession.
Scholars have marshaled large cross-national datasets to show that the relationship between transparency and perceptions of corruption holds beyond Brazil.Perceptions of corruption do indeed worsen following the adoption of FOI laws (COSTA, 2013;VADLAMANNATI e COORAY, 2016).When adequately contemplated, this result should come as no surprise.Logically, perceptions of corruption will be heightened during the initial decades following the implementation of transparency policies, but should level out and drop as officials respond to the greater probability of being detected.A greater probability of being punished (as well as detected) should only help accelerate this process.
A second rationale to explain why transparency's impacts go unperceived and therefore undervalued has to do with the way people attribute causality to political phenomena.'Impacts' rarely tend to be attributed to factors other than agency (human or institutional action) or systems, such as groups of institutions or economic systems.In other words, as either an input or output, transparency does not excite the popular imagination to the same extent as a leader, institution or system.In summary, transparency has diffuse and indirect impacts that tend to bear positive fruit only over extended periods of time.Put differently, transparency's impacts imply a long wait and they are hard to trace.That does not mean we cannot measure them, and much more work is needed in this respect.Yet 'direct' causality is difficult to prove, especially in the short term.Even though transparency has been marketed as an antidote to corruption (BALL, 2009), it is clearly a long term and roundabout process.
The diffuse, indirect and long-term nature of transparency's impacts do not make transparency -or information in general -popular among unreflective or impatient leaders and voters.The more materially needy a political community, the more electoral cycles tend to make unreflective leaders and voters impatient for material advances.Under these con-ditions, transparency and the promise of better information contain comparatively minor allure.In many part of Brazil, such is the case.

Brazil's Information Ecosystem
The paradox here is that the political communities least likely to invest in the foundation of good information (i.e.transparency) are those most in need of good information.Uneducated populations tend to be less engaged with news (LECHELER e DE VREESE, 2017) and more likely to believe one-sided or ideologically polarizing information.Transparent government information can counteract one-sided or polarizing information, but only if two conditions are met.First, governments must be in possession of incentives to produce such information, and second, the information must be widely disseminated.It is unlikely that a large plurality of Brazil's political communities meet these two conditions.But before providing a rationale for substantiating this claim, the current task at hand is to examine the elements of a healthy and transparency ecosystem.

An Educated Populace
Brazil is not an educated populace in two ways that are important for a country's 'information ecosystem'.The first and most basic is formal education.This is a painful theme, and there no need to belabor the point beyond a few basic statistics.In 2013, Brazil's Institute for Geography and Statistics found that over a quarter of Brazilians were either illiterate (8.3%) or functionally illiterate (17.8%).Average years of schooling for Brazilians over 25 years old hovers around 7.5, with only 15 percent of citizens having gone on to higher education after high school.These data speak to basic capacity, the capacity to produce and consume good information, thereby contributing to the health of a transparency ecosystem.The obvious solution is to invest more and better, which is obviously easier said than done.
The other way that an 'educated populace' impacts the transparency ecosystem has to do with a public-regarding socialization.Socialization is, after all, at least as important an education as formal schooling.My family, including my brother, sister and I, wrote let-ters to our Canadian Ministers of Parliament as older children, adolescents, and adults; and they responded.With our parents we attended political party primaries, and community meetings and rallies, many of which put an end to projects that we saw as economically or environmentally threatening to our community or country.My parents did this because their parents did this, and so on and so forth.These patterns might not be considered 'the norm' in Canada, but they certainly exist at scale.The active citizenship described above also occurs in Brazil, but apart from a few regional examples, it hardly can be said to exist at scale.Information matters to Canadian citizens because it can be acted upon with transformative consequences.Information matters to Brazilian citizens too, but much less so for the great majority.There are good reasons why: acting on information can be dangerous and citizens may be ignored by political leaders or coopted media outlets.
