In contemporary international relations (IR) theory, there is perhaps no subject more contested than the democratic peace. The origins of democratic peace theory can be traced back to Immanuel Kant’s Perpetual Peace, in which he argues that democracy has a pacific effect on states. Specifically, he argues that because the people bear the burdens of war, a state whose government is directed by the people will be more peaceful than a state whose government is autocratic. In the last twenty years, Kant’s simple proposition has been debated at length. Modern political scientists have studied the democratic peace using many methodologies and have come up with innumerable explanations for the supposed phenomenon. Dyadic democratic peace theory, according to which democracies are only peaceful in relation to one another, is the most prominent version of democratic peace theory. Monadic democratic peace theory has been largely rejected by analyses showing that democracies, overall, fight wars almost as often as autocracies. Most work on the democratic peace comes from quantitative studies, which are useful starting points for the formation of a comprehensive theory. At the same time, however, this methodology has major limitations. This paper will discuss the pros and cons of both quantitative methods and historical case studies and show that the preferable methodology for studying the democratic peace is not one or the other, but a careful reconciliation of the two. The result of the discussion will be a revitalization of monadic peace theory in probabilistic terms. First, however, the paper will discuss why finding the truth about the democratic peace matters in the first place.
What is at Stake
One reason that this debate is important is immediately apparent: it has profound implications for policymaking. A leader who believes in the democratic peace is likely to encourage the spread of democracy around the world. This conviction might even lead the leader to use military force. Modern American neo-conservatism is clearly grounded in notions of the democratic peace. Liberals, while disinclined from unilateral and more militant policies, also often support policies that expand democracy around the world.
Policymakers who reject the validity of the democratic peace might have very different foreign policy objectives. If democracy has no bearing on a state’s peacefulness, already democratic states should not seek regime change or democratization for the sake of peace abroad, but instead only pursue their strategic interests. This thinking is characteristic of realism. In an anarchistic global order of rational states, regime type does not matter because all states compete for the same things: relative power and security. A realist policymaker would therefore reject democratic peace theory and execute a foreign policy concerned only with ensuring her state’s security and augmenting its relative power.
Here, the crux of the significance of the democratic peace is apparent. To liberalism, states are not identically rational, but distinguishable from one another based on regime type, among other factors. Moreover, states are not doomed to compete in the anarchic international system. Anarchy can be mitigated, if not overcome, by factors such as economic interdependence, international law, and democracy. Though Sebastian Rosato claims to have discredited democratic peace theory in his article, “The Flawed Logic of Democratic Peace Theory,” he acknowledges that, if true, “it undermines both the realist claim that states are condemned to exist in a constant state of security competition and its assertion that the structure of the international system, rather than state type, should be central to our understanding of state behavior” (Rosato 2003, 585). The controversy over democratic peace therefore centers on the fact that its validity would greatly discredit realism while its invalidity would strongly reaffirm realism’s fundamental claims about international politics.
An Evaluation of Methodologies
Due to this debate’s high stakes, an evaluation of the methodologies used to advance democratic peace theory is warranted. There is not one universally accepted methodology for studying the theory, and so the task of selecting a methodology is an additional complication for scholars attempting to assess the democratic peace. On one level, the democratic peace can be presented as an empirical observation; through multiple quantitative studies conducted for the last 150 years, scholars have shown that democracies nearly never fight one another. It appears that the dyadic democratic peace is an empirical reality. Although this is not uncontested by any means, the larger debate lies in finding the true causal mechanism for this correlation. Because the very existence of the democratic peace is established based on quantitative studies, such large, statistical studies are often also favored in the search for true causal mechanisms.
