The era of democratization is not over

In the latest issue of the Journal of Democracy, (PDF), Marc Plattner makes the provocative claim that “the era of democratic transitions is over, and should now become the province of the historians.” By that, he seems to mean that we should not expect new waves of democratization similar in form and scale to the ones that have occurred before. I think Plattner is wrong, in part because he has defined “wave” too broadly. If we tighten up that concept a bit, I think we can see at least a few possibilities for new waves in the not-too-distant future, and thus an extension of the now–long-running era of democratization.

In his essay, Plattner implicitly adopts the definition of waves of democratization described by Samuel Huntington on p. 15 of his influential 1991 book:

A wave of democratization is a group of transitions from nondemocratic to democratic regimes that occur within a specified period of time and that significantly outnumber transitions in the opposite direction during that period of time.

Much of what’s been written and said about waves of democratization since that book was published accepts those terms and the three waves Huntington identifies when he applies them to the historical evidence: one in Europe from the 1820s to the 1920s; another and wider one in Europe, Latin America, and Asia from the 1940s to the early 1960s; and a third and so-far final one that began in Portugal in 1974, has been global in scope, and now appears to have stalled or ended.

I find Huntington’s definition and resulting periodization wanting because they focus on the what and don’t pay enough attention to the why. A large number of transitions might occur around the same time because they share common underlying causes; because they cause and reinforce each other; or as a matter of chance, when independent events just happen to cluster. The third possibility is not scientifically interesting (cf. the Texas sharpshooter fallacy). More relevant here, though, I think the first two become banal if we let the time lag or chain of causality stretch too far. We inhabit a global system; at some level, everything causes, and is caused by, everything else. For the wave idea to be scientifically useful, we have to restrict its use to clusters of transitions that share common, temporally proximate causes and/or directly cause and reinforce each other.

By that definition, I think we can make out at least five and maybe more such waves since the early 1900s, not the three or maybe four we usually hear about.

First, as Plattner  (p. 9) points out, what Huntington describes as the “first, long” wave really includes two distinct clusters: 1) the “dozen or so European and European-settler countries that already had succeeded in establishing a fair degree of freedom and rule of law, and then moved into the democratic column by gradually extending the suffrage”; and 2) “countries that became democratic after World War I, many of them new nations born from the midst of the European empires defeated and destroyed during the war.”

The second (or now third?) wave grew out of World War II. Even though this wave was relatively short, it also included a few distinct sub-clusters: countries defeated in that war, countries born of decolonization, and a number of Latin American cases. This wave is more coherent, in that all of these sub-clusters were at least partially nudged along by the war’s dynamics and outcomes. It wouldn’t be unreasonable to split the so-called second wave into two clusters (war losers and newly independent states) and a clump of coincidences (Latin America), but there are enough direct linkages across those sets to see meaning in a larger wave, too.

As for the so-called third wave, I’m with Mike McFaul (here) and others who see at least two separate clusters in there. The wave of democratization that swept southern Europe and Latin America in the 1970s and early 1980s is temporally and causally distinct from the spate of transitions associated with the USSR’s reform and disintegration, so it makes no sense to talk of a coherent era spanning the past 40 years. Less clear is where to put the many democratic transitions—some successful, many others aborted or short lived—that occurred in Africa as Communist rule collapsed. Based partly on Robert Bates’ analysis (here), I am comfortable grouping them with the post-Communist cases. Trends in the global economy and the disappearance of the USSR as a patron state directly affected many of these countries, and political and social linkages within and across these regional sets also helped to make democratization contagious once it started.

So, based on that definition and its application, I think it’s fair to say that we have seen at least five waves of democratization in the past two centuries, and perhaps as many as six or seven.

Given that definition, I think it’s also easier to see possibilities for new waves, or “clusters” if we want to make clearer the distinction from conventional usage. Of course, the probability of any new waves is partially diminished by the success of the earlier ones. Nearly two-thirds of the world’s countries now have regimes that most observers would call democratic, so the pool of potential democratizers is substantially diminished. As Plattner puts it (p. 14), “The ‘low-hanging fruit’ has been picked.” Still, if we look for groups of authoritarian regimes that share enough political, economic, social, and cultural connections to allow common causes and contagion to kick in, then I think we can find some sets in which this dynamic could clearly happen again. I see three in particular.

The first and most obvious is in the Middle East and North Africa, the region that has proved most resistant to democratization to date. In fact, I think we already saw—or, arguably, are still seeing—the next wave of democratization in the form of the Arab Spring and its aftermath. So far, that cluster of popular uprisings and state collapses has only produced one persistently democratic state (Tunisia), but it has also produced a democratic interlude in Egypt; a series of competitively elected (albeit ineffective) governments in Libya; a nonviolent transfer of power between elected governments in Iraq; ongoing (albeit not particularly liberal) revolutions in Syria and Yemen; and sustained, liberal challenges to authoritarian rule in Bahrain, Kuwait, and, perhaps, Saudi Arabia. In other words, a lot of countries are involved, and it ain’t over yet. Most of the Soviet successor states never really made it all the way to democracy, but we still think of them as an important cluster of attempts at democratization. I think the Arab Spring fits the same mold.

