Transparency and Historical Reconciliation were Campaign Promises
When Brazilian President Dilma Rousseff promulgated the country’s new freedom of information law on November 18th, she signed on to a measure that is among the strongest in Latin America, and perhaps more importantly, she endorsed a law that made inordinate improvements over bill (5228) introduced by her predecessor, President Luiz Inácio Lula da Silva, in 2009.
Laws, Decrees, and Pending FOI Initiatives across the Region
As the blurb at the side of my blog describes, I am in the midst of writing a book on the adoption of freedom of information laws across Latin America. Part of the study requires to me to ‘score’ the strength of freedom of information laws across the region. For my dissertation, I originally scored the bill 5228, introduced by the administration of President Lula da Silva in May 2009. That bill earned a 1.8 out of a possible score of 3, qualifying it as ‘moderately weak’ in the benchmarks established by my index.
By contrast, I recently finished scoring the newly enacted law, 12.527, and it earned a 2.4, qualifying it as ‘moderately strong.’ Although the bill has some significant weak points, it mostly reflects the trend towards internationally accepted better practice standards. Brazil’s achievement deserves a caveat: the laws now coming out generally tend to be legally stronger than older laws because they reflect more evolved standards and stronger pressures to join the fray of standard-bearing adopters. This is certainly not a rule, but I can marshall good empirical evidence to make the case.
Brazil is the 10th country to adopt a modern, comprehensive freedom of information law in Latin America. Six other countries have yet to adopt such a measure. These include Argentina and Bolivia, which possess presidential decrees establishing FOI in the Executive branch; and Colombia and Costa Rica, which also possess certain administrative allowances for disclosure.  Until Brazil enacted legislation on the 18th of November, it stood among the ranks of Paraguay and Venezuela, countries that possessed no generalized disclosure mechanisms of note.
In this Long Post
The current post analyzes six aspects of the Brazilian law, characteristics by which all freedom of information laws are measured. I examine the Brazilian law’s 1) scope, 2) procedures, 3) appeal system, 4) exceptions to disclosure and classification system, 5) the duties of government agencies to promote the law and publish information, and 6) the sanctions and protections for officials administrating access to information requests.
Because Brazil’s law regulates the country’s constitutional provisions for freedom of information (articles 5, 37, and 216), it regiments all branches and levels of government: the Executive, Legislative, and Judicial branches of government, and Federal, State, and Local governments. Positively, it also regiments non-profits receiving government funds and state-owned-enterprises. The law’s scope is just about as wide as it can get. This clear breadth is a vast improvement over the bill Lula introduced in 2009. That measure did not even define what government entities would be regulated by the law.
While the breadth of the law is impressive, it is also impressively daunting. Mexico discovered the scale of this challenge shortly after amending its constitution in 2007 to guarantee the right of universal access to information. In Brazil, it remains to be seen whether the territorial units will write their own disclosure policies, and what standards will regulate them. In the case of Mexico, there is an established ‘minimum standard’ for subnational units to follow. No such standard has yet been discussed in Brazil.
Questions of implementation and regulation aside for the moment, the law adopts a strong statement of openness in Article 3, referring to the “fundamental right of access to information,” and establishing the famous rule: “publicity as the general precept, and secrecy as the exception.”
The most significant problem with Brazil’s law is not the scope, but the oversight of its ambitious scope. Unlike other countries in the region, such as Mexico, Chile, or even Honduras, the Brazilian law provides no agency to define the implementation, regulation, and enforcement of the law. Even though it is widely assumed that the Comptroller General will oversee the law, the text does not specify responsibility. It merely states (Art. 41) that the President will appoint the oversight agency. As it stands, the Comptroller is only responsible for deciding appeals. The regulatory vacuum that the law implies is thus its most obvious shortcoming, and measures should be taken to establish a well-endowed, independent central authority. The Comptroller is inappropriate to the task; its multiple attributions and dubious independence from the presidency make it a sub-optimal oversight mechanism.
The law’s procedures (Articles 10,11, 12, 13, and 14) are clear, user-friendly, and fall within international better standards. Officials have 20 days to respond to a request, and they may extend this timeframe by 10 days if they alert the requester in writing. Officials also have obligations to help requesters file requests, convey misplaced requests to the appropriate entity, and inform requesters of their options for appeal in the case of a denial. The procedures also stipulate that denials must be justified (Art.11); fees can only be levied on the materials used for photocopying; the legally-defined poor are exempt from paying any costs; and requesters can receive responses in various forms. Perhaps the law’s most innovative feature is its emphasis on formats that are open, machine-readable, and non-proprietary (Art. 8 VI:3 ii & iii). This provision will help technologists, data-based journalists, and skilled citizens to re-use public information for analysis or applications with greater ease.
Promotion and Duty to Publish
The law obligates governmental agencies to actively publish a wide assortment of information, independent of requests (Art. 8). Active transparency obligations include official contact details of all employees, financial operations, spending, procurement contracts, and answers to commonly asked questions, among others. Unfortunately, the law specifies no agency for promoting the law, as discussed at the end of the section on ‘Scope.’ The President must appoint this agency (Art. 41), who will annually present a report of activities to Congress, as is better practice. A key point is that the heads of all government entities are expected to specify their information officers within 60 days of the law’s enactment. I have already talked to my colleague Fabiano Angélico about this provision, and beginning on January 18th 2012 we are determined to find out just how successful the process of assigning information officers has been.