The oligarchic governments and dictatorships that pattern Brazilian history have interrupted the formation of social capital via public-regarding socialization.As a professor, I see that Brazilian Millennials appear to have internalized the incipient democratic values their parents absorbed during the transition to civilian rule in the mid 1980s.How deep these values run, however, is questionable; the parents of Millennials drew only shallowly from their respective parents due to the dictatorship.The basic history lesson is that if one cannot act on information to generate positive change (such as during a dictatorship) than information -at least the sort that may feed transformations -ceases to be of great consequence.Analogously, if whom you know is a greater determinant of your station in life than what you know, then information also loses some of its value.All of this simplifies the issue greatly, perhaps too much.The general point, however -the extent to which social and political contexts permit information to be acted upon -is nevertheless an important means of explaining the value different societies place upon information.

A Vital News Media
The second element of a healthy transparency ecosystem is a vital news media.I define a vital news media by negative freedoms here, which I refer to as 'the Four-Rs': freedom from repression and relational restrictions (e.g.no access to policymakers), and freedom from regulatory and revenue-based distortions.Although violent repression of Brazilian journalists and media owners is high by international standards, 3 particularly in the hinterland, it is episodic rather than systemic.Journalists in Brazil are generally free from repression, although their independence is another question, and one that is extremely relevant for evaluating a country's transparency ecosystem.Yet far from being a systematic evaluation of Brazil's media independence or the Four-Rs, the modest ambition of this section is to point out a few regulatory and revenue -based distortions.
On the regulatory front, Brazil is clearly out of line with better international practices.The basis of the country's broadcasting law is from the 1960s, criteria for allotting concessions are vague and are decided by Congress.There are good reasons why advanced democracies confer authority to one central regulatory body and exclude legislators from decisions on concessions -the politicization of ownership.Unsurprisingly, ownership of media outlets may be Brazil's weakest regulatory link.Here, rules are predicated on de jure understandings of ownership, rather than being inferred on a de facto basis (MENDEL e SALOMON, 2011, p. 12, 21, 68).This detail means that politicians can exercise de facto control of media outlets while placing friends or relatives as de jure owners.It also means that large media corporations can use affiliates to sidestep rules regarding vertical and horizontal concentration.Finally, to maximize opportunities for regulatory failure, media governance is split among nine different entities, whose responsibilities frequently overlap (MENDEL e SALOMON, 2011, p. 20-21).It is little wonder that media ownership is notoriously opaque (INTERVOZES, REPORTERS WITHOUT BORDERS, 2019).
Governmental manipulation of news media coverage through the discretionary allocation of government advertising contracts constitutes a second major informational distortion in Brazil.Advertising expenditures by governments tend to represent a large percentage of total advertising revenues in Latin America, particularly in areas that are 3 For details, refer to the organization, Reporters Without Borders.
economically depressed -i.e. the hinterland.Government advertising contracts can distort the quality of media coverage in significant ways.Positive allocations go to media outlets that provide favorable coverage, and negative allocations hurt those who dare to criticize government.Media outlets may even blackmail politicians, threatening negative coverage unless advertising contracts flow.
By implication, 'buying the news media' with government advertising funds is an effective way of skewing the electoral playing field and skirting accountability responsibilities.Soon-to-be-published research shows a strong, statistically significant relationship between government advertising expenditures in Brazil's largest municipalities and mayoral reelection (KOPP e MICHENER, 2017).In other words, greater spending on advertising is associated with electoral returns, even when controlling for multifarious other factors, such as spending on infrastructure projects.This finding has significant implications, especially because news media independence should become more vulnerable as city-size decreases.Given that Brazil is a country of small cities, where only 310 of more than 5500 municipalities contain more than 100,000 people, the impact of government advertising expenditures on the quality of news media coverage should give significant reason for concern.The solution is to regulate and cap allocations and total expenditures on advertising, but as is the case with regulation -solutions are easily available but politically difficult to secure.Some will argue that the above problems are not all that grave, because traditional media's influence has waned in the age of social media.There is truth to the claim that social media can eclipse the importance of traditional news media.One paradigmatic example is the recently reelected Governor of Maranhão, Flávio Dino.Challenging the Sarney political and economic dynasty -whose allies control most of the state's big media -in back-to-back elections, Dino has relied almost exclusively on alternative media vehicles, social media, and government transparency.In terms of transparency, one year after Dino assumed power Maranhão went from a 2.2 on the Comptroller General's (CGU) Transparency Index to a maximum score of 10, and the number of requests reported by the state government more than tripled.Dino's messages and accomplishments are getting through to Maranhenses and there is little doubt that the introduction of governmental transparency obligations and social media have vastly ameliorated the state's transparency ecosystem.
