Why do democratic states not fight each other? A systemic approach to the democratic peace

TABLE OF CONTENTS

Acknowledgements

Abstract

Introduction

0.1 Presentation of the subject
0.2 Methodology and literature review
0.3 Definitions

Chapter I The monadic level of analysis
1.1 Empirical research on the monadic level
1.2 The three main axioms and the rationalist approach
1.3 Structural constraints
1.4 Normative constraints
1.5 Synthetic approaches
1.6 Concluding words

Chapter II The dyadic level of analysis
2.1 Empirical research on the dyadic level
2.2 The shared democratic institutions and the credibility of signals
2.3 The shared democratic culture
2.4 A peace of deterrence
2.5 Concluding words

Chapter III The systemic level of analysis
3.1 Alternative approaches to the systemic analysis
3.2 International Organisations
3.3 The international organisation of democratic states prevents war
3.4 Concluding words

Conclusion
4.1 Summary of the argument
4.2 Critical outlooks

Bibliography

Books and chapters in books

Journal articles and papers

Internet resources

Abstract

There is no regularity in international relations that is as imperturbable as the democratic peace. From the beginning of the statistical research in 1816 until today, no clear-cut case of war between two democratic states has been recorded. The democratic peace has obstinately kept the secret of its causal mechanism. No convincing theory as to its cause has been widely accepted.

It is the aim of this dissertation to provide an alternative explanation for why democracies do not fight each other. Empirical research can only account for correlation but not for causal mechanisms. This dissertation thus concentrates on the theoretical explanations. Scholars developed approaches to account for the democratic peace ranging from constructivist through the sociological to game-theoretical methodology. They focus on the single democratic state, the relation between two democratic states and, recently, the international system itself. This dissertation critically examines a number of such theories which vary in methodology and focus. Especially, arguments by Russett, Doyle and Müller are given attention, but, to a greater or lesser extent, they are flawed or insufficient.

At the same time, this dissertation points out a number of special characteristics of democratic states of importance. Pulling those together, an approach is proposed based on the assumption that the international system itself bears a major responsibility for the democratic peace. Supporting an approach by Hasenclever, it is argued that international institutions set up by democratic states are especially capable of mitigating conflicts and thus prevent war. Together with the special features of their member-states, such organisations account for the peaceful behaviour of democracies.

Introduction

0.1 Presentation of the subject

Today, in the spring of 2006, nations around the globe engage in armed conflict with one another. Statistically, such wars between states are quite unlikely events. However, being of such magnitude, wars shape the international system to an unmatched degree. It is therefore not surprising that the scientific community is committed to investigate the reasons why they occur and methods for preventing such horrible events. One particular strain in this field of research is the democratic peace theory.

Historically speaking, wars are an ancient form of conflict resolution. Today’s democratic form of governing states has been established only recently. However, this particular form of government has proven to be the most efficient counter to armed conflict. A clear-cut case of war between two democratic states has yet to take place. It was as early as 1917 when the American President Wilson stipulated clearly and precisely that “A steadfast concert for peace can never be maintained except by a partnership of democratic nations.”^[1] A belief in the peacefulness of democratic states became manifest long before the academic discussion started. In hindsight, Wilson seemed to be right. In 2001, the then President of the European Commission, Prodi, stated that the EU is the single best conflict-preventing concept of the world. “Die EU ist das beste Konfliktverhütungsmodell der Welt”.^[2] The European Union as a union of democratic states succeeded in guaranteeing peace in the once heavily contested European mainland for over 50 years now. What had Prodi and Wilson in mind when they made their statements? Does the stable peace between democracies merely reflect the peaceful inner constitution of democracies? Can it be because of the strict built-in checks and balances? Can the special relationship between two democratic states account for it? Or is it, after all, an outcome of the international organisation of democratic states? It is the main focus of this dissertation to answer these questions to find why democracies do not fight each other.

It will be argued that a shared culture or similar norms and values amongst democratic states are insufficient to account for the peace between democratic nations. The democratic peace can only be explained in combination with the special organisation of democratic states in inter-democratic international organisations. This argument will be established over the course of three analytical chapters. Laying the foundation for it, the unique inner constitution of democratic states will be examined first. The special checks and balances democracies implement to control their executive branches and the democratic culture that rules out violence as a conflict resolution mechanism are discussed in the first chapter. Some scholars argue that democracies behave in a more peaceful way in general. It will be established that this viewpoint is highly disputed. This dissertation argues that the set-up of democratic states offers important conclusions but cannot account for the democratic peace on its own.

In a next step of the argument, the second chapter is dedicated to the special relationship between two democratic states. Unique mechanisms operate in this relation. Scholars propose that a culture of coexistence between two democratic countries enables them to resolve conflict peacefully. The democratic leaders’ mutual fear of war is supposed to be another factor contributing to the democratic peace.

These theories are analysed and it will be found that neither can thoroughly explain the democratic peace. However, the special relationship between democratic states still holds conclusions for the overall argument.

In a final step, chapter three introduces the systemic level. It focuses on the international organisation of democratic states. It is found that a special relationship between democratic international organisations and their member states is of major importance for the democratic peace. This chapter addresses the systemic level of the democratic peace with recourse to conclusions from the sub-systemic level of chapters one and two. In conclusion, it is argued that the democratic peace can most convincingly be explained by the inner constitution working in conjunction with the special relationship of democratic states to form a set of interrelations between international organisations and their democratic member states.

0.2 Methodology and literature review

This dissertation will critically examine propositions by scholars and add alternative approaches to find what explains best the democratic peace. It is designed as a critical literature review. The literature used has been chosen with the aim to provide a sensible overview over the scientific discussion and the different approaches to explain the democratic peace. The literature-selection places emphasis on theoretical approaches rather than empirical explanations. It is necessary to find causal mechanisms rather than empirical regularities. Therefore, it is the theoretical analyses of the democratic peace which will be in the focus of attention. As a work on the theory rather than on the empiricism of the democratic peace, this dissertation does not include any original empirical research. Nevertheless, the empirical foundation of the democratic peace is a prerequisite for the theoretical work. Thus, chapters one and two will include sections on empirical research, conducted by other scholars, to lay a ground for theoretical work in this dissertation.

In chapters one, two and three this dissertation uses comparative methods to analyse various books and articles as well as secondary literature on popular works. The strengths and weaknesses of the approaches will be] shown and balanced against each other. The research in the field of the democratic peace theory is broad and does not follow consistent conceptions. Scholars approach the puzzle of the stable peace between democratic states from very different angles. Some rely on empirical research or game-theoretical assumptions, such as Ray^[3] or Bueno de Mesquita.^[4] Others use sociological or constructivist methods and analyse behaviour and ideas of inhabitants and governments of democratic states, such as Russett^[5] or Müller.^[6] To examine critically all these different approaches in one coherent analysis, a regulative framework will be required. Therefore, this dissertation is based on a distinction between monadic-, dyadic- and systemic approaches.^[7] Such approaches may vary in their methodology (e.g. empirical or sociological) but their focus will be on one specific unit of analysis. However, no pre-selection has been made on whether one specific unit of analysis is most promising for the research.

Monadic approaches investigate the inner constitution of democratic states; the unit of analysis is the single state itself. Dyadic approaches refer to the special relationship between two democratic states. The unit of analysis here is the democratic dyad. The international system and the impact of international organisations on the democratic peace is the main focus of the systemic approach. Scholars who focus on this level use conclusions from other levels to underpin their argument but their unit of analysis is the international system. In the course of the dissertation, the reader will be made familiar with a number of works of influential scholars and their various ideas and methodologies. Some of the most important ones are as follows.

Russett has over years been one of the leading scholars of the democratic peace theory. He unites empirical and theoretical research in his books. Ideas by Russett are presented in every chapter of this dissertation as his main works “Grasping the Democratic Peace, Principles for a Post-Cold War World”^[8] and “Triangulating Peace. Democracy, Interdependence and International Organizations”^[9] include monadic, dyadic and systemic elements. Bueno de Mesquita favours a quite different methodology. He uses game-theoretical assumptions. Being original and convincing in his article “An Institutional Explanation of the Democratic Peace”^[10], he successfully manages to avoid fallacies other scholars fall for on a dyadic level. His ideas are utilised in chapters one and two of this dissertation.

The discovery of the systemic level for the democratic peace proposition happened only recently. Few works have been published so far. Interesting research has been set up by the Hessische Stiftung Friedens- und Konfliktforschung^[11], a German foundation. The empirical research is designed to find out about the reciprocity between international organisations and the democratic peace. The theoretical groundwork has partly been provided by the associated scholar Hasenclever in his article “The Democratic Peace meets International Institutions. Überlegungen zur Internationalen Organisation des Demokratischen Friedens“.^[12] Some of his ideas are illustrated in the third chapter. This dissertation follows the upcoming trend and awards explanatory power to the systemic level in its own right. Most traditional research has been preoccupied with the monadic, the dyadic level or a combination of both. So far, these approaches have been unable to produce a convincing explanation. It may well be that the systemic level adds the missing link.

0.3 Definitions

Whatever explanation one favours for the absence of war between democratic states, it is important to define the terms in question. In particular, these terms are democracy and war. However obvious the significance of these terms may seem, it is imperative to set a guiding definition to ensure that research is not conducted on completely different terms.

The definition used in this dissertation and the most common definition of the term war in scientific literature has been provided by Singer and Small in their “Correlates of War” project.^[13] To classify as a war, a conflict needs to take a toll of at least 1000 battle fatalities. This threshold rules out accidental conflicts and deliberate actions by local commanders. To qualify as a combating political unit a participant in the conflict must dispose of a minimum 1000 uniformed military personnel in active combat or suffer at least 100 fatalities. This definition is important to designate political units as participants in wars with more than two parties involved (as in coalition warfare). The Singer and Small definition captures all clear-cut wars and leaves out a few ambiguous cases. However, a problem arises with today’s conduct of war. Interstate wars only account for a mere ten percent of all armed conflicts. Singer and Small’s definition does not include guerrilla wars or genocide, which is what takes most lives. This will be kept in mind for a later assessment.

The term democracy is used to describe a form of control over the government by the public. In addition, democracy implies that a number of rights are granted to the population such as civil liberties, a high voting franchise, contested elections and separation of powers. As such and other features of an ideal democratic state can only be found to a certain degree, democracy needs to be defined in a more flexible way. This dissertation favours the “Freedom House” seven-point scale that ranges from truly democratic over partly democratic to autocratic.^[14] Many other definitions of democracy are used in the literature. The definitions used in the “Polyarchy Dataset” by Vanhanen serve as another popular guideline.^[15] In general, the research on the democratic peace is more robust against changes in definitions of democracy than of war. Thus it is common to use varying definitions of democracy in the literature as the effect is not dramatic. Whenever necessary, a comment will be given to explain the effect of such definitions on the various analytical levels.

Chapter I: The monadic level of analysis

The first level this dissertation analyses is the monadic level. Researchers who focus on individual states ask how the organisation of the particular democratic state affects its behaviour in international relations. Some scholars of the monadic analysis of the democratic peace claim that democracies behave in a more peaceful way in international relations than other states due to the very nature of the democratic organisation of society. Therefore, it might well be that democracies do not fight each other because they are peaceful in general. But such claims on the monadic level have thus far been highly controversial. Controversy develops mainly concerning two subjects: what is the real character of a democratic state in terms of its use of international force (empirical) and what is an appropriate explanation for its behaviour (theoretical).

The current scientific agreement is that democracies are (statistically) by no means more peaceful in international relations than are autocratic regimes, the democratic peace is a phenomenon between democracies. A minority has lately voiced doubts about this general statement. New empirical research^[16] has found that there are indicators of a slightly higher peacefulness of democracies over non-democracies in general. The basic empirical work on the monadic level of analysis will be briefly outlined at the beginning of this chapter. Differences do occur but an in-depth analysis of the structure and the empirical research design responsible for the varieties would be a step too far at this point. Considering that empirical research on its own may only account for observations but not for the underlying theoretical foundation, it is more important to evaluate the strengths and weaknesses of the explanatory approaches.

Despite the differences in empirical findings, monadic analysis offers, at least at the surface, plausible explanations for a more peaceful approach by democracies in international relations. In addition, the theoretical foundation laid down in the monadic theory is the basis for the given that democracies do, by and large, not fight each other. This dyadic finding, which is largely confirmed, is thus closely linked to the monadic level of analysis. Therefore, it is important to understand the main implications of the monadic argument(s) even if one would disagree with the assumption that the inner organisation of democracies accounts for their peacefulness.

1.1 Empirical research on the monadic level

Despite the rather robust empirical finding that the democratic-democratic dyad offers an extremely peaceful relation in the international realm, the monadic analysis of the genuine peacefulness of democracies is controversial. The scientific community offers various prudent empirical research. The outcomes nevertheless differ. Contemporary scholars like Henderson found that democracy “actually increases the likelihood of interstate war.”^[17] This rather extreme position finds few advocates. The most frequently adopted position is sobering: “Polyarchy has no independent effect on war-proneness.”^[18] Predominant from the 1940s until the beginning of the 1990s, this position has been supported by a wide range of scientists.^[19] A brief summary of works declaring no difference in behaviour between democracies and non-democracies can be found in Henderson.^[20]

The contemporary research offers a more optimistic view about the relation between peace and democracy. “[...] not only was there a separate peace among democratic states but democracies were more peaceful than autocracies generally.”^[21] Similar statements are given by Bruce Bueno de Mesquita and David Lalman^[22] or by Benoit.^[23] A comprehensive summary is offered by Ray.^[24] New approaches by Rousseau^[25] incorporate the question of which states initiate a military dispute. Democracies are more often the victim of aggression than initiators of disputes themselves. Still, a closer look at the process of bargaining in these conflicts needed to be taken to ensure that democracies do not just provoke more often. Taking into account the near past, the assumption of democratic states as victims of violence in international relations is very questionable.^[26]

Obviously, scholars find it hard to identify a coherent empirical framework to overcome inherent problems. Prominent researchers like Russett have already changed their minds in the process and are now supporting the optimistic position. As even the best empirical research cannot ensure that no mistakes slipped in, persuasive statistical results still need to be handled with care. In addition, no serious effort has thus far been made to rule out the assumption of a spurious relation between regime-type and conflict-proneness.^[27] Confounding variables such as wealth or stability may bias results.^[28] Furthermore, democracy is a matter of definition. The world is not divided into democracies and autocracies but many states share features of both systems in common, democracy is a “matter of degree rather than of kind”^[29]. Altering definitions may especially in the research on monadic analysis contribute to different outcomes.^[30]

It is therefore reasonable to assume that empiric research alone cannot offer a coherent analysis. The theoretical approach to the monadic level might offer further guidance in the search for an explanation of the phenomenon of the democratic peace.

1.2 The main axioms and the rationalist approach

Three main axioms form the body of most of the present theoretical monadic approaches. Firstly, it is assumed that the main preference of every political leader is to stay in office. Secondly, every human being is assumed to act in a rational way, guided by its preferences, trying to accumulate maximum gains with a minimized effort. Thirdly, the preferences have to remain constant, that is, assumptions are only valid ceteris paribus. Especially the last axiom has caused major critiques of various monadic explanations.^[31]

Kant already saw a constraint on leaders of a republic state deriving from public opinion, based on the assumption that the population is reluctant to go to war as they would have to carry the burden.^[32] Incorporated in this conclusion is the second axiom, the assumption of the homo economicus^[33], the rational human being. Given the preference of any administration is to stay in office, which is unquestioned throughout the literature, leaders of a democratic state need to take into account that their first goal has to be the accumulation of votes. Therefore, they are (at least at times close to an election) bound to a policy appreciated by the electorate. As the population is reluctant to go to war, so will be the administration. This simple rational-choice approach developed into a rationalist-utilitarian argument. Citizens want to avoid costs and risks. War entails severe risk to life and property. “A prudent citizen anticipates all these risks, compares it with the risks, costs and benefits of peace and checks them against the possible benefits of a war.”^[34] Expecting a rational behaviour, the citizens then will decide against war. Nothing can offset the price of death.

A difference in conflict settling behaviour between democracies and non­democratic state can be expected when taking this argument into account. Yet it is not sufficient to explain the peaceful behaviour of democracies because being coherent in general, the rationalist-utilitarian argument suffers from being highly dependent on the validity of its axioms. How far changed preferences and a different calculation of costs and gains by the citizens^[35] can turn this argument upside down will be discussed in a later part of this chapter.

