Pax Democritus?

There is a lot of talk on the internet boards in support of the United States invasion of Iraq based on the theory of Pax Democritus, or the Democratic Peace Theory. The theory is that since democracies do not wage war against democracies, if a democracy were to be established in Iraq, Iraq would no longer be any threat to the United States. The theory is the work of a Professor R. J. Rummel, based largely on the POLITY Data Set. The POLITY Data Set rates various countries on how close to a liberal democracy they are, giving them a score between 10 for liberal democracies and –10 for illiberal non-democracies.

One problem is a logical one; because an event hasn't happened doesn't mean it won't happen. Every scientific theory is only as valid as its next test, leading to the joke in every Physics I class about the physicist who drops a book every day to see if that would be the day the book doesn't fall. If it doesn't fall, the physicist will win the Nobel Prize. This does nothing to disprove the theory, as every theory in science, from Gravity to Relativity to Evolution is accepted based on the preponderance of data and the absence of contradictory data.

But is it true that two democracies have not gone to war with each other? If it can be shown that two democracies have gone to war with each other, that will disprove the theory. Unfortunately, every time an event is shown that can disprove the theory, advocates of the theory are quick to try to ignore or explain away that new data. In science, contradictory data demands attention; data exclusion is the hallmark of bad science.

Attempts to disprove the theory have run afoul of several problems of definition. The first of which is determining when a country is a democracy. Many examples of democracies at war are resolved to "Ah, but that one isn't a democracy." It's a classic example of the No True Scotsman fallacy. "No Scotsman puts sugar on his porridge." "My uncle Angus puts sugar on his porridge." "Ah, but no true Scotsman puts sugar on his porridge." In cases of a war between democracies, proponents of the theory are quick to discount one of the participants as "not a democracy."

A subset of this problem is the definition of liberal democracy, an even more ephemeral concept. The POLITY Data Set rates countries not on the sole criterion of whether or not they are democratic. The highest scores go to liberal democracies. Intermediate scores go to liberal non-democracies and illiberal democracies. The lowest scores go to illiberal non-democracies. Some proponents of the theory try to talk about the liberal democracies only, but that severely limits the discussion. In 1816, the year the Data Set begins, the liberal democracies of that era all had slavery within either their own or their colony's borders. Universal suffrage was not well known either. In fact, before WWI, it is hard to say that by today's standards any liberal democracies existed. After WWII there exists the special case of the liberal democracies of the world being allied against a mutual external threat, so if the case is limited to liberal democracies, there is too little data to reach a conclusion.

The dataset begins on the year 1816, and therefore the theory states that since 1816 two democracies have not gone to war with each other. 1816 is a convenient number because 1815 was the end of the Napoleonic Era and therefore the beginning of the modern world. After the fall of Napoleon, liberal ideas spread across Europe. Starting at the year 1816 excludes from the data set all conflicts that occurred before 1816.

The second problem is the theory's definition of "war," which states that there must be at least 1000 deaths. Any conflict with fewer that 1000 deaths is considered not to be a war. According to Prof. Rummel, a threshold needed to be established, but this is a case of the Fallacy of the Beard — when can you say someone really has a beard? Was it a war when the United States invaded Haiti? Panama? How about the bombing of Libya? None of those were wars as defined by the theory.

Another aspect of the definition of "war" is that internal conflicts are not counted. Civil Wars, Wars of Secession, guerilla resistance, and any other conflict contained within the border of a single country are not counted.

While there is a very comprehensive list of democracies at war with each other, I'll concentrate on five specific examples and why they are excluded.

The first example is the 1812 war between the United States and Great Britain. Advocates of the theory insist that the dataset starts at 1816. The problem is that neither the United States nor Great Britain were significantly different in 1816 than they were in 1812. Both are listed as democratic in 1816, so both should be listed in 1812. The war of 1812 is excluded because it occurred before 1816. Professor Rummel instead excludes this example by saying that in 1812 Great Britain was not a democracy, even though for that time only the United States was more of a democracy.

The second example is the War Between the States in America. Both the United States of America and the Confederate States of America had nearly identical constitutions, and both had their own currencies, elected officials, scheduled elections, and foreign policy. But the Democratic Peace Theory insists that the War Between the States doesn't count because it was an internal conflict. The Confederate States of America isn't even given a POLITY rating, although it is likely that its score would be the same as that of United States since it was, for it's time, very liberal and democratic. The citizens of both the U.S.A. and the C.S.A. regarded the C.S.A. as democratic.

The third example is World War One. While many of the participants were not democracies, such as Tsarist Russia or the Ottoman Empire, there are two participants that beg for attention: Great Britain and Germany. Advocates of the Democratic Peace theory have to work hard to exclude Germany under the Kaiser as not being one of the world's democracies, because this is the biggest example of the theory failing.

Germany is considered to be "not a Democracy" because it was ruled by a monarch and a parliament, just like Great Britain. It is true that the monarch had actual powers, just like in Great Britain. No definition can easily exclude one and include the other. Germany is given a much lower rating in polity (a 2 instead of an 8) but it should be remembered that this theory discusses democracies, not liberal democracies. There exists no reason to exclude this example.

The fourth example is the Ruhr invasion of Germany by France. While advocates of the Theory exclude Kaiser Germany because it wasn't democratic enough, advocate of the theory exclude the invasion by Third Republic France of Weimar Germany as it not being enough of a war. In order for any act of war to be counted, it must result in at least 1000 deaths. Professor Rummel doesn't even mention this incident in his discussion of the theory.

For those not familiar with the Ruhr invasion, the story is sad and simple. The French government was unhappy with the enforcement of the Treaty of Versailles, and was also unhappy that Germany wasn't punished enough. So French troops entered the Ruhr area and declared it a separate state, complete with a separate government supported by French troops. After pleading to the League of Nations, the German government threatened war if French troops did not leave the Ruhr. The British government informed the French government that if war broke out, Britain would not support France this time. The French government got the message, and withdrew troops. The independent Ruhr government collapsed, the quisling leaders were executed, and Ruhr rejoined Germany.

The fifth example occurs during World War Two, with the fighting between Finland and Great Britain. This is the one that proponents of the theory do not want to examine. Some try to say that the Finnish were not fighting for the Nazis but only against the Soviets. However, the British bombed Finland and sunk Finnish merchant ships in the Baltic Sea. Finnish deaths did come to 1000 military and 2000 civilian, but that is the sum of both fronts.

The first example doesn't count because it was before 1816.

The second example doesn't count because it was a civil war.

The third example doesn't count because Germany wasn't really a democracy.

The fourth example doesn't count because it wasn't really a war.

The fifth example doesn't count because, even though Britain bombed Finnish properties and territories, causing casualties, they supposedly weren't at war with each other, just each other's allies.

Five examples, and in each case proponents of the theory insist that the examples don't count. The data set is modified to eliminate counter examples. In science, that is called "fixing the data" and is frowned upon. These five examples either have to be eliminated, or the theory is disproven. It is not safe to say that a successful democratizing of Iraq will result in the neutralization of a threat, as democracies have indeed gone to war with each other.

August 31, 2005