Beyond that, though, I also see the possibility of a wave of regime breakdowns and attempts at democracy in Asia brought on by economic or political instability in China. Many of the autocracies that remain in that region—and there are many—depend directly or indirectly on Chinese patronage and trade, so any significant disruption in China’s political economy would send shock waves through their systems as well. I happen to think that systemic instability will probably hit China in the next few years (see here, here, and here), but the timing is less relevant here than the possibility of this turbulence, and thus of the wider wave of democratization it could help to produce.

Last and probably least in its scope and impact, I think we can also imagine a similar cluster occurring in Eurasia in response to instability in Russia. The number of countries enmeshed in this network is smaller, but the average strength of their ties is probably similar.

I won’t hazard guesses now about the timing and outcome of the latter two possibilities beyond what I’ve already written about China’s increasing fragility. As the Arab Spring has shown, even when we can spot the stresses, it’s very hard to anticipate when they’ll overwhelm the sources of negative feedback and what form the new equilibrium will take. What I hope I have already done, though, is to demonstrate that, contra Plattner, there’s plenty of room left in the system for fresh waves of democratization. In fact, I think we even have a pretty good sense of where and how those waves are most likely to come.

Advocascience

I think comparative politics has a bigger problem with conflicts of interest than scholars who work in this field generally acknowledge. I don’t think the problem can be eliminated, but I imagine that talking about it more can help, so that’s what I’m going to do.

When you hear the term “conflict of interest,” you probably think of corporations paying for studies that advance their commercial interests. I know I do. It’s easy to see why studies on the effectiveness of new drug therapies or the link between pollution and cancer, for example, warrant closer scrutiny when they’re funded by firms with profits riding on the results. You don’t have to be a misanthrope to believe that the profit motive might have shaped the analysis, and there are enough examples of outright fraud to make skepticism the prudent default setting.

That’s not the only conflict that can arise, though. What I think many scholars working in comparative politics don’t appreciate as much as we should is that it’s also possible for political values and advocacy to play a similar role, and to similar effect. When a researcher’s work deals with issues on which he or she has strong moral beliefs, that confluence can hinder his or her ability to identify and fairly weigh relevant evidence. Confirmation bias is hard to overcome, especially in studies that rely entirely on an author’s interpretation, as many qualitative studies do. The problem is even more intense if the researchers’ personal life is interwoven with her work. Certain conclusions may be more palatable or appealing to people with certain values, and it can be professionally and personally damaging for researchers to report findings that suggest the work their friends and colleagues are doing may not be all that useful, or may even be counterproductive.

The example I know best comes from one of my primary research interests, comparative democratization. Some of the best-known and most respected researchers and organizations in this sub-field routinely engage in advocacy through op-eds, policy briefs, and meetings and speaking engagements with advocates and development professionals. One of the leading journals on this topic, the Journal of Democracy (JoD), is published for the National Endowment for Democracy, a U.S. government-funded organization that supports the U.S. government’s efforts to promote democracy around the world. In contrast to conventional academic practice, most submissions to JoD are commissioned by the editors, and they aren’t formally peer-reviewed.

Perhaps it’s just a coincidence, but for the past 20 years or so, the main themes to emerge from research on this topic are that democratization has all kinds of ancillary benefits—peace, wealth, and freedom from terrorism, to name a few—and that the kinds of the things the U.S. government and the advocates it supports generally do to advance democratization are helpful. In other words, scholars’ studies often reach conclusions that affirm the value of U.S. policy and their own advocacy, which is intimately connected to their personal beliefs and relationships.

That happy alignment doesn’t automatically invalidate those studies, of course, but I think it does warrant closer scrutiny than it now gets. I have great respect for many of the people working in democratization studies, and I happen to share their moral convictions that democracy is the best form of government and that every human being deserves citizenship. Still, let’s be honest: we feel better when we believe our research is helping people we admire change the world for the better, and we’re more likely to get that positive feedback when our findings validate the work those people are already doing. The effects of this feedback loop on the questions we ask, the designs we adopt to answer them, and the conclusions we reach may not be trivial. I think we should talk more about it, both in a general way and whenever evaluating specific pieces of research.

It would be unfair and probably unethical of me to conclude without pointing out that similar issues arise when scholars do consulting work, as I have for the past 15 or so years. Even if a client asks for as fair and objective a study as possible, interpersonal and financial concerns can shape the design of the analysis and interpretation of the results. For example, if you’re paid handsomely to develop a system to forecast event X, you have a financial interest in saying that you can indeed forecast event X and that you can do it well. We can ameliorate this problem by being as transparent as possible about our funding, data, and methods, but we can’t eliminate it, and we’re usually not the best judges of our own motives. Contract research like this occupies a pretty small space in comparative politics right now, so I don’t think this is having much effect on the field at the moment, but I think it’s important for me to note it, given the career path I’ve taken.

Follow

Get every new post delivered to your Inbox.

Join 7,090 other followers

%d bloggers like this: