What Do We Know about Democratic Transitions? A Listsicle of 9 Judgments

A year ago, a client asked me to summarize what we can learn from prior research about how transitions from authoritarian rule unfold. Looking back now on the list of nine judgments I produced, I think it holds up fairly well, so I thought I would post it here for a wider audience.

If this were an academic paper, I’d need to start with a bunch of definitions, to document the sources for each of these inferences, and to respond to competing claims. But it’s not an academic paper, it’s my blog, so I’m going to keep it simple and just offer my own thoughts with the occasional reference. The text of the list I produced a year ago is bolded, and I’ve added some annotations in plain text about my confidence in those earlier judgments and related reflections on recent events.

1. Popular protest during the transition period does not diminish prospects for democratization; in fact, it probably enhances them. High confidence. The idea that popular moderation facilitates democratization has been thoroughly debunked by Nancy Bermeo, Ruth Berins Collier and James Mahoney, and Elisabeth Jean Wood, among others.

2. Having the military involved in politics during the transition period is not necessarily a bad thing. Much depends on the intent of that intervention, and early signs that military leaders support democratization probably bode well for a democratic outcome. High confidence in the first part, low confidence in the second. I suppose it’s better to have juntas start out saying nice things about democracy than not, but we really don’t know if those early noises tell us much about where the transition is headed.

3. The transition process does not have to be inclusive or otherwise democratic for a democratic regime to result. High confidence. By definition, the process needs to end with free and fair elections for democracy to follow, but the process leading up to those elections doesn’t need to be broadly inclusive. It rarely is, yet democracy still happens.

4. Elite pacts are neither necessary nor sufficient for a democratic outcome to occur, and their presence probably does not much affect the odds that democracy will survive if it is established. High confidence; see this recent post for more on why.

5. Political rivals do not have to be evenly matched for a democratic transition to produce democracy. If anything, the opposite is probably true: democracy is more likely to emerge in cases where de facto power tilts heavily in the direction of groups that favor democratization for its own sake. Moderate confidence. This is a major conclusion of now-ambassador Mike McFaul’s 2002 book, and I think the results of the Arab Spring so far support it as well, but power is hard to measure and the sample is small.

6. Cohesive opposition movements help to keep transitions on a path toward free and fair elections in ways that fragmented oppositions do not. Moderate confidence, although again, I think the trajectories of the Arab Spring cases so far bear this out.

7. Choices about institutional design—separation of powers and electoral systems—do not have a big effect on the outcome of the transition process, in part because they are endogenous to it. High confidence. As Philippe Schmitter puts it, “Allegedly democracy-unfriendly institutions are symptoms, not causes.”

8. Democratic transitions are more likely to produce a democratic regime in countries that depend heavily on foreign aid than in ones that don’t. Moderate confidence. It depends on where the aid comes from, what other interests the donor countries have at stake, and whether or not alternative sources of state finance are available. On Egypt, for example, I’m skeptical that U.S. military aid provides much leverage.

9. The most likely outcome of a democratic transition nowadays is a competitive authoritarian regime, either because initial elections will be unfair by design or because the party that wins those elections will quickly use state resources to advantage itself in future contests. Highest confidence. Democracy is hard to produce and relatively easy to undo. Just ask the Iraqis, or the Nicaraguans, or the Hungarians, or…

Leave a comment


  1. Michael Ross

     /  January 19, 2012

    Great list. I wonder if you had any thoughts about the validity of the now-ubiquitous claim that economic inequality ex ante has a large effect on the outcome – with either low (Boix) or moderate (Acemoglu & Robinson) inequality raising the likelihood of a successful transition. It’s not an easy one to test, but I’d be very interested if you had any views.

    • Thanks for reading and commenting, Michael. I don’t have strong views on those hypotheses for the reason you mention: they are hard to test (well), and the original purpose of this list was to identify claims for which there was a solid body of empirical evidence. I know those hypotheses have been tested, but I don’t trust the tests because I don’t trust the data on inequality. Even if you think the time series that folks have ginned up are plausible, they’re definitely not reliable enough to produce confident inferences about any but the largest differences. That’s how I see it, anyway. If you see it differently, I’d be very interested to hear your take.

  2. Just wondering: do you think the size of the middle class has any influence on the speed of the transition? (the so-called Moore hypothesis). Are there any necessary structural prerequisites for the transition to succeed?

    There is much speculation about the recent protests in Russia (as an expression of the discontent of the emerging middle class) and about the tendencies of the emerging Chinese middle class.

    • I think those kinds of structural preconditions matter more for whether or not democracy sticks if it’s tried than they do for the outcome of the transition period. Also, based on some of the work I cited about popular protest, I don’t think the middle class is usually as influential in the transition process as the wider population of urban paid labor, to include blue- and white-collar workers and state bureaucrats.

  3. The wild card in all this is the explosive pace of communications of all kinds. I think that changes the dynamic everywhere, e.g., it is far easier for moderates and minorities to find each other and form coalitions than previously. Also, censorship is all but impossible. Finally, in any state with a reasonable electronic infrastructure, the whole world is constantly watching.

    What a world.

    • I agree that technology is probably changing some of these dynamics, but I don’t think those changes are unilaterally favorable to democratization. As Yevgeny Morozov argues, states are adopting and adapting to these technologies as well, and they can be used for surveillance and disinformation, too. So I think the effects are mixed so far.

  4. Thanks, a thoughtful entry. I was wondering if you would make any regional provisos on these universalist statements. Number nine is supported by the case of Ukraine as well, among others.

    • I would not add regional provisos, but it’s not because I don’t think variation across space and over time matters. Quite the opposite. Nearly all of the items on the list are negative, and that’s not an accident. As I said in a chapter I contributed to a recently published Routledge Handbook of Democratization, I think the search for covering laws of democratic transitions is misguided. We’re better off looking for mid-range theories that deal with the contingent effects of specific factors during specific historical phases.

  5. Jaroslav Petrik

     /  February 17, 2012

    I wonder where does the nature of the pre-transition regime come into your list. A huge body of literature (Stepan, Linz, Brooker, etc.) and a whole now dead discipline (‘sovietology’) spent a great deal of efforts studying which types of regimes are likely to democratize and which are lost causes. The shape of the house you’re trying to rebuild matters as much as, and largely determines, the way you rebuild it.

  1. Lidos #2 – destruição criativa
  2. Recomendados #2 « destruição criativa

Leave a Comment

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

WordPress.com Logo

You are commenting using your WordPress.com 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 )

Google+ photo

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

Connecting to %s


Get every new post delivered to your Inbox.

Join 5,754 other followers

%d bloggers like this: