Brazil: Lessons Learned
Last updated on December 17, 2013
Key findings and lessons learned from Brazil’s experience are summarized here. They are organized according to the key issues and considerations outlined in the Overview of this guidebook.
Although Congress formally creates the rules governing elections, the TSE is by far the most powerful actor in designing legislation governing elections. Usually when Congress has passed legislation contrary to the preferences of the TSE, the TSE has successfully convinced Congress to repeal the legislation or convinced the Supreme Court to suspend it.
The institutional structure of election management in Brazil makes it difficult for external actors to independently influence and evaluate the use electronic voting. This stems from the fact that the TSE both implements elections and adjudicates electoral disputes. This arrangement creates a clear conflict of interest, since the TSE’s own actions are often involved in any disputes involving election technology. This problem is further exacerbated by the fact that the only judicial body higher than the TSE, the Supreme Court, is partly composed of ministers of the TSE. As a result of this institutional architecture, it is virtually impossible for outside actors to successfully challenge decisions made by the TSE through the legal system.
While the TSE has taken steps to make electronic voting accountable, these steps have not completely addressed issues of accountability.
Robust forms of external auditing and evaluations are not provided. Opportunities to examine the source code or other aspects of the system are highly controlled and, given the complexity of the system, insufficient time is given for adequate vetting of the code and related systems.
There is no practical way for political parties or candidates to dispute election outcomes, primarily due to the lack of VVPAT. Despite repeated attempts of congressional actors to modify the system to include a VVPAT, the TSE has successfully resisted such changes.
Given these factors, some stakeholders have pointed out that greater access for non-governmental actors to examine or audit source codes would be beneficial for the election process in Brazil, and would enhance accountability of electronic voting.
Security and Secrecy
In comparison to the paper ballot system, where fraud was relatively widespread, electronic voting has substantially improved the integrity of the vote count. The vast majority of electorate and political elites view the system as reliable and trustworthy, although there are some exceptions, particularly in the academic community.
However, the limits placed by the TSE on full audits of the source code, equipment, and election outcomes breed distrust amongst academics and civil society groups interested in government transparency.
Critics of the system have pointed out several potential flaws with the encryption and software verification mechanisms, but the TSE rarely responds to these criticisms directly, which lowers trust in the system among interested parties.
Most of the TSE’s security efforts are aimed at protecting against an external attacker. Critics of the system argue that an internal attacker is also possible and that the TSE has not adequately described safeguards against such an attack.
The voter verification system is linked to the voting machine, which is against international best practices. Congress attempted to sever this link through a change in the law, but the TSE succeeded in convincing the Supreme Court to suspend the law. The TSE argues the link is necessary to prevent voters from voting multiple times.
While the TSE states it is transparent during some parts of the electoral process, this is not always sufficient in meeting international best practices and gaining the trust and confidence of key stakeholders.
In some cases, transparency was restricted because of sensitivity and secrecy of information, particularly with regard to access source code. When external parties have access to the source code, basic tools used to test and search the code are not permitted.
Most importantly, no independent observers are allowed to observe the electoral processes in Brazil. Civil society organizations have to work through political parties to gain accreditation as observers. Only the OAB has access as an independent organization, but it claims it does not have sufficient expertise to assess the processes fully.
In the experience of Brazil, there are no challenges and recounts carried out, as there is no VVPAT. Even though the trust of citizens and political parties in the system is very high, this does not guarantee a fully accountable process.
The vast majority of non-governmental interviewees recognize the need of VVPAT for Brazil elections and there have been legislative attempts to introduce VVPAT.
Since 1996, Brazil began implementation of electronic voting in a staged manner. Beginning in 2000, all elections have been fully electronic and EVMs are accepted by all stakeholders.
Since 1996, the country has built over 16 years of experience in electronic voting and, as a result, has very few problems associated with EVMs. Error rates are very low; although some modifications were made, the same basic system adopted in 1996 is still in use.
The development of electronic voting in Brazil was somewhat inclusive in 1996, and involved input from technology experts and vendors; it still lacks sufficient input from non-State actors.
The voter education in Brazil seems to be effective, as polls show that voters have sufficient levels of information on the electoral system and on usage of electronic voting machines, and electronic voting retains widespread support among Brazilian voters.
Electronic voting has greatly improved inclusiveness for low literacy voters. After the adoption of electronic voting, the fraction of blank and invalid votes has dropped dramatically, as the new system has proven easier to use than the paper ballot system.
Most parties and voters trust the electronic voting system and rate it highly, particularly the fact that it produces results quickly and reduces uncertainty over election outcomes.
A few political parties, civil society activists and members of academia view the TSE as too closed and unaccountable. This distrust stems mainly from the fact that attempted reforms to introduce voter verifiability have been blocked by the TSE. Highly-restricted access to the electronic voting source code also contributes to this mistrust.
A range of stakeholders, including civil society activists and academics, contend the TSE would be well-served to open up the electronic voting process for further auditability, and explore ways VVPAT can be introduced in a cost-effective manner.
Case Study Report on Electronic Voting in the Netherlands: Background