In lieu of an abstract, here is a brief excerpt of the content:

  • ExchangeMisunderstanding Gradualism
  • Thomas Carothers (bio)

I greatly appreciate the serious, thoughtful responses by Edward Mansfield and Jack Snyder and by Francis Fukuyama to my January 2007 Journal of Democracy essay entitled "How Democracies Emerge: The 'Sequencing' Fallacy." Unfortunately, Professors Mansfield and Snyder attribute to me ideas that I do not express in that article, mischaracterize some of the points that I do make there, and fail to engage at all with some of my fundamental arguments. The very first sentence of their response hints at the problems to come. For nowhere in my article do I, as they claim, "acknowledge the now widely recognized fact that countries taking early steps on the journey from dictatorship toward electoral politics are especially prone to civil and international war, violent revolutions, and ethnic and sectarian bloodshed." I make no such acknowledgement because in fact I reject this view, and I find the evidence for it set out in their book Electing to Fight unpersuasive.

Their use of highly refined statistical methods to make their case is undermined, in my view, by some significant problems with their classification of cases. Thus they somehow categorize the 1982 Falklands War—a conflict started by the Argentine military junta—as an example of the warlike propensities of democratizing countries. Similarly, most observers of Balkan politics will be surprised to learn from Electing to Fight that it was the democratizing character of Serbia under the thuggish Slobodan Milošević that was responsible for Serbia's militaristic predations in Kosovo and elsewhere.

More generally, Mansfield and Snyder base a sizeable part of their [End Page 18] case for the dangers of democratization on the internecine conflicts that flared up after the Soviet Union and Yugoslavia each broke apart. Blaming democratization for the wars that sprang from the collapse of multiethnic authoritarian federations—the messy, often violent process of establishing the borders of the newly emergent states—is a significant category mistake. As Francis Fukuyama stresses in his comment, the overwhelming bulk of the many wars that have raged in Europe over the last half a millennium have been related not to democracy, but to state-building.

Throughout their response, Mansfield and Snyder steadfastly reassert the preferability of sequentialism and their disapproval of what they call "out-of-sequence" democratic transitions. They do not, however, actually address, let alone refute, my two core arguments. To recapitulate, these are:

1) Outside East Asia, autocratic governments in the developing world have a terrible record as builders of competent, impartial institutions due to deep-seated conflicts between autocracy and the second phase of state-building (that is, going beyond establishing a monopoly of force to creating effective institutions); and

2) Although emergent democratic governments struggle with this second, institution-strengthening phase of state-building and with rule-of-law development, they have some important advantages when it comes to tackling these tasks.

Mansfield and Snyder maintain a strong skepticism about the wisdom of holding elections in places that lack what they believe are necessary preconditions for democracy. Yet this view overlooks the decades of authoritarian rule in Africa and other parts of the postcolonial world that have left so many states in such terrible condition. Prescribing the deferral of democracy—and consequently the prolongation of authoritarian rule—as a cure for the ills of prolonged authoritarianism makes little sense. In this regard, it is telling that a probing recent study of African politics since the early 1990s finds that those countries which moved early toward elections and persisted with elections thereafter have done better at consolidating all aspects of democracy than those countries that delayed holding elections.1

Mansfield and Snyder try to soften their sequentialism in two ways. First, they say that elections need to be deferred only until competent state institutions "have begun to take root" or even just until the development of such institutions has "barely begun." But then they return to their more standard sequentialist dictum that "it is necessary to have impartial state institutions" before risking elections.

The gap between barely beginning to have competent, impartial state institutions and having them in place and functioning well is enormous, and bridging it can take fifty years, a century...