Flexibilisation, parliamentarisation and presidential breakdowns: lessons for Brazil

From time to time, or actually rather seldom, ones research may become relevant for real political events. With the impeachment threat looming over President Dilma Rouseff, some observers have revisited some of the arguments I made with Einar Berntzen in our article in Comparative Politics in 2008. In that article we study and categorise the many presidential breakdowns in Latin America.

Taking as point of departure Juan Linz's thesis that presidential regimes are not as well equipped as parliamentary regimes to tackle political conflicts, we argue that the presidential breakdowns rather show that presidential regimes in solving their crises seem to have overcome its deficiensies. The perils of presidentialism according to Linz were rigidity due to the fixed terms of parliament and president, and dual democratic legitimacy since both parliament and president enjoy direct democratic legitimacy. These problems make it difficult to solve conflicts between parliament and president since it is difficult to remove either institution democratically, and there is no institution that by definition decides over the other. In parliamentarism conflicts are solved by parliament (or its majority) having the power to nominate and remove the government, and the government normally, and within some set limits, has the power to call for early elections.

The article then argues that the removal of presidents before the end of the electoral term demonstrates that the fixed terms are not as fixed as we believed, and that instead of ending in democratic breakdowns, which Linz predicted, conflicts in presidential regimes in Latin America now end in presidential breakdown with democracy surviving. We further argued that some of these presidential falls resemble parliamentary modes of conflict resolution. Whereas we argued that impeachment was a presidential form of conflict resolution, since impeachment of the president is constitutionalised in most presidential regimes in Latin America, we said that other forms of parliament voting out the president resembled votes of no confidence. Types of this could be votes of incapacity, or abandonment of the presidency. The cases were here two in Ecuador (1997 and 2005), and one in Bolivia (in 2005), the latter being an example of a president losing a vote of confidence which led to presidential resignation, in addition to failed attempts such as in Peru in 1991. We further pointed to another aspect, which was the call for early elections during times of crisis as another sign of choosing parliamentary solutions to presidential crises. This latter aspect has often gone unnoticed. We counted (back in 2008) early elections in Bolivia (twice), Dominican Republic, Argentina and Peru. In the article Einar Berntzen and I interpreted rather positively these presidential breakdowns (or interrupted presidencies as we call them) since they did not end in democratic breakdown, and we argued that the flexible solutions (not always envisioned constitutionally at the time) had helped avoid full democratic breakdown.

After our article, the use of impeachment has gone from only dealing with unconstitutional and undemocratic behaviour (Raúl Cubas in Paraguay in 1999) and corruption, to seemingly deal with more strictly political matters in the case of President Lugo in 2012, and Dilma Rouseff in Brazil today. With Mariana Llanos and Detlef Nolte, we analysed the fall of President Lugo highlighting the aspect of the politics of impeachment. In this case Lugo did not fall because he was corrupt, but because 2/3 of Congress argued he did not perform his duties well. Using political motives to spur an impeachment process could then also make impeachment procedures resemble a type of vote of no confidence used in parliamentary regimes, and one could argue that impeachment was used "parliamentary-style". In Brazil today, it also seems that impeachment is much more politically motivated than in the case of for instance Collor in 1992 (a case of corruption), and therefore many have argued that impeachment is not warranted in this case. Again we see impeachment used as a vote of no confidence, but with a 2/3 bar rather than 1/2.

Pérez-Liñán revisited the parliamentarisation arguments in an article in La Nación where he pointed out, correctly, that an impeachment is not a vote of no confidence, and this is also one reason the current efforts to impeach Dilma have been criticised. Originally, as I mentioned, we never thought of impeachment as a form of vote of no confidence, or parliamentarisation, but in the cases of Lugo and Dilma impeachment resembles votes of no-confidence. As such these two cases may strengthen our original argument that presidential regimes copy parliamentary regimes when solving conflicts, even though impeachment as vote of no-confidence may be harder to defend normatively for the very reasons Pérez-Liñán mentions (but is still better than our the alternative outcomes provisioned in the literature we discussed in our article: democratic breakdown). At any rate it is interesting to observe that presidentialism as we thought it, is evolving even though we may disagree with its direction. (For pragmatic reasons, however, I find that if a 2/3 majority in both chambers are willing to remove the president I think they should be allowed to do so no matter the reason. This is simply because if 2/3 of the representatives want the president removed, the president is most likely not able to govern anyway and her/his survival may then secure un-governability. A new president, all else equal, should have a higher probability of obtaining congressional majorities). 

But, if we are to revisit the flexibilisation or parliamentarisation argument for the current crisis in Brazil, I much rather suggest considering one of the parliamentary formula we actually discussed: that of holding early elections (for congress and president). Early elections, however, is not envisioned constitutionally, and requires constitutional reform. Further, parliamentarians are not known for voting in favour of something that will leave them without (a well-paid) job. The solution is therefore maybe unlikely, but one could promise parliamentarians immunity in exchange for early elections, which would increase parliamentarians' incentives. Early elections instead of just impeaching Dilma has the benefit of sharing the blame equally between parliament and president which seems on the face of it more just in the case at hand (because in Brazil no one is without sin even though everyone casts stones). Several observers, including The Economist, have already suggested it, and it seems a somewhat more just solution than putting all blame on Dilma. 

Another criticism of the impeachment solution is that the removal of the president in Dilma's case will not solve much. I agree. This has to do with the type of crisis in Brazil. Should Dilma fall the lava jato scandal and investigation will still be there, and more importantly the poor economic performance will not automatically go away with Dilma. In another article on the topic I point out that presidential breakdown - the early removal of the president - seemed to have worked well (that is stabilising the political situation and avoiding more breakdowns) when the crisis is related mainly to the person inhabiting the presidency (for instance corruption in the case of Collor). Then you "solve" the problem by removing the president, and people protesting in the streets go home. Presidential breakdown also seems to have worked well when the president has been of the semi-authoritarian type such as Fujimori in Peru or Balaguer in the Dominican Republic. Again the motive for people going into the streets is removed with the removal of the president, and in these cases removing the president also holds the benefit of democratising the country (Venezuela today could become a similar case). However, when the problems have been political and/or economic, or both, such as in Ecuador and Bolivia in the early 2000s, and in Brazil today, then removing the president does not solve all that much (even though democratic breakdown has been avoided). The underlying problems persist, people will come out in the streets again, and often instability continues (think Bolivia and Ecuador in the years before Morales and Correa took over). Things tend to stabilise first after a new president implementing a policy switch have come into power. This occurred right away in Bolivia in the 1980s with Paz Estenssoro and also in Argentina with Menem (no more presidential falls), but took some time and some more presidential falls in Bolivia and Ecuador in the 2000s.

Taking lessons from these more politically motivated presidential breakdowns we find that a potential successor to Dilma may quickly also be in peril unless he manages to drastically change the political and economic course of the country in a manner that sufficiently satisfies people and encourages them to home rather than taking to the streets.

Some of my articles on the topic

Marsteintredet, Leiv (2014). Explaining variation of executive instability in presidential regimes: Presidential interruptions in Latin America.  International Political Science Review.  ISSN 0192-5121.  35(2), s 173- 194 . 

Marsteintredet, Leiv, Mariana Llanos & Detlef Nolte. 2013. Paraguay and the Politics of Impeachment. In Journal of Democracy 24(4), October (pp.110-123).

Marsteintredet, Leiv & Einar Berntzen (2008). Reducing the Perils of Presidentialism in Latin America through Presidential Interruptions. In Comparative Politics 41(1), October (pp. 83-101).