On Quantitative Analyses
Quantitative studies, especially those with large sample sizes, can appear to demonstrate “laws” of social science similar to the “laws” discoverable by similar methods in the “harder” or “natural” sciences. Historical case studies, on the other hand, are by nature limited in scope, and extrapolating universal truths from them is therefore more difficult. Quantitative studies are useful for discovering trends over periods of time and finding interesting correlations between variables. They do, however, have important shortcomings one must consider. In the introduction to their book Democratic Wars, Anna Geis, Lothar Brock, and Harald Müller state that such “statistical tests do not inquire into causal mechanisms, they establish correlations that can plausibly be interpreted as causation.” They point out that although “quantitative studies have produced a rich store of statistical data which spell out regularities in the behaviour of democracies; their pay-off… is limited when it comes to explaining war” (Geis, Brock, and Müller 2006, 4). Quantitative studies highlight correlations between independent and dependent variables but they cannot explicate how, or whether or not they are causally related. Scholars who use such a methodology must theorize about the causal mechanism based on the correlations they find, but considerable quandaries arise from this social-scientific method.
One such quandary is the controversial assumptions researchers make in order to perform quantitative analyses to begin with. In evaluating a large dataset of conflicts between states, it is necessary to make certain assumptions about the states involved and code variables in particular ways. Slight variations in assumptions and codings can make a huge difference for the findings of quantitative studies. In his article “How Smart and Tough Are Democracies?” Alexander Downes demonstrates that a relatively slight adjustment in assumptions can produce new, contradictory findings. Downes examines Dan Reiter and Allan Stam’s discovery, presented in their book Democracies at War, that democracies “win nearly all of the wars they start, and about two-thirds of the wars in which they are targeted by other states” (Downes 2009, 9). Downes argues that Reiter and Stam’s codings and assumptions are flawed. While they code states as either targets or initiators, Downes reasonably claims that a third category of states, joiners, should be included. These are states that get involved in wars only after initial proceedings, often a significant variation from being either the target or initiator of war. In addition, he asserts that military conflicts that end in a draw—omitted by Reiter and Stam—should be included in the dataset (Downes 2009, 10-11). By making these changes and then computing the statistics over again, Downes arrives at an entirely new conclusion: “[I]n statistical terms the substantive effects of democracy are indistinguishable from zero, meaning one cannot reject the null hypothesis that democracy has no effect on the likelihood of victory and defeat for any type of belligerent” (Downes 2009, 26-28). It is therefore clear that the assumptions and codings one uses to quantitatively analyze the democratic peace are of great significance. Downes’s article does not deny that democracy affects the probability of winning a war, but that quantitative studies are inherently problematic because their findings can be greatly disrupted by relatively slight adjustments in assumptions.
In addition to suffering from problematic assumptions and codings, quantitative studies can fail to describe certain important events, making the causal theories that accompany the studies less valuable. Such studies often fail to explain particular events because of simplifications in the way those events are described by the social-scientific method. For example, the democracy/autocracy dichotomy is much too simplistic to describe states in the real world. Even the sliding scale between most democratic and most autocratic that many IR scholars use can be insufficiently sophisticated. Moreover, some factors that influence outcomes, such as the magnitude of each military conflict, alliances, geographical location, access to resources, etc., are always necessarily ignored, even in multivariable analyses. Because of such simplifications, findings are frequently prone to historical counterexamples.
Because quantitative studies seek to prove generalizations and are not particularly concerned with the details of a given military conflict, they would appear immune to individual counter-examples. Presented with a case that contradicts the finding of a quantitative study, a researcher could dismiss the challenge as an unrepresentative exception. However, Geis, Brock, and Müller assert, “if the theory has any validity, it should stand the test of ‘salient cases’ as well as random evidence” (Geis, Brock, and Müller 2006, 5). These authors define salient cases as “those major events involving the use of military force that have a decisive impact on the course of history” (Geis, Brock, and Müller 2006, 5). Importantly, most quantitative studies involving large numbers of military conflicts do not have a mechanism that weights the magnitude of each conflict. Therefore, a counter-example to a theory may be a salient case in reality while of only minimal significance to the results of a quantitative study. It seems logical that the Barbary Coast War should carry less weight than a more salient case such as World War One, in a study of military conflict. It should be more important that a theory accurately describe the latter than that it describe the former. Therefore, in a comprehensive quantitative analysis, the theory posited should be tested against salient case studies.
On Historical Case Studies
Though extensive quantitative studies have their faults, historically specific case studies are not necessarily better. In-depth examinations of individual cases also have their limitations. Such studies are valuable in that they highlight exactly how variables interact in a given instance, so that it becomes clear how they work in reality and not just in theory. While quantitative studies are prone to over-simplification and often contain weak assumptions, historical case studies tend to be too focused on individual cases to be representative of a general trend in state behavior. It is difficult, if not impossible, to extrapolate universally valid theories of state behavior from a single or even several case studies.
Nonetheless, Robert Ivie and Oscar Giner seek to do precisely this in their article, “Hunting the Devil: Democracy’s Rhetorical Impulse to War.” The authors argue that there is a demonic side to democracy, “an impassioned ogre of mob violence…who reflects the common fear and shared anxiety about democracy,” and that this aspect of democracy is prone to exploitation by leaders desiring to go to war (Ivie and Giner 2007, 581). The article calls into question the pacific influence of democracy on states because it asserts that democracy has an inherently demonic quality that can make democratic states more warlike. Ivie and Giner make this argument with exclusive reference to the United States and rely almost entirely on President Bush’s war rhetoric before and during the Second Gulf War for evidence. Citing examples of Bush’s moralistic rhetoric about the war effort, the authors conclude “democracy thus served ambiguously as war’s purpose and provocation” (Ivie and Giner 2007, 584). While this article perhaps provides some interesting insight into how democracy might permit war mongering, it is ill equipped to theorize about democracy’s effect on states in general. Ivie and Giner do not give sufficient reason to believe that this phenomenon exists outside of the United States or perhaps even outside the case of the invasion of Iraq in 2003. It could very well be the case that the United States is an outlier in this case when it comes to democratic states and is therefore not a representative case.
Methodological Synthesis and Monadic Democratic Peace
An ideal methodology would take into consideration the pros and cons of both quantitative and historical outlooks. It would include a quantitative study that is considerably broad in scope so that the resulting theory is widely applicable, but also be refined so that it acknowledges and accounts for the important differences between democratic states. Moreover, the theory derived from the quantitative study would need to be tested against salient cases in order to demonstrate that it can explain the most prominent instances of military conflict.
The two types of methodologies explained above, when combined, actually provide a basis for belief in monadic democratic peace theory. The most well known findings from purely quantitative analyses suggest that democracies rarely, if ever, fight one another, but that overall they fight just as many wars as non-democracies. Nevertheless, many critics of democratic peace theory doubt the existence of the dyadic peace based on individual cases that contradict the quantitative literature’s finding. For example, in his article “The Flawed Logic of Democratic Peace Theory,” Rosato uses the example of American interventions against democratic countries during the Cold War as evidence against the dyadic democratic peace. He states, “American interventions to destabilize fellow democracies in the developing world provide good evidence that democracies do not always treat each other with trust and respect when they have a conflict of interest” (Rosato 2003, 590). If there were a dyadic democratic peace, the United States would not have intervened so frequently in other democracies. However, it is actually possible to reconcile these two seemingly opposed views and in doing so support a monadic democratic peace in probabilistic terms.
In “The Antimony of Democratic Peace,” Harald Müller reveals that only a handful of democracies account for a vast majority of military conflicts involving democracies “at a level of 7 and higher on the combined autocracy/democracy scale of Polity IV,” a highly democratic rating on a commonly used sliding scale measurement of how democratic or autocratic a state is (Müller 2004, 495). He finds, through a statistical study of the period 1950-2001, that “of the 283 discrete [military] involvements by…stable democracies, 75.6 percent of these were carried out by just four countries (19 percent of the whole group): Israel, the United States, India, and the United Kingdom” (Müller 2004, 495). His finding suggests a distinction between those democracies that do the vast majority of the fighting and those that almost never fight. His assumption that only democracies with scores above a certain level on the Polity IV scale matter is perhaps assailable, but the point remains that some democracies do much more fighting than others.