Yet single-case studies such as the above tend to be considered 'outliers' for a reason.Compared to traditional media, social media seems to have been most effective in 'anti moments', such as the anti-corruption crusade of the Ficha Limpa Campaign, the 'anti-system protests' in 2013, or the 'anti-candidate election' won by President Jair Bolsonaro.In most jurisdictions and on most days, however, traditional media dominates quotidian discussions of issues that directly affect citizens' lives.Social media is entertainment for most people, and public-regarding news issues may be commanding relatively little attention.Moreover, the traditional media tends to enjoy widespread dissemination on social media and, on certain platforms, such as Twitter, exercises "vastly more agenda-setting power than other actors do" (HARDER, SEVENANS e VAN AELST, 2017).Therefore, despite the importance of social media, regulatory and revenue-based distortions occurring in the traditional media deserve policy interventions.
Finally, social media itself may be part of the problem of an uninformed or mal-informed public.That is because social media provides the illusion of 'being informed' in several senses.First, researchers have found that social media has inculcated a "newsfinds-me perception", which leads citizens to engage in less active news searches, a phenomenon that leads to lower levels of political knowledge over time (ZÚÑIGA, WEEKS e ARDÈVOL-ABREU, 2017).Second and better known are social media's 'echo chamber' effects.Not only are we being exposed to likeminded thinkers on social media, but algorithms actively curate news to appeal to our preexisting expressed news preferences.In sum, we lack exposure to issues that might be important but with which we are not familiar, and we lack exposure to 'both sides of coin' viewpoints.While social media offers promise, it should be seen not a substitute for traditional media, but rather as a complement.

A Proactive Government
The third critical element of a transparency ecosystem is a proactive government.Proactive is the operative word here on several dimensions.First, proactive means publishing actionable and significant information, beyond the call of what is explicitly required by law.Of critical importance here are agendas and evaluations.Second, government should balance proactive work, such as monitoring, evaluation, and adaptation, with reactive work, such as 'controlling' accounts, processes, or people.Third, proactive governments invest strongly in transparency oversight and promotion initiatives.This section covers each of these points in turn.

Publishing Actionable and Significant Information: Agendas and Evaluations
The transparency obligations of the Brazilian state are vast are growing, embedded in laws as diverse as the Transparency Law, 4 the Freedom of Information (FOI) law, 5 and Regulatory laws on state-owned enterprises, 6 government-funded nonprofits, 7 and the national sports system, 8 among others.Yet the basic legal transparency obligations Brazil's public organizations are relatively straightforward, especially when compared with more sophisticated and demanding FOI laws, such as those of India or Mexico.Apart from 'frequently asked questions', Brazil's FOI law only requires that five other items be actively disclosed, essentially corresponding to, a) expenditures, b) procurement contracts, c) agreements and accords, d) organizational details, such as contact, structure and location, and e) programs and actions.
All of these are important, but one in particular deserves further specification -programs and actions.Here is what government does; an agency's 'programs and actions' should convince citizens that it has a plan, shows results, and is evaluating and learning from its strengths and weaknesses.They should 'sell' the agency's value proposition to the taxpayer and provide space for citizens to engage.From a policy perspective, 'programs and actions' implies policy agendas, formulation, adoption, and implementation processes, and the evaluation and adaptation of policies and processes.
Two of these policy stages are especially critical and deficient -agendas and evaluations.Policy agenda setting and evaluation processes are the alpha and omega of organizational behavior, where an organization decides how to express its mandate and then learns from its performance in order to improve or redirect its efforts.Agendas consist of actions, policies, plans, and goals to be considered or carried out.Evaluations analyze and summarize institutional performance in terms of service-oriented goals (targets), programs, and administration.Agendas and evaluations have several important features in common.Both should be guided by clearly justified principles or criteria.Both should give stakeholders the right to have a say in their development and results.They are the beginning and end of policy cycles, respectively, and both are relevant for understanding goals.Finally, agendas and evaluations are of such importance that they should be featured not only on individual agency websites (e.g.ministries), but also on the websites of information oversight agencies, such as Brazil's CGU.