[...]

^[1] W. Wilson in his congressional address “War Message”, delivered 2 April 1917 in Washington D.C., United States.

^[2] R. Prodi in an interview with the German newspaper Frankfurter Allgemeine Zeitung in the 22 November 2001 issue, p. 14. Both quotations are being taken from A. Hasenclever, “The Democratic Peace meets International Institutions. Überlegungen zur Internationalen Organisation des Demokratischen Friedens“, Zeitschrift für internationale Beziehungen, 9, 1, 2002, pp. 75-111.

^[3] See for example J. Ray, “Democracy. On the Level(s), Does Democracy Correlate with Peace?”, in J. Vasquez, What Do We Know About War? (London: Boulder, 2000), pp. 299-316.

^[4] See for example B. Bueno de Mesquita & J. Morrow, R. Siverson, A. Smith, “An Institutional Explanation of the Democratic Peace”, American Political Science Review, 93, 4, 1999, pp.791-807.

^[5] See for example B. Russett, Grasping the Democratic Peace, Principles for a Post-Cold War World (Princeton: Princeton University Press, 1993).

^[6] See for example H. Müller & J. Wolff, “Dyadic Democratic Peace Strikes Back”, paper prepared for the 5th Pan-European International Relations Conference, The Hague, 9-11 September, 2004.

^[7] This distinction is sometimes artificial in the sense that arguments use assumptions from more than one level of analysis. Such arguments will be divided into parts in a sensible way to fit into the framework of this dissertation.

^[8] B. Russett, Grasping the Democratic Peace, Principles for a Post-Cold War World (Princeton: Princeton University Press, 1993).

^[9] B. Russet & J. Oneal, Triangulating Peace. Democracy, Interdependence and International Organizations (New York/London: Norton, 2001).

^[10] B. Bueno de Mesquita et al, “An Institutional Explanation of the Democratic Peace”.

^[11] The foundation’s webpage in German can be found at http://www.hsfk.de/index.php?language=de (accessed 12th April 2006). The particular research-design can be found at the program’s webpage at http://www.hsfk.de/group.php?id=4&language=de (accessed 13th April 2006).

^[12] A. Hasenclever, “The Democratic Peace meets International Institutions. Überlegungen zur Internationalen Organisation des Demokratischen Friedens“.

^[13] The webpage of this empirical research project can be found at http://www.correlatesofwar.org (accessed 14th April 2006).

^[14] The webpage of Freedom House and the scale can be found at http://www.freedomhouse.org/template.cfm?page=1 (accessed 15th April 2006).

^[15] An outline of this project can be found at the Finnish Social Science Data Archive at http://www.fsd.uta.fi/aineistot/taustatietoa/FSD1289/Introduction_Tampere.pdf (accessed 15th April 2006).

^[16] For recent work see K. Benoit, “Democracies Really Are More Pacific (in General)“, Journal of Conflict Resolution, 40, 12, 1996, p. 636-657; also see B. Russet & J. Oneal, Triangulating Peace. Democracy, Interdependence and International Organizations. For preceding work see R. Rummel, “Libertarianism and International Violence”, Journal of Conflict Resolution, 27, 1983, p. 27-71.

^[17] E. Henderson, Democracy and War. The End of Illusion? (London: Boulder, 2002) p. 73.

^[18] B. Russett & R. Monsen, “Bureaucracy and Polyarchy as Predictors of Performance: A Cross-National Examination”, Comparative Political Studies, 8, 1, 1975, p. 27.

^[19] See S. Chan, “Mirror, Mirror, on the Wall...Are the Freer Countries More Pacific?”, Journal of Conflict Resolution, 28, 1984, p. 617-648; also see E. Weede, “Democracy and War Involvement”, Journal of Conflict Resolution, 28, 1984, p. 649-664.

^[20] See E. Henderson, Democracy and War. The End of Illusion?.

^[21] B. Russett & J. Oneal, “The Classical Liberals Were Right: Democracy, Interdependence, and Conflict, 1950-1985, International