It therefore seems logical to consider what differentiates democracies that fight from those that do not. Müller suggests that there are specific reasons why each democratic country is a militant or pacific democracy. This, however, fails to counter Rosato’s criticism of the dyadic democratic peace. Looking carefully at each of the four warlike democracies listed above, a theoretical explanation for the distinction between warlike and pacific democracies is not forthcoming, so long as democracy is the only influence on state behavior considered. The problem of American interventions against fellow democracies during the Cold War is left unresolved on a theoretical level.
A monadic democratic peace is possible if we consider democracy as only one of several influences affecting states’ foreign policy. Such an understanding of the democratic peace would be more probabilistic than absolute. Instead of making an argument for how democracy can uniquely influence a state to be peaceful in a democratic dyad and potentially warlike in a mixed dyad, it is possible to accept the simpler argument that democracy has a ubiquitously peaceful affect on states while affirming the “counterexamples” to traditional dyadic peace theory by recognizing that at least a few democratic states are militarily inclined, even against other democracies in rare instances. This is true because there are other influences on a state’s behavior that, depending upon the democratic state in question, can overcome the peaceful influence of democracy and cause the state to act belligerently in the international system. This hypothesis supports a probabilistic monadic democratic peace because it asserts that democracy is always a peaceful influence on states. If democracy is the dominant influence for a state, that state will act peacefully. Moreover, this hypothesis accounts for the differences between democracies and acknowledges that, as Bruce Russett and John Oneal argue in their book Triangulating Peace, “it is important to situate each country within its politically relevant environment” (Russett and Oneal 2001, 50). There are many influences on a given state’s foreign policy, but there are several prominent influences that stick out in addition to regime type.
Two factors stressed by realists, geographic location and relative power, are examples of other leading influences on a state’s foreign policy and may either reinforce or counteract the peaceful influence of democracy. Geographic location is significant because as the distance between two states increases, they are less likely to get involved in a military conflict with one another. Russett and Oneal give two reasons why prospects for peace improve as distances between states increase. One is that “it is hard to exert great military power at a substantial distance.” The second is that “widely separated states” have little to go to war over as “the great majority of international wars arise over territorial issues, chiefly the location of a disputed border… or the ownership of valuable resources” (Russett and Oneal 2001, 86). Relative power is also important because great powers are more war-prone than weaker states. In his article “Democracy, War and Expansion through Historical Lenses,” Russett explains of great powers that, “whether democratic or autocratic, their widespread interests and intervention capabilities allow them to be militarily effective beyond their geographical neighborhoods, and so they are more likely than small states to get into conflict” (Russett 2009, 14). These realist influences demonstrate that states’ decisions about war and peace can be influenced by factors unrelated to regime type.