Much has been written about evaluations, and their theoretical and practical importance should be clear; individuals and organizations require feedback to do better.By comparison, agendas frequently receive less attention even though they determine the path-dependent trajectories of policies or institutions towards success or failure.It is before implementation is undertaken -during the agenda-setting phase -that citizens and collectivities can help guide policy function and design, expressing preferences for the goals and actions an agency prioritizes.An investment in a policy's agenda-setting processes is analogous to an investment in a child's education and health: lacking beneficial inputs, a policy may perform poorly or, worse, society will end up paying for its deficiencies throughout its life-cycle.
Authorities may worry about disclosing agendas for fear of losing control.It is acceptable for executive final decision-making processes to be kept private; transparency can stifle the expression of honest opinions.Yet officials should open-up processes that are pre-decisional (formulation and agenda-setting processes) and post-decisional (what was decided and justifications) to public scrutiny.This recommendation means that agendas should be public and open to input.
In Brazilian institutions, agendas and evaluations are difficult to find, and their transparency varies wildly.For the sake of brevity, the focus here is on agendas alone.Federal institutions are required to have a "Menu of Services" (Carta de Serviços), but these are often difficult to find and say more about citizen services than policy agendas.A positive example of agenda transparency is Brazil's Federal Ministry of Education, which has a National Education Plan, an extensive battery of details on 'programs and actions', and most programs contain details on elaboration processes and plans for implementation.Brazil's Federal Public Prosecutor (Ministério Público, henceforth MPF) and Federal Audit Court (Tribunal de Contas da União), by contrast, do not even provide a section entitled 'programs and actions' (or any variation therein).In the case of the MPF, for example, policy priorities are handled by 'Chambers' (Câmaras, under the heading 'activity ' [atuação]).The 5th Chamber dedicated to "Combatting Corruption", possesses no clear articulation of agendas or activities at all.One is inclined to ask whether clear principles and criteria or the pursuit of news media attention guide agendas.Other Chambers, such as the 3rd Chamber on Consumer Protection, contain administrative performance targets rather than broader service-oriented targets or goals.This appears to be a common motif in Brazil: goals and agendas tend to be focused on administrative goals rather than service-oriented goals.Oftentimes these goals can be overlapping or it is simply a question of framing, other times it seems that the ultimate purpose of government, public service, has been misplaced.

Balancing Proactive v. Reactive Information Production
What is proactive v. reactive is, admittedly, a question of perspective.The perspective here is that one overwhelming feature of Brazilian government is a reactive preoccupation with 'control' (controle) in the form of auditing and prosecuting.The diversity and accretion of controle is such that government adminis-trators complain of committing irregularities (i.e.illegalities) just to get honest work done.But controle is much more than bureaucratic rules, it is an entire complex in the Brazilian state that consumes enormous resources and therefore imposes significant opportunity costs on other, more proactive forms of governance.
Table 1 illustrates data on equivalent institutions in the United States (U.S.) and Brazil, focusing on supreme auditing institutions.California and São Paulo are compared because they are both national economic engines and have populations of similar sizes (40 v. 45 million, respectively).On a per capita ba-sis, Brazil's federal government spends nearly twice what the U.S. does on its supreme auditing institution, the Tribunal de Contas da União (TCU).As a percentage of the total budget, the TCU consumes more than four times the share of its U.S. counterpart.It is true that the TCU audits all parts of Brazil, and not just the federal government.But this function is replicated by state and municipal TCs, auditors, and comptrollers.If the TCU really were a national institution, why would the State of São Paulo's TCE 'need' to spend nearly 8.5 times more per capita than California's Legislative Analyst Office and State Auditor, and consume a share of the budget nearly three times as large as its U.S. comparator?Budget data from respective PDF copies of official budgets for each jurisdiction.
To provide a notion of what these numbers mean in relative terms, the last two columns of Table 1 illustrate budgets for national agencies that support the development of private enterprise, the Department of Commerce in the U.S. and that of the Ministry of Industry, External Commerce and Services, in Brazil.It is important to note that these institutions are engaged primarily in proactive forms of governance.In other words, they are presumably developing and promoting more than they are defending, protecting, or investigating.As the numbers show, Brazil invests nearly as much in the TCU as it does in its ministry of commerce, while the U.S. invests more than 12 times more in its Commerce Department than in its Government Accountability Office.The examples here are illustrative, but are clearly significant in what they say about governance priorities.
The expansiveness and expense of Brazil's accountability complex becomes ever more astounding as one continues to count the ostensibly overlapping controle institutions at the national, state, and occasionally, at the municipal levels.In addition to the auditing courts (TCs), these include auditor generals, comptroller generals, fiscal prosecutors, as well as federal and state prosecutor offices that consume more than three times the budget (R$ 6.73 billion) of the TCU.
The cost-benefit calculus of Brazil's controle complex is rarely broached and research is desperately needed in this area.The logic of this complex appears to have a stronger basis in path-dependent traditions of accommodating political allies and the country's elite with jobs than proven necessities or benefits.Returning to the U.S.-Brazil comparison, if one visits the website of the U.S. Government Accountability Office, the largest heading on the homepage states, "Our Highest Priorities to Help Agencies Save Money and Better Serve the Nation". 9If one proceeds to the link "About the GAO", one encounters a cost-benefit statement in the first paragraph: "we identified $73.9 billion in financial benefits in fiscal year 2017 -a return of about $128 for every $1 invested in us."There is no discourse even remotely comparable on the TCU's website.Here, one is met with a query box headed by the phrase, "what are you looking for?" (o que você procura?).The homepage features campaigns and news headlines with a strong emphasis on controle, such as the program #eufiscalizo (#ipolice), a participatory citizen accountability program (policing, checking) for irregularities in primary and secondary public education. 10The message seems to be that controle is an end in itself, not a public value proposition.
Having argued that Brazil places an inordinate emphasis on reactive forms of governance -controle -what of a proactive government, one that discloses useful, actionable transparent information?Discussions surrounding agendas and evaluations, at the beginning of this section, provided one answer.A second answer is to be had in looking at oversight and the promotion of transparency.

Transparency Oversight and Promotion
Brazil's investment in government transparency is relatively middling.At the federal level, the comptroller general (CGU) has less than half the budget (R$ 1.04 billion) of the TCU.But transparency is only one of the CGU's many concerns.Although the CGU has done a commendable job, the government needs to invest more in transparency.
budgetary concessions for the CGU's mounting transparency responsibilities.From 2004 to 2011, the CGU's budget increased at an annualized average of approximately 17 percent, with inflation averaging 5.3% annually11 during the same period.In 2011, not only did Brazil enact a FOI law but it also joined the Open Government Partnership (OGP), serving as Co-chair alongside the United States.As with the FOI law, the CGU shouldered much of the responsibility for the OGP.One might assume that these two added responsibilities would have increased the CGU's budget after 2011.Yet from 2011 until the onslaught of Brazil's worst ever recession, which began in 2015, annualized budget increases averaged 6.5 percent (2011-15), nearly three times less than the previous seven years.This annualized increase was below inflation, which averaged 7.1 percent 12 during the same period.The CGU's budget is divided among the institutions many divisions, which include 1) the Executive Secretary, 2) Secretary General for Auditing (controle interno), 3) Secretary of Transparency and Corruption Prevention, 4) the Inspector General (corregedorias), 5) The Ombudsman General, and the 6) Secretary for Combatting Corruption.Several of these units are dedicated to activities associated with transparency.For example, one of the Ombudsman General's three subunits resolve FOI appeals.The Ombudsman has also set up a National Network of Ombudsmen, one of whose tasks is to coordinate activities associated with the FOI law.Yet only the Secretary of Transparency and Corruption Prevention is centrally and explicitly dedicated to the FOI law.
