[First published April 7, 2006] This morning I came across two blogs (here, and here) that relied on Matthew White’s page to dismiss the democratic peace. Since White continues to have influence on the democratic peace debate, I have a few words to say on his statistics.
Thanks to Dean Esmay for referring me to Matthew White’s page that raises questions about the democratic peace. I know of White’s useful Historical Atlas of the Twentieth Century , and have used his statistics in my own research. He is careful, thoughtful, and systematic in what he presents, so when he questions the democratic peace, he has to be answered.
First, he presents the pros and cons about the various possible exceptions to the democratic peace. Keep in mind that the democratic peace, among other propositions, says that democracies don’t make war on each other. So, a true negative example thunders against this. Many have been proposed such exceptions, such as the War of 1812, the Boar War, WWI and Germany, democratic Finland being allied with Hitler in WWII, and the American Civil War. The sheer number of these exceptions and the weight of all the pros that White provides gives the impression that there has to be something to at least one or more of them. I have not studied them all, but those I have spent some time on in my own research, such as Germany in WWI, the case of Finland, the Boar War, and the Civil War simply cannot be treated as true exceptions. Others who have investigated these possible exceptions, in addition to the rest of them on White’s list, agree. In particular, I point you to Bruce Russett’s Grasping the Democratic Peace, James Lee Ray’s Democracy and International Conflict, and Spencer R. Weart’s, Never At War. Russett and Ray are political scientists, Weart is an historian. See also my democratic peace bibliography and my Q & A, which answers questions about some of these supposed exceptions (use the search command to find them).
After going through the exceptions, White concludes that the democratic peace depends on the definition of democracy and war. Researchers know this, of course, and have done different things about it. One is to collect their own data according to very clear, replicable criteria, while others have used data on democracy and war that have a wide reputation for their validity. Two sources especially have been important. One is the statistics on war collected by Melvin Small and J. David Singer, such as their data on wars from 1816 to 1992. I have used this in my research (see the table in the upper right here) as have hundreds of others. I should say that Small and Singer do not accept the democratic peace, which makes their classification of wars and democracies since 1816 particularly important. For democracy, in addition to the Small and Singer classification, which I am one of the few to use, there is the very popular and respected Polity data, which provides a scale for measuring the degree to which a country is democratic or autocratic. For an additional data set used in replicating the democratic peace, go here.
What is noteworthy about all these different data on democracy and war whose definitions or criteria slightly differ, is that those using them have come out with the same conclusions: there is a democratic peace. Replications have well established this to the point that students of international relations say it is the best-tested proposition in the field and almost has the status of a law.
Now, Mathew White lists 39 wars 1945-1999, and says that six “might have been between democracies,” which means they might not have been, but still he makes much of it in calculating the probability of this happening by chance. Rather than deal with his “might have been,” I’m going to actually collect data from two sources on democracy and international violence between countries. The source I will use for violence is compiled by Monty G. Marshall on “Major Episodes of Political Violence 1946-2004;” for democracy, I will use Freedom’s House’s “All Country Ratings from 1972-2003” (Sorry, I can’t find it on their stupidly remodeled website). Freedom House is not a proponent of the democratic peace (I don’t recall them ever mentioning it), so we can treat their data as independent of this proposition. Similarly with Marshall, who along with Ted Gurr, is the author of the Peace and Conflict Survey 2005 that I referred to in a former blog for ignoring the democratic peace.
From Marshall’s data, I’ll include as violence any that is indicated in his data as “international.” This is a hard test, since it includes violence short of war. From Freedom House, I will use their Free (F) rating of a country for a year as defining a liberal democracy in terms of civil liberties and political rights.
First, how many liberal democracies are there versus the total number of countries. For five years spans after 1972 and ending with 2003 (year, number of liberal democracies, total number of countries):
1972, 43, 148
1975, 39, 158
1980, 50, 162
1985, 55, 166
1990, 64, 165
1995, 75, 191
2000, 85, 192
2003, 87, 192
Now, for the classification of violence between types of regimes (F = free, PF = partly free, NF = not free, where F-F = between free countries, etc.)
F-F = 0
F-PF = 6
F-NF = 11
PF-PF = 5
So, between which countries is there the least violence? Between liberal democracies. Which countries are the most violent towards each other? Nondemocracies. All as precisely predicted by the democratic peace.
A note on statistical tests. Think of this subjectively. Here you have all these liberal democracies for each of thirty-one years, and none of them have violence between them. This is not a matter of just five or ten democracies, but by the end of the 1990s, there are over eighty. This number is not my reckoning, but that of Freedom House. And by Marshall’s data, in spite of so many democracies, none had violence between them vs. 20 cases of violence between the nonfree ones during these years.
Now, some people don’t like subjective statistics, so lets calculate the probability. There are 46 cases of international violence, and six alternative ways that could occur (e.g., F-F, or PF-PF). Let the number 1 stand for the F-F alternative, and the other five numbers for each of the others. Throw a six-numbered die 46 times, and what is the probability that it will never come up with a 1? The probability that it will not come up a 1 in one throw is 5/6. So, the probability of no 1 in 46 throws is 5/6 to the 46th power (assuming each case of violence is independent), which is a probability of happening by chance of 8.02E-36, or about the probability of one being hit by a meteor.
Obviously, there has to be something more than chance here. And what is that something? Surprise. It is two countries having democratic governments. That is, the democratic peace.
Link of Note
“DOES DEMOCRACY CAUSE PEACE?” By James Lee Ray. In Annual. Review of Political Science 1998. 1:27-46.
The idea that democratic states have not fought and are not likely to fight interstate wars against each other runs counter to the realist and neorealist theoretical traditions that have dominated the field of international politics. Since the mid-1970s, the generation of new data and the development of superior analytical techniques have enabled evaluators of the idea to generate impressive empirical evidence in favor of the democratic peace proposition, which is reinforced by substantial theoretical elaboration. Some critics argue that common interests during the Cold War have been primarily responsible for peace among democracies, but both statistical evidence and intuitive arguments cast doubt on that contention. It has also been argued that transitions to democracy can make states war-prone, but that criticism too has been responded to persuasively. The diverse empirical evidence and developing theoretical bases that support the democratic peace proposition warrant confidence in its validity.
RJR: It is Ray who should be referenced on the democratic peace, and not Matthew White. But, that is too much to expect out of the isolationist libertarian crowd that frequently quotes him.