Latin coup datapoint of the day

By Felix Salmon
June 29, 2009
The 1960s saw 12 successful coups in the region; the 1970s saw 9; in the 1980s there were 6; in the 1990s there were 3; and so far this decade there has only been one, in Ecuador in 2000. ... For all the fears about the future of Latin American democracy, the fact is that the region is far more democratic now than it has been at any point in the past, and that even when there are coups, like the present one in Honduras, the military moves quickly to install a new civilian leadership, rather than ruling as a junta. " data-share-img="" data-share="twitter,facebook,linkedin,reddit,google" data-share-count="true">

The Wikipedia list of successful coups d’├ętat is a very useful resource. You can argue the toss on many of the specifics (was Alberto Fujimori’s dissolution of the Peruvian parliament in 1992 a coup?) but the big picture is clear: Latin American coups are increasingly rare things.


The 1960s saw 12 successful coups in the region; the 1970s saw 9; in the 1980s there were 6; in the 1990s there were 3; and so far this decade there has only been one, in Ecuador in 2000. Honduras might double that figure to two, but that misses a more important datapoint — that if you look just at the big three countries in the region — Argentina, Mexico, and Brazil — there hasn’t been a successful coup since 1976, 33 years ago.

It wasn’t that long ago that all Latin American politicians had to carefully mollify the military, lest they suffer a coup. Today, that really isn’t a worry in most Latin countries, even though the military naturally leans right while politics in the region is tending leftward. What’s more, the important countries have fewer coups than the small ones: Bolivia had no fewer than 11 successful coups between 1930 and 1980, only for Haiti to take on the mantle with further coups in 1986, 1988, and 1991. But coups in bigger countries, like Venezuela, have been notable for their lack of success: Chavez failed with his own attempt in 1992, and then successfully rebuffed another ten years later.

For all the fears about the future of Latin American democracy, the fact is that the region is far more democratic now than it has been at any point in the past, and that even when there are coups, like the present one in Honduras, the military moves quickly to install a new civilian leadership, rather than ruling as a junta. That’s the silver lining in today’s cloud. (Full list of datapoints here.)

More From Felix Salmon
Post Felix
The Piketty pessimist
The most expensive lottery ticket in the world
The problems of HFT, Joe Stiglitz edition
Private equity math, Nuveen edition
Five explanations for Greece’s bond yield
9 comments so far

You had me til the last paragraph, Salmon. Installing a puppet suit&tie is no different from having a military uniform as head of state. You gotta live it to believe it.

Anyway, neat numerical breakdown.

Posted by otto | Report as abusive

What Otto Said. Military overthrow of an elected government is =definitionally= a coup, except in the Bizarro World inhabited by the WSJ editorial page and, apparently–strangely–Reuters.

(Though you give away the game when questioning whether Fujimori’s coup can legitimately be described as such. Is this just what happens when we have people who we raised swearing allegiance to “Queen and Country”–in that order–attempt to interpret Democratic Government?)

Is it a “coup” if it’s done to prevent a government from becoming less responsive? It seems like Honduras may have been an immune response by a Latin democracy infected with El Maximo Syndrome. I guess we’ll see what happens.

Posted by Sterling | Report as abusive

Those data points and the subsequent blah-blah are missing one significant point: America was behind or supporting most of those coups, installing gov’ts that were American puppets in the name of fighting Russian advances. That would help explain the drop off as time went on – the countries had already ‘expelled’ the leftist parties and groups. Since the USSR dissolved, excluding Haiti, there have been only Chavez’ failed attempt in 92 and the successful coup in Ecuador in 2000. All this to say, that left to there own devices, there would almost certainly NOT be this plague of military overthrows of gov’t in Latin America.

Posted by the Shah | Report as abusive


The recent time trend in Latin America is as you say – a lower incidence of coups, military take overs, call them what you will. People have over and over again talked about some fundamental shift to democracy in Latin America. But I would submit that I see no fundamental reason for that trend to continue and potentially reverse itself. Certain changes that appear permanent are probably a lot shakier than they appear. The basic political logics that may generate coup support are still present even if latent in much of Latin America.

Saying that coups are a thing of the past is very much like saying that sovereign defaults are a thing of the past. Many an observer are only familiar with recent political history just their friends in the realm of sovereign finance who seemed to forget history time and time again as you have pointed out in your postings on more than one occasion if I recall correctly.

As Taleb might say, it is precisely the situations when we rely on our accumulated experience that we underestimate the possibility of extreme change of the sort that may be on the horizon for Latin American polities. While I am no Latin America expert I don’t recall any sustained public discussion in major outlets about coup risk in Honduras but perhaps I don’t read enough non financial news.

During the Cold War, political scientists thought that the coup and other forms of violent political change were the norm in Africa. Instead of institutionalized processes of political change like elections, there seemed to be institutionalization of violent change. As “democratization” spread more recently some people started to wishfully believe that democracy would take root in a meaningful way. Even if coup frequency declined in Africa that substantively meant nothing about the nature of governance on that continent. Most observers probably would not disagree with such a claim.

The change in many Latin American countries seemed to be of greater magnitude (especially with the reinforcing economic performance as South America left behind the 80s crisis), but the recent shift to the hard left suggests to me that the political climate has changed and there is a greater opportunity structure for a re-radicalization of politics and political change.

Scholars who study coups often argue that legitimacy of the civilian government is key to good civil-military relations and the prevention of coups and other forms of political violence. Sources of legitimacy vary, but pressures from a weakened international economy may make it more difficult for civilian leaders to deliver economic stability and performance. In countries where the military has traditionally played a very important role in domestic politics changes that threaten the military or seem to threaten political stability seem to encourage officers to consider toppling the civilian leadership. Perhaps those might be reasons to expect the unexpected.

Posted by Matthew | Report as abusive

I’m going to assume you’re an academic.

And the snarky curt response to that graph is that the US has been too busy overthrowing governments on other continents to worry about South America much. And, we decided it was easier to ship the majority of our third world production to South Asia and keep whatever problems arise nicely sequestered by the world’s largest ocean.

Not to say that South America isn’t our backup plan.

Posted by mark | Report as abusive

This wasn’t a coup. In Honduras, as in many countries (see article 272 of the Honduran Constitution and article 142 of the Constitution of Brazil), it is the Armed Forces job to guarantee the obedience of the constitution and any one of the three branches of government can order it to do so. The President is only the commander-in-chief if he obeys the Constitution.

The Supreme Court of Honduras simply obeyed the Constitution (Articles 239, 184, 185, 186) and ordered the army to arrest and expel president Zelaya.

Zelaya is the one who was trying to give a coup. He disobeyed a Supreme Court ruling and was planning to go ahead with an illegal referendum on Sunday. He was trying to extend his term limits, which is forbidden by the Constitution.

I applaud the Honduran democracy! I wish that if one day a crazy president gets elected in my country, Brazil, our institutions work just as perfectly to get rid of him and defend democracy and the Constitution.

Posted by Souza | Report as abusive

Re Academic: I presume verbiage & reference to “experts” gave me away.

I agree that post Cold War US policy shift in S Amer. is responsible for some of what we see now. I don’t share your conclusion about the “back up plan” but agree that if US foreign policy behavior changes – we may have a repeat of prior experiences. The left-govt bloc along with the Colombia mess could easily generate a set of excuses for wild interventionism.

Posted by Matthew | Report as abusive

I agree with Souza on this. How is it not a coup when the current President is attempting to formulate pseudo dictatorial powers by defying the constitution of said country?

If the legislature and SC of that country declared the actions illegal per their constitution and that president decided to go forward with them (reminds me of a certain US President and his penchant for signing statements on legislation through which he essentially said he wasn’t going to follow the bill he was signing into law) is that not in and of itself a coup?

Posted by Steve | Report as abusive
Post Your Comment

We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see