In 2017, the FGV's Public Transparency Program conducted interviews in the CGU in order to understand what was being spent on Brazil's FOI Law.The written response staff provided researchers listed piecemeal programs and costs totalling a sum below R$1 million reais.This sum is obviously implausible, both because it does not include fixed costs such as personnel and infrastructure, among others, but also because it cannot conceivably add up all the value that different parts of the CGU contribute to the FOI law.Yet if one takes a broadly estimative approach and divides the annual budget for the CGU into about five parts, 13 one might conjecture that the federal government's transparency obligations consume approximately R$200 million reais.That sum would fall shy of Mexico's National Institute for Access to Information's budget in 2018 (R$ 235 million), and would be especially low in per capita terms -Brazil has approximately 40 percent more people than Mexico.Chile's budget for its Consejo de Transparencia (Transparency Council -2019) equates to R$ 41 million, which is approximately one-fifth of the Brazilian number, but in a country only one-twelfth as populous.In other words, if Chile's population were the same as Brazil's, and its budget proportionate to current spending, total investments in its information commissioner would more than double the Brazilian estimate of R $200 million.In short, Brazil's federal government underinvests in transparency when compared to other regional leaders, and especially when compared to investments in Brazil's reactive forms of governance (controle).
Despite this underinvestment in transparency, the CGU has a very impressive track record.Most important among its accomplishments, the federal government has exhibited high levels of responsiveness in multiple evaluations.Research is needed to look into the quality of responses on a large scale, but on its face, the Brazilian federal government compares favourably with more advanced democracies in terms of FOI performance.The CGU has also encouraged many states and about a quarter of Brazil's municipalities to adhere to Brazil's Transparency Program.This advance is partly due to the Brazil Transparency Scale, an index that gauges the transparency of municipalities and states.Many subnational jurisdictions have adopted the CGU's excellent e-SIC system, which is statistically associated with elevated rates of responsiveness.Furthermore, the CGU has led very productive Open Government Partnership initiatives.These include the passage of new legislation (law 13.460/2017) that provided a legal basis for protecting the identity of FOI requesters.The CGU has implemented this provision, which is a first in Latin America.
13 Given that there are six secretaries, one being centrally concerned with the FOI law and the other, somewhat.Refer to preceding paragraph.
As recent work has emphasized, however, the federal government is not Brazil's major challenge in terms of transparency.That distinction goes to subnational governments and rule of law institutions, especially state courts and public prosecutor offices.Given previous discussions surrounding education and news media independence, it is not difficult to understand why subnational governments score much lower in terms of compliance and implementation.Incentives are very different on the local level.Whereas the federal government is performing for an international audience, local governments play to a captured audience, captured by biased news, clientelistic networks, pacted coalitions of parties, and low levels of interest and public awareness in transparency.
What is to be done about these information deserts?Carrots extended by the CGU or other federal or state bodies are not as compelling as one would hope.Local officials have their own logic; to be transparent is to be vulnerable, and as previous sections have argued, the costs of ignoring transparency obligations tend to be low in the absence of an educated populace or an independent news media.Oversight is the Brazilian FOI law's major oversight; unlike the laws of other countries, law 12.527/2011 does not name one institution as responsible for the law's promotion, regulation, and oversight.Several articles refer to the CGU's role in the appeals process, but resolving appeals is a far call from oversight.In different parts of the country the FOI law is the responsibility of the Tribunais de Contas, or the Ministério Público, or the Comptrollers, or no one's, as each passes the hot potato to the other.The natural guardian of the FOI law is the defender of fundamental rights, the Ministério Público, but its many responsibilities and poor transparency role modeling make it a suboptimal option.Whatever the option it chooses, if Brazil's governments are to do their part in promoting a robust transparency ecosystem, they need to address the question of oversight.As opposed to being another 'reactive' form of governance, an oversight institution responsible for transparency is to free information, making sure that the maximal principle of Brazil's FOI law is followed, where secrecy is the exception and openness the rule.

