Spiro, D., 1994. The Insignificance of the Liberal Peace. INTERNATIONAL SECURITY, 19, 50.
“Ever since Michael Doyle’s 1983 essay pointed out that no liberal democracy has ever fought a war with another liberal democracy, scholars have treated pacifism between democracies as ‘the closest thing we have to a law in international politics…The purpose of this article is to raise critical questions about newly rediscovered law of peace among democracies” (50).
“I will argue that the absence of wars between liberal democracies is not, in fact, as significant pattern for most of the past two centuries. Studies that do claim significance for the absence of wars between democracies are based on analyses that are highly sensitive to the ways that they select definitions of the key terms of democracy and war, and to the methods they choose for statistical analysis. I argue that much of the quantitative literature on democracy and war has little to do with the theories it seeks to confirm, and that the results rest on methods and operationalization of variables that undergo contortions before they yield apparently significant results” (51).
Spiro argues that “democracy” is a term that is difficult to measure, and those who try to create a taxonomy of democratic countries eventually end up with conflictual lists. This is the same for war, as Mansfield argued in 1988.
Spiro then examines the statistical significance of zero, by that, the author means to question the significance of only exploring dyadic groupings that are not at war. Because dyadic war is such a rare occurance in the international system, it is not unexpected to not find war between countries labeled as democratic. He finds a generally that, through random probabilities, and because the cases of international war are so slim, the odds of two states randomly not going to war are great, barring some historical outliers (WWI, WWI, etc.).
There is also a specific critique of Maoz and Russett.