By considering influential factors affecting foreign policy besides regime type it is possible to make sense of the war-proneness of democratic states such as the four considered by Müller to be the most militant. According to the hypothesis stated earlier, the militancy of those four states does not threaten the validity of monadic democratic peace, nor does it suggest either democracy’s pacific or militant influence. Instead, it implies that each of those four states have other influences that have counteracted the pacific effects of democracy on a number of occasions. For the United States, it is plausible that its status as the world’s superpower for the last 65 years has had a larger influence on its foreign policy than it being democratic. Because it has interests in every corner of the world and has sought to protect them and retain its superpower status, the United States has had many more opportunities to undertake military conflicts than weaker democratic states. Perhaps of even greater significance is that the United States was engaged in the Cold War against the only other potential global hegemon, the USSR. This rivalry certainly outweighed the influence of democracy when it came to the foreign policy of the United States during the Cold War. Similarly, the United Kingdom is one of the most powerful countries in the world. For much of the twentieth century Great Britain sought to maintain its great power status while it slowly and painfully released its colonial possessions. For Israel and India, geographic location has played a larger role in their war-proneness. Israel is surrounded by autocratic countries and militant non-state actors that have opposed its very existence. India is in a difficult geographic neighborhood with many autocratic states, as well as its bitter rival Pakistan, a nominally democratic state. India and Pakistan have had a history of fierce military conflict due to their religious differences and the virulent nature of their partition. For each of these four states, there are powerful geopolitical influences other than democracy that have counter-acted and overcome the pacific effects of democracy in many instances. From Müller’s analysis we know that these four states are responsible for the vast majority of military conflicts involving democracies, which implies that for most democracies, the realist influences mentioned above either work in tandem with democracy to have a pacific influence on state behavior or are insufficiently salient to overcome democracy’s influence on foreign policy. Simply put, democracy is not a deterministic factor when it comes to state behavior in relation to matters of war in peace. It is, however, a powerful pacific influence on state behavior.
Democratic peace theory is difficult to assess. That so much attention is paid to the subject is indicative of its significance. Whether or not there is a democratic peace has grave consequences for policymaking, but also for the equilibrium in IR theory between realism and liberalism. If there is a democratic peace, the foundational assumptions of the realist school of thought are severely undermined and liberalism should surge to the forefront of IR theory. If democratic peace theory is invalid, realism triumphs over its greatest challenge from liberalism in the last twenty years. The methodologies used to analyze the topic are therefore of great significance. Largely, studies on the democratic peace are either statistical or historical. By considering the weaknesses of each methodology, it is possible to analyze the democratic peace and conclude that there is a probabilistic monadic democratic peace in the sense that democracy will always increase the prospects for peaceful state behavior as it has a ubiquitously peaceful effect. There are, however, other influences that can overcome the effect of democracy in certain cases, causing a state to be more warlike than democracy would, by itself, allow. This way of conceptualizing the democratic peace, with democracy as one of several influences on state behavior, is more descriptive of state differences and is therefore an improvement on theories derived from simplistic quantitative models. Moreover, it is preferable to theories based on historically specific case studies because it is universally applicable and not based on an individual case that could be labeled an exception. Instead of fully flushing out a new theory, the hypothesis presented here constitutes a proposal that both statistical and historical case studies are needed to formalize a proper theory about the democratic peace.
1. Brock, L., Geis, A., & Müller, H. 2006. Democratic Wars: Looking at the Dark Side of Democratic Peace. Palgrave Macmillan, New York.
2. Downes, A. B. 2009. ‘How Smart and Tough Are Democracies?’ International Security, vol. 33, no. 4, pp. 9-51.
3. Giner, O. & Ivie, R. 2007. ‘Hunting the Devil: Democracy’s Rhetorical Impulse to War’, Presidential Studies Quarterly, vol. 37, no. 4, pp. 580-598.
4. Müller, H. 2004. ‘The Antimony of Democratic Peace’, International Politics, vol. 41, no. 4, pp. 494-520.
5. Rosato, S. 2003. ‘The Flawed Logic of Democratic Peace Theory’, American Political Science Review, vol. 97, no. 4, pp. 585-602.
6. Russett, B. 2009. ‘Democracy, War and Expansion through Historical Lenses’, European Journal of International Relations, vol. 15, no. 1, pp. 9-36.
7. Russett, B & Oneal, J. 2001. Triangulating Peace: Democracy, Interdependence, and International Organizations, W. W. Norton & Company, New York.
8. The White House Archives. November 11, 2003. “President Bush Discusses Iraq in Veterans Day Address.” <http://georgewbush-whitehouse.archives.gov/news/releases/2003/11/20031111-10.html>