Conclusion
In answer to the question posed at the beginning of this essay, "does transparency make a difference in Brazil?" the answer is that transparency has always made and will continue to make a difference.It has made a difference in a way that disappoints most Brazilians; although visible and inferable information (i.e.transparency) is the lifeblood of an educated populace, a vibrant media and an effective government, information in Brazil continues to be constricted by impingements.These are frustrating, yet especially superable in the current context.In this context, information and communication technologies have facilitated access to alternative sources of information and, forced by global norms, governments have committed to adopting transparency measures.Although formal governmental commitments to transparency may not translate into real compliance, compliance can be forced by citizens who persist in their informational demands until reaching legal conclusions (DISTELHORST, 2017).The same is to be said for reforms that can help improve Brazil's educational and media policies move forward.This paper has only glossed over a few of the dilemmas faced by transparency ecosystems, dilemmas common to many countries -not just Brazil.Many of these dilemmas are old, such as regulating the news media or nurturing an educated citizenry, but others, such as governmental transparency, are new.Because of the newness of transparency in Brazil and their newness in world historical terms, there is an urgent need to get them right -now.Proactive policymakers should take the plasticity of transparency policies seriously.Just because they are codified in law does not mean they should not be adapted and innovated upon.For example, a privacy law (13.709)was recently enacted in Brazil, and this new law provides for the establishment of a privacy regulator.The intelligent path is to ensure that privacy is handled by the same institution that regulates transparency.This system is in place in nearly all countries with information commissioners, from Mexico to Canada.Privacy and transparency are two sides of the same coin, and therefore must be treated with balance and moderation.If priva-cy and transparency fall under the jurisdiction of potentially competing institutions, decisions are likely to favor one at the cost of another.Knowing Brazil's history, privacy would tend to displace transparency.This conjecture finds basis in the imperiousness of laws and court decisions that, even today, continue to privilege concerns of 'honor' over freedom of expression and information.
The precarious position of information and transparency within governmental structures is at odds with its societal and organizational importance.In this sense, it is not just transparency policies that need adapting, but also our governance structures.Information remains low on the totem pole of government priorities.The clearest indicator, dollar investments, has been discussed at length.An equally important indicator, however, is the allocation of authority.
Greater authority should go to institutions and public servants that manage the public's information.In the case of the Brazilian government, the CGU is clearly in a position of inferiority.For example, the CGU reportedly has trouble obtaining information from the Treasury (Receita Federal) because of claims of domain-authority on issues of 'fiscal secrecy'.Much information withheld by the Treasury is, unsurprisingly, about fiscal exemptions -i.e.rent-seeking.As the CGU and the Treasury both respond to the authority of the President, the Treasury nearly always prevails.In the case of SERPRO, a state-owned enterprise that digitizes government data, the CGU has to pay for access to SERPRO's 'enriched' government information -an absurd but verifiably true fact.That more authority over information is allotted to institutions such as the Brazilian Treasury (Receita Federal) or SERPRO than the guarantor of public information has an element of the absurd.
The question of authority is central to valuing information.Brazil has an urgent need to place greater value on information.Brazil therefore needs to move in the direction of Mexico, India, El Salvador, Chile, and over 30 other countries in establishing an independent information oversight agency that has binding power over other parts of the state.Right now, the CGU's power lies in suasion, which is clearly insufficient.

4
Lei complementar 131/2009; 5 12.527/2011 6 13.303/2016 7 13.019/2014 2018, the newly elected President of Mexico, Andres Manuel López Obrador, gave his inaugural address, decrying the cost of Mexico's National Institute for Transparency (INAI).López Obrador asked what good the INAI had done, since corruption was greater than ever (ZAVALA, 2018).Here is the most basic of fallacies.Transparency should logically increase the salience of corruption, which it has clearly accomplished in the Mexican case.So too in Brazil.The difference, perhaps, is that in Brazil corruption has been prosecuted by Brazil's powerful Ministério Público, whereas the more politically vulnerable Procuraduria de la República in Mexico has been less assertive.

TABLE 1 .
COST OF SUPREME AUDIT INSTITUTIONS AND COST OF MINISTRIES OF COMMERCE IN BRAZIL AND THE USA Sources: population figures for countries from World Bank, for states, IBGE in Brazil and US Census Bureau.