Exceptions to Disclosure and Classification
The law provides the standard battery of exceptions to disclosure and also provides for a severability clause, meaning that citizens can have access to classified documents, albeit with sensitive portions ‘blacked-out.’ One standard exception is missing from a list that is otherwise in fine condition: ongoing legal cases. These cases are typically withheld from the public until concluded. Exceptions to disclosure could be strengthened by balancing tests, to see whether harm caused by information justifies withholding, or whether the public interest of possessing the information in question outweighs the public interest of maintaining it outside the realm of public knowledge. As with most other freedom of information laws, the measure does not override the secrecy statutes of other laws, i.e. subjecting the secrecy provisions of other laws to the freedom of information law’s list of exceptions. This is unfortunate, but lamentably the norm.
Finally, the most glaring deficiency of the law’s classification scheme is its three tiered secrecy provisions and its deviously long reserve periods. ‘Ultra-secret’ information is withheld for 25 years before being subject to a re-classification or de-classification process. The other two tiers are ‘secret’ (15 years) and ‘reserved’ (5 years). Twenty-five years is an outlandishly long reserve period, and three tiers of secrecy are wholly unnecessary. Chile’s reserve period stipulates 5 years, Mexico’s 12, and the highest in the region –before the Brazilian law was enacted – was Uruguay’s at 15 years. There is no rhyme or reason for information to be denied a re-classification process after 5-10 years.
Appeals and Complaints
The appeals and complaints system promises Kafkaesque journeys into the dark recesses of the Brazilian bureaucracy. Well, perhaps not that bad. Positively, requesters may ask for a written explanation when access to information is denied (Art. 14). Having this justification should greatly expedite the appeals process, as the need for a pre-ruling investigation should be all but obviated.
One problematic aspect of the appeal regime are the timeframes: requesters can lodge an appeal within 10 days, which is all fair and well, but decisions must be rendered within a very demanding 5 days. This period of time is hopeful, but unrealistic. The likelihood is that public sector entities will constantly transgress this timeframe, which may leaves citizens cynical and upset, and detract from the integrity of the law. Once a public sector surpasses the allotted timeframe, all bets are off as to when they may be disposed to resolve an appeal. Thus as my colleague Andrew Ecclestone points out in one of this post’s comments, it would have been better to set a more realistic timeframe for resolving appeals.
Petitioners must obligatorily appeal to the “hierarchically superior” official, as a first recourse. Failing a positive response, the requester can then lodge an appeal with the Comptroller General, which again, must be answered within 5 days. If that fails, the requester may appeal to the ‘Mixed Commission on Information Re-Validation.” No timeframe for a decision is given. The law makes no mention of recourse to the courts, and I am uncertain as to whether the courts will accept a case once the Comptroller has denied an appeal. This would seem to be the logical progression, especially given that the Comptroller should only have constitutional say over requests to the Executive Branch. I am interested to figure out how the courts fit into the appeal process—so if you have any ideas, please let me know.
All told, however, the appeals process does not look promising. Internal appeals are notoriously ineffective, and appeals to an institution that lacks the attributes of a strong oversight mechanism are not ideal. Finally, there is no established timeframe for the third and final appeal process.
Sanctions and Protections
Finally, the Brazilian law scores reasonably well on sanctions and protections. Article 32 provides a comprehensive list of offences for which officials can be punished, such as hiding or destroying information. The list even includes “using bad faith when analyzing access to information requests.” The punishments are dictated by a host of legislation on administrative conduct. Unfortunately, the law does not provide whistle-blower protections or protections for those who disclose sensitive information in good faith. But again, this shortcoming is so common as to be the norm among the region’s laws.
So there it is. There is much more, but these are some of the more important points. The law is much better than the original bill (5228) introduced into Congress by Lula in 2009. But one of the problems is that Lula set the bar so low that it was difficult to make that paradigm jump to a really exceptional law. As it stands, the law is good; it’s strong on its breadth, its duties to publish, and its exceptions from disclosure and sanctions and protections are mostly in order. Where the law really falls is on the issue of oversight – there is no central oversight agency to promote, implement, regulate, enforce, and protect the law and the disclosure process. It also comes up really short on classification: a three tiered secrecy scheme with a ghastly 25 year reserve limit.
Clearly, the surrender of secrecy will be an ongoing struggle in Brazil. The strong pull towards obfuscation and impunity gives reason to question whether Brazil can make good on the law. But here’s a new beginning. The question now is whether the enormity of implementation can be achieved in 6 measly months. According to a recent article, top officials from the Comptroller General’s Office are asking the same question.
 For more information on the scoring, the index, and benchmarks, please see the appendix to my dissertation.
 Colombia possesses a law from 1985 that can best be described as ‘skeletal’ and does not even include any reference to appealing denied information. It is therefore not ‘modern.’ Costa Rica guarantees freedom of information through an assortment of administrative statutes, but still has no comprehensive law.