The Brazilian impeachment (constitutional provisions)

As almost anyone who would read this blog surely knows, the Brazilian Chamber of Deputies voted to impeach President Dilma Rousseff on 17 April. The case now moves to the Senate, which could try and remove her if it concurs with the charges.

The impeachment process in Brazil is similar to that of the USA, but differs in very important detail. First it takes a two-thirds vote in the first chamber, unlike the US where the House of Representatives impeaches (brings formal charges) against the President by majority. In both countries, it takes two thirds of the second chamber (senate) to convict and remove.

The key provisions of the Brazilian constitution are from Article 51:

It is exclusively the competence of the Chamber of Deputies… to authorize, by two-thirds of its members, legal proceeding to be initiated against the President and the Vice-President of the Republic …

And Article 52:

It is exclusively the competence of the Federal Senate… to effect the legal proceeding and trial of the president and vice-president of the republic for crime of malversation…
Sole paragraph. in the cases provided for in items i and ii, the chief Justice of the supreme federal court shall act as president and the sentence, which may only be issued by two-thirds of the votes of the federal senate, shall be limited to the loss of of ce with disquali cation to hold any public of ce for a period of eight years, without prejudice to other applicable judicial sanctions.

Another nontrivial different from US procedure is that the president of Brazil can be suspended from office upon impeachment, even before the Senate trial has begun. The conditions for impeachment and trial are further laid out in the section of the constitution concerned with executive authority.

Article 85:

Those acts of the president of the republic which attempt on the federal Constitution and especially on the following, are crimes of malversation:
I – the existence of the Union;
ii – the free exercise of the legislative power, the Judicial power, the public Prosecution and the constitutional Powers of the units of the Federation;
III – the exercise of political, individual and social rights; IV – the internal security of the country;
V – probity in the administration;
VI – the budgetary law;
vii – compliance with the laws and with court decisions.
Sole paragraph. These crimes shall be de ned in a special law, which shall establish the rules of procedure and trial.

I have not seen the law referenced in that “sole paragraph”, so I do not know what further conditions or elaborations may be stated therein. I also have not seen the bill of impeachment itself, so I do not know if it cites on of the above causes. However, the reporting on the impeachment generally emphasizes charges over manipulation of accounts. More specifically (in Portuguese) that she did not respect the law of fiscal responsibility.

Article 86:

If charges against the president of the republic are accepted by two- thirds of the chamber of deputies, he shall be submitted to trial before the supreme federal court for common criminal offenses or before the federal senate for crimes of malversation.
Paragraph 1. The President shall be suspended from his functions:
ii – in the event of crimes of malversation, after the proceeding is instituted by the federal senate.

I welcome discussion from those following the case closely regarding the process, and evaluations of the charges themselves.

17 thoughts on “The Brazilian impeachment (constitutional provisions)

  1. Pingback: Brazil: Early elections instead? | Fruits and Votes

  2. Hi Matthew,

    Sorry to leave a reply on an unrelated topic but I couldn’t find an email for you. I’d be interested to hear your thoughts on the South Korean parliamentary elections last week. I know that FPTP heavily distorts the correlation between votes and representation but the figures are astonishing.

    The ruling Saenuri party – panned in the international media as having been defeated – lost on seats but won 1.8m more votes than the Minjoo party that technically finished first. Furthermore Minjoo didn’t even finish second in terms of votes. They were third! A whole 300,000 votes behind the People’s Party which took home a paltry 38 seats. Surely this is one of the biggest distortions in FPTP history in an established democracy?

    I’ve not seen anything online about electoral reform in South Korea but I’m reminded somewhat of the 1981 Maltese election where the ruling party took less votes but still had a majority of seats and provoked a six-year long constitutional crisis. The system was later rebalanced. I don’t see any such outpouring of anger though in S.Korea?


    • Matt, the Korean result was unusually distorted only if you only consider the list votes. But Korea uses two ballots, and the FPTP district vote totals were very different and much more in line with the seat results. There is a plurality reversal, but just barely, and without a majority of seats. Perhaps the really unusual thing is the amount of vote splitting.

      The full results:
      Minjoo – 37.0%, 110 seats
      Saenuri – 38.3%, 105
      People’s – 14.9%, 25
      Justice – 1.6%, 2
      Independents – 7%, 11

      Minjoo – 25.5%, 13 seats
      Saenuri – 33.5%, 17
      People’s – 26.7%, 13
      Justice – 7.2%, 4

      Total seats
      Minjoo – 123 seats (41%)
      Saenuri – 122 (40.7%)
      People’s – 38 (12.7%)
      Justice – 6 (2%)
      Independents – 11 (3.7%)


    • Hi Matt. I’m afraid I’m not Shugart, but I’ll try and answer your question.

      South Korea uses a mixed-member majoritarian electoral system. About 50 seats are elected using party-list proportional representation, and the remaining 250 are elected in single member districts using first-past-the-post. Different ballot papers are used for each segment.

      The vote totals you quote are from the national party-list contest. In this contest, Minjoo won 13 seats, the People’s Party won 13, and Saenuri won 15.

      However, in the single-member districts, the vote totals were quite different. Wikipedia says Saenuri won 38% to 37% for Minjoo, and the People’s Party only won 15%. In terms of district seats, Minjoo won 110, Saenuri 105, and People’s 25. While there is clearly some disproportionality there, it does not seem too dramatic to me by the standards of FPTP.


  3. The risks of copying and pasting text from a PDF – “fi” and fl” show up as blanks.
    I must say, that seems a very shaky translation. Though I do like the term “malversation”.


    • I’ve always found that the dividing lines between MMP and MMM to be a little narrow and confusing. However, I could certainly identify some methods that have been used that are more proportional than ordinary MMM, but not MMP.

      Hungary’s variant of MMM, which subtracts the majority for the victorious candidate in a constituency (defined as the vote for the second-place candidate, plus one) from the national vote total for that party in order to allocate party list seats. Had these votes not been subtracted at the last Hungarian election, Fidesz would have been about seven seats better off, so it’s clearly not that proportional, but it makes a difference.

      Italy used a similar system, except that all votes cast for the victorious candidate were subtracted from the party list of that victorious candidate. This worked quite well until the two major parties affiliated their candidates to decoy lists, which made it virtually the same as mixed-member majoritarian.


      • Re MMP: I like the Hungarian idea. Kind of like a Vickrey auction, where the highest bidder only has to pay $1 more than the second-highest bid (no matter how much higher the former actually bid).
        To eliminate decoys in MMP, I’d go for:
        1. If the winning district candidate was endorsed by, and/or appears on the list of, an over-5% party.
        2. If not — the runner-up candidates have the right to specify, in writing, within (say) 72 hours after the result is declared, which over-5% list they believe the winning district candidate should be “billed to.” If they disagree, the list nominated by the defeated candidates with the most votes combined.
        So if (eg) independent Tony Windsor were to win a district with, say, 43%, and the National and Liberal candidates (with 17% and 10% respectively) were to certify that he should be counted as Labor for the purposes of list-seat allocation, while the Labor candidate (with 20%) insisted that Windsor is actually in the Liberal or National column, the former would prevail.


      • Well, yes, Miguel, but that would still make it mixed-member majoritarian. I think Rob was referring to compensatory mechanisms that could be added to MMM to make it more proportional.


  4. On the Hungarian system, I devised something similar, but structured it so it could work without political parties.

    Essentially, most deputies are elected by single member districts by plurality, while some are elected from larger multi-member districts or regions, also by plurality. Each candidate running in one of the smaller single member districts is listed on the ballot along with a running mate candidate for its regional multi-member district. A candidate for one of the multi-member districts would have several running mates, potentially one for each single member district it encompasses.

    A vote for a ticket of a candidate for a single member district and a candidate for the regional tier gets counted for the single member candidate if he/ she wins, up to the number of votes for the strongest runner up plus one (the votes needed to elect). Otherwise it gets counted for the regional candidate. Essentially it works the same way as in Hungary, a party running in the multi-member tier will have most of the votes for its successful single-member candidates deducted.

    By technically not requiring parties to work, the entire question of fake parties that bedevils these sorts of systems is sidestepped. It also allows for independent candidates. It also echoes the old English tradition of having borough MPs and county MPs.

    My views on proportionality have changed. I think the goal should be “proportional enough”, without seeking much of correlation between votes and seats. By “proportional enough” the system should allow for minor parties with substantial support to win seats, and makes legislative majorities very unlikely for parties that get less than 45% of the vote. Once you achieve those goals, you can focus on what else you want the system to accomplish without having to worry about proportionality. This leads to hybrid mixed member systems such as Hungary or South Korea. I would call the South Korean system mixed member non-proportional (MMNP), or single member plurality plus (SMPP).

    I will put my Brazil comments in another thread.


    • Ed: Your system is certainly mixed-member; however, I’m not sure exactly how proportional it would be, especially in a small district magnitude. My preference for preventing overhang, if that is considered desirable, is the German system of expanding the legislature to make any overhang proportional relative to the rest of the legislature; the Scottish system can incentivise tactical voting on the party-list ballot, which I consider undesirable, and the German system will discourage tricks like decoy lists enough to outweigh the potential disadvantage of a large legislature.

      Also, to be pedantic, the term normally used for South Korea’s system of unlinked SMDs and party-list PR is mixed-member majoritarian.

      Rob: I can only assume one vote is needed for Ed’s hypothetical system.


    • Ed, you say that is a model without political parties, but as soon as you have teams of individuals and an allocation rule that takes said teams into account, you have parties–whatever you choose to call them.


  5. What about Mexico’s MMM system, isn’t that a bit more proportionate than say the Japanese MMM system? It sounds as if South Korea wants to give the 3rd largest party representation, but not alot, but only wants a 2 and a half party system and not an atomized party system similar to the Netherlands and/or Finland.


    • Mexico’s system achieves a partial compensation via caps on how over-represented a party can be. Hungary’s does so by adjusting votes in the list tier according to outcomes in the nominal. But both allocate the seats in parallel once the caps or vote-adjustments are established. They do not first set an overall target for seats, combining nominal and list tiers, before determining how many of the latter a party obtains, as is required for ‘MMP’ as I understand the term.


  6. Pingback: Brazil: President ousted by Senate | Fruits and Votes

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.