Review of Power Kills
Ted Galen Carpenter's Review
Ted Galen Carpenter's rebuttal to Rummel's reply
Other Related Democratic Peace Documents On This Site
Ted Galen Carpenter, an otherwise well-respected analyst of American foreign and defense policy at the Cato Institute, wrote a miserable review of a tremendously important book by R. J. Rummel. I don't know whether Carpenter is carrying some grudge against Rummel, merely skimmed the book he was reviewing, or is unfamiliar with the issues at hand, but all are in evidence in his review. There are at least a dozen specific problems.
1. First, this was not a review of Power Kills, but an extended out-gassing of Carpenter's indigestion over one element of Rummel's book, specifically the "democratic peace" proposition (that well-established democracies don't war against each other). In so doing, he ignores the most significant point of Rummel's book: democracy is a powerful, general method of non-violence. Democracy solves, by controlling power better than any other form of government, the problems of genocide and mass murder, internal violence, bilateral violence, as well as war. In Power Kills Rummel provides the larger theory and evidence of the intimate connection between democracy and non-violence, in all spheres of human action. This is critically important stuff, especially in today's world where it appears that interstate war is taking a back seat to internal wars and mass violence. But most of this was ignored in the review. Carpenter did note that Rummel's Death By Government (Transaction, 1994) documents how the less libertarian a government, the more of its own citizens it kills, so he has understood a few things, but not everything. Why the single-minded focus on the inter-democratic peace? Didn't he see the big picture? Carpenter wrote the equivalent of a critical review of Einstein's General Theory, while only examining the Special Theory ÷ and making an inadequate examination at that.
2. Carpenter makes ad nauseum use of ad hominem attacks: "greatly exceeds his grasp," "egregious oversimplification," "almost contemptuous" (well, was it or wasn't it?), "borders on perversity," "vague and slippery treatment," etc. If Carpenter is intentionally mean-spirited, he should chose less prestigious journals for his tantrums; if he's not, then he's a poor reviewer who lets his emotions preclude analysis.
3. What's with this insulting phrase that accuses Rummel of "robotic invocation of statistical data"? What's wrong with invoking data? One of the strong points about classical liberalism (and today's libertarian political movement) is the use of evidentiary arguments with plenty of data. This approach has been used very successfully to challenge subjective liberal assertions about everything from outrageously overstated dangers of everyday life (a distorted view of risk justifying government intervention) to the alleged success of innumerable wealth redistribution programs, from welfare to the arts. Empirical analysis should be a powerfully important component of any political agenda. Did Rummel use the wrong statistical tools, or apply them incorrectly. If so, Carpenter should have told us. He didn't.
Carpenter himself notes that extraordinary claims require extraordinary evidence. Well, Rummel's evidence is extraordinary and his widely respected studies have long strongly supported libertarian positions. What is it about Carpenter's epistemology that makes him so dismissive of data?
4. Rummel's theoretical work and his strong statistical analysis have been further validated by historian Spencer Weart, in his book Never At War: Why Democracies Will Never Attack One Another (Yale, 1998). Weart examines every recorded instance of violent conflict among democracies and comes to the remarkable conclusion that they have never gone to war against each other. This is no longer a curious and debatable statistical construct, but a strong (perhaps even absolute) historical rule. Curiously, in Carpenter's Rejoinder, he writes that "Rummel's vision of a world of peaceful democracies is enticing, but like all utopian visions it exists only in the realm of fantasy." Given Weart's exhaustive historical research, it is clear that Rummel's conclusion that democracies are peaceful with one another is no fantastic utopian vision, but an extremely well grounded practical and realistic one.
5. Carpenter engages in sophistry. He poses a hypothetical and treats it as a fact. Because a possible war between a democracy and a totalitarian country (specifically a "hot" war between the US and USSR superpowers) would have been worse than an actual war between two totalitarian belligerents (China and Vietnam) this proves wrong Rummel's proposition that inter-totalitarian wars are the most violent of all.
Rummel's point was, quite obviously, about regime types in the aggregate. One case does not an argument make--certainly Carpenter knows this. On the other hand, one case can indeed unmake a case (known in epistemological circles as falsifiability). Perhaps Carpenter's dreaded fear of "statistical data" prevents him from using real events. He could have done something like compare the extent of violent conflict between two countries while they were under different types of government, if he had wanted to be persuasive. He didn't. He isn't.
6. Carpenter is obtuse. In his rejoinder to Rummel, Carpenter explains how covert action by democracies against other countries (most of them nominal democracies) proves that there is no inter-democratic peace. But Rummel (and others) have dealt with this issue, and with a little bit of thought, Carpenter himself could have figured it out: covert action is not war, it is action short of war. Furthermore, covert actions are by their nature secretive and non-democratic, and so, according to Rummel's theory on democracy and peace, we should even expect them to be violent. But even here Carpenter misses the point. Covert actions by democracies, although conducted secretly and non-democratically, are done within a larger democratic (and therefore limiting) context--so they are minimally violent compared to war.
7. Carpenter does adequately sum up Rummel's three explanations for the overall peaceful nature of democracies, including Rummel's "spontaneous social field" work that echoes much of Hayek's "spontaneous order." Yet having apparently grasped Rummel's broad-ranging theoretical foundations about the multi-faceted pacific nature of democracy, all he offers as criticism are a few dubious historical counter-examples that only address one facet of democracy, namely it's war nexus. One of the impressive features of Rummel's research is his discovery that democracy is a whole cloth, with interwoven strands of nonviolence and peace at all levels. Democracy, as a mechanism to control power, is a force for peace from the micro-levels of interpersonal relations to the macro-levels of international relations. Does Carpenter not understand this? Rummel's theoretical work is very important and should have been engaged. It was not.
8. Carpenter tells us that Rummel's central target is "structural realists" who believe that "conflict is inherent in an international system that has no central authority and is made up of nation-states with conflicting interests." But this is far too vague: even idealistic pacifists would agree with it as a description (prescriptions are, of course, another matter). Rummel himself argues (The Conflict Helix, Transaction, 1991) that conflict is inherent in the international system, and will be even when the whole world is made up of democracies. But Carpenter makes the mistake of equating "conflict" with violence and war, and misses Rummel's distinction that while there will always be conflict, between democracies there isn't any war, and minimal violence--especially when compared to any other kind of government. So it isn't Carpenter's strawman definition of "realists" that Rummel is critical of, but rather the realist's insistence that the type of government a country has is irrelevant to its international relations. Does Carpenter believe this is true? If he does, he must present evidence to support this supposedly realist assertion, rather than just say the contrary isn't so.
9. Carpenter is either unaware of or willfully ignores much of the current work being done on the democratic peace. If he is not knowledgeable, he should have admitted his shortcomings, or not have written the review. And if he is familiar with the debate, then he himself is being perverse in not acknowledging it.
Carpenter referred to his former colleague Chris Layne's "important" article "Kant or Cant: The Myth of the Democratic Peace" in International Security (Summer 1994), yet failed to mention subsequent articles that refute Layne, such as Bruce Russett's extensive reply "Yet It Moves." Furthermore, he completely ignores Zeev Maoz's landmark article in the same prestigious publication, "The Controversy over the Democratic Peace: Rearguard Action or Cracks in the Wall" (Summer 1997). Maoz actually tests Layne's (and others) "critical cases," with findings that support the democratic peace over alternative explanations, such as "realism." Maoz also points out that Layne's argument (that "balance of power" explains near misses--when democracies did not go to war against each other--better than "regime type") does not lead to contradictory predictions than the democratic peace, but rather to complementary ones (realism isn't wrong, just woefully incomplete). So how can Carpenter say that realism leads to different conclusions? If Carpenter thought Layne's article was important enough to mention, why did he not refer to Maoz's even more important article?
10. Carpenter attacks Rummel for not dealing in depth with various historical cases. This may seem a fair point, but ignores Rummel's repeated references to the work of others who have done detailed historical analysis. James Lee Ray's extensive treatment of "'Wars' Between Democracies" (chapter 3 in his book Democracy and International Conflict: An Evaluation of the Democratic Peace Proposition, University of South Carolina, 1995), which Rummel refers to, comes to quite different conclusions than does Carpenter. And how could he have ignored the famous study, also referenced by Rummel, "Peace among Democratic and Oligarchic Republics" (Journal of Peace Research, 1994), by historian Spencer Weart, in which he stated that "A striking lack of wars between well-established democracies prevailed not only among modern states, but also among earlier regimes commonly described as democracies...."? Carpenter's tangential and gratuitous attack on the work of Bruce Russett and Michael Doyle (without even telling us which works he is attacking!) is grossly inadequate criticism. To attack Rummel for not having duplicated quite adequate historical work done by others is pretty weak. I assume Carpenter himself builds on the work of others, as do all scholars.
11. Carpenter continues his disproportional attacks on Rummel in his rejoinder, where he accuses Rummel of "creative historical revisionism," presumably on the basis of a disagreement about whether Germany under Wilhelm was "democratic." But there is plenty of valid historical analysis supporting Rummel on this ö so why the vitriolic taunting when a simple note of disagreement would have been in order?
12. Finally, this review has done a disservice to Rummel, but more importantly, the cause of liberty. A fair-minded reader must be perplexed about the vehemence of Carpenter's attack. What on earth is going on here? Sure, there is an on-going debate about the democratic peace, with one segment of assorted "realists" contemptuously dismissing an important and persuasive new movement in political science. But this review goes far beyond the methodological quibbles of scholarly discourse; it is an ill-tempered and intemperate broadside.
Curiously, Carpenter's attack on Rummel from the "realist right" is reminiscent of the vitriol cast on Rummel from the "pacifist left" for his call to strengthen US military forces in the late 1970s. Back then, the "higher purpose" that allegedly trumped democracy was anti-militarism. What is Carpenter's? Certainly it isn't liberty, or the defense of the US.
Perhaps this was meant to be leveled at someone "on the other side of the barricades." But Rummel's work--all of it--is a powerful defense of liberty. This is certainly reflected in the interview ("Rudolph Rummel Talks About the Miracle of Liberty and Peace") in the staunchly libertarian journal The Freeman: Ideas on Liberty (July 1977). Maybe Carpenter should get out and read more.
Carpenter's ideological cosmology seems so constricted that he sees only his concept of "realism" as congruent with the defense of liberty, which it clearly is not. Carpenter excoriates Rummel for attacking "realists" (which he indeed does), but fails to mention Rummel's libertarian conclusion about the democratic peace: "This solution has been around for centuries and in one form or another was integrated into the classical liberal view of government: the government that governs least governs best; and freedom promotes peace and welfare."
The relationship between democracy and peace, long hinted at and intuitively understood by great thinkers throughout history, from Thucydides to Kant to the Enlightenment classical liberals, has finally been proven rigorously, methodically, and unambiguously. It does not negate "realism" but incorporates and transcends it, just as Copernicus' views did Galileo's, and later Einstein transcended Newton's understanding of reality. Carpenter should join the 21st century.
*Written on 25 September 1998, San Diego, California. Carpenter's review appeared in The Independent Review (Winter 1998), and his rejoinder to Rummel's reply (Summer 1998). See the top right sidebar for links.
Go to top of document