More on Qadaffi’s Death: Violent Leader Removal Increases Likelihood of Democratization

by Joshua Tucker on October 21, 2011 · 9 comments

in Comparative Politics,Transitions

In response to my request for research on the effect of the death of dictators on the future prospects of the country in question, Michael Miller of the Australian National University sent along the following comments:

You pose some very interesting and timely questions related to Qaddafi’s violent ouster and what this implies for Libya’s democratic prospects. I have some research here directly on this question.

The gist is this: On average, the violent removal of an autocrat (whether by coup, rebellion, assassination, threat, or foreign assistance—it doesn’t seem to make much difference) makes it three times more likely that a country will democratize in the immediate future. About half of democratic transitions occur within five years of a violent ouster, and another quarter after a peaceful turnover between autocrats. Hence, there’s a big association between an autocratic leader leaving office and autocracy ending. I argue the main reason is that violence indicates and contributes to regime weakness. The periods of chaos following violence, when elites are divided and citizens are engaged, provide the best possible openings for democratic actors to make their demands.

But here’s the rub: That opening only matters if there exist democrats in the country and they have sufficient support and power to win over the next wave of opportunistic autocrats. For this reason, I find that the aftermath of violence is when socioeconomic conditions matter the most for predicting democratization. In particular, average income predicts
democratization if and only if there’s been a recent violent turnover. In other words, violence shakes up the system, but what you get out of it is largely a product of structure.

What does this say for Libya’s democratic prospects? If we only consider economic development, they’re quite good. Libya is wealthier (~$13,000/capita) and more urbanized than most people think, even accounting for oil wealth. This simple model suggests about a 59% likelihood of democratization within five years. However, if we add some political characteristics (like the current Polity score and the regional level of democracy), the picture is much more pessimistic, around 10%. Again, this would be lower still without Qaddafi’s death. Obviously, other factors, like international support and the democratic trajectories of Tunisia and Egypt, will also play a role.

The central point remains that violent leader removal does increase the likelihood of democratization. For Libya’s citizens, this is their best chance in decades to achieve self-rule.


David October 21, 2011 at 12:50 pm

Sounds like something the Tea Party’s been saying for a while. About Obama.

idiot October 21, 2011 at 1:35 pm

Um, Joshua Tucker? The comment by Hume’s Bastard indicates that there is the Archigos database, and it had a peer-reviewed article by its maker, HENK E. GOEMANS, attached to said database that dealt with a similar matter (leadership death)[1]. Granted, Miller used the exact same Archigos database, with one difference, he reduced the sample size of countries under consideration.

What Goemans’ paper add is that, in the third wave, punishing leaders after an irregular transfer of power reduces the chance of democratization. Successful democracies usually not punish their leaders after an irregular transfer, and if so, they do so via imprisonment or exile, not death. If we are to believe that the Libyan government had assumed power once they took Tripoli and that Qaddafi was overthrown, then Qaddafi must have been punished (i.e., murdered) after losing power. That, in and of itself, reduces the chance of democratization occurring, even if the method by which Qaddafi was overthrown (irregular transfer) increased the chances of democratization. I want to know how Miller handled such a possibility.

As a side note, I am skeptical of this paper: Miller admits that “[a]ccording to [his] paper’s coding … there exist only 13 cases since 1875 of democratization concurrent with the violent ousting of an autocratic leader. Further, only five of these cases occurred without an irregular turnover in the previous five years. It is far more common for democratization to follow an irregular turnover after a short period of instability and flux.” That’s why he said democratization will occurs in a 5-year period after an irregular turnover. It may seem possible that an irregular turnover might “assist” in democratization in a semi-long term, but I admit bias in thinking that other factors probably play roles other than irregular turnover if it has to take that long.

Michael Miller October 21, 2011 at 2:39 pm

Hi, I’m the author of the post. Just a quick reply. What Goemans, Gleditsch, and Chiozza look at in the paper you reference is the stability of the ensuing democracy, not the likelihood of democratization. Putting the two stories together, Libya’s chances of democratization are increased by violence, but the resulting democracy would likely be fragile.

On my own findings, there are only 13 cases of democratization in the exact same year as a violent turnover. Most democratization processes that are triggered by violence take a couple years to come to fruition. (Sometimes this is just to allow for founding elections or a constitutional referendum, and sometimes there is an unsuccessful attempt to found a new autocratic government.) This is the most likely time frame for Libya or Egypt, although Tunisia might be #14.

Masha October 23, 2011 at 2:55 pm

Going off of your original post, I think the more serious consideration is how democratic and liberal the political factions of the current opposition are. You said, considering this factor, Libya’s chances of democratization lessen considerably (1o%?) Still, we all agree that Qadhafi’s elimination paves the way to democratization.

That being said, dictators can be removed without being killed (as the two posters underneath mentioned) and bloody beginnings to democracies often have a way of haunting people.

Unfortunately for Libya, they didn’t have a chance at a peaceful transition since Qadhafi and his regime were uncompromising and they lacked the instituions to give him a fair trial. But compromised transitions have ended up in more solidified democracies, because the process of democratization can, in and of itself, be democratic and foretell how consolidated a democracy will be.

My question is, if Qadhafi hadn’t died but had been removed some other way, (simply replaced and then exiled, perhaps) would Libya’s chances at a more SOLID democracy have increased?

Guillermo Aveledo October 21, 2011 at 10:25 pm

Just to be clear: “violent removal” does not necessarily entail the execution of the authoritarian ruler, right? A violet coup, a violent invasion, cannot lead unequivocally to a violent demise of an autocrat, or should it?

David Tomlin October 23, 2011 at 12:32 am

Second the previous. The title of the post is misleading. If Qadaffi’s life had been spared, his removal would still have been violent. Nothing in the post is pertinent to his death.

Joshua Tucker October 23, 2011 at 8:10 pm

David: Point well taken. Probably should have said “More on the aftermath of Qadaffi downfall” or something like that. Still, the death at the least punctuates the point that the downfall is irreversible at this point.

Michael Miller October 23, 2011 at 9:23 pm

Yes, agreed, what my post was getting at is the more general point that autocratic removals involving significant violence and bloodshed are conducive to democratization, with Qaddafi’s demise capping the turnover in Libya. A further distinction is between violent turnovers that end in the dictator’s death and those that don’t. I don’t believe that has been looked at before, although Jones and Olken have a 2009 paper looking at assassinations.

Out of curiosity, I ran the numbers distinguishing the two types of turnovers (using Archigos data). The Qaddafi route does provide a marginally higher likelihood of democratization, but the difference is not significant. So the point remains that Libya’s democratic chances are at a peak, and Qaddafi’s bloody end is unlikely to hold it back.

Truthful James October 24, 2011 at 11:36 am

I do hate to see simple regression analysis used in complicated situations. The authors need to use “democratization” as independent variable and include, violent overthrow, common ethnic background, common religions background, literacy levels, wealth, the presence of a middle class,and other possible variables to reach such a conclusion.

I would also suggest a re-reading of Crane Brinton’s excellent “Anatomy of a Revolution” to take into account over time the cycles through which a revolution goes through and most particularly the time necessary to complete a cycle, not complete a revolution.

Finally we have the problem of defining the country. Our crusading efforts to impose the principal of Judeo Christian democracy bears little weight in Muslim countries, whose development through 14 centuries resembles Europe at the same relative point in time. Democracies in these countries will, ceteris paribus, occur, only after Reformation and the Age of Enlightenment. So far, a Muslim John Locke has not made known his presence.

Hope springs eternal; simple statistics do not provide proof.

Comments on this entry are closed.

Previous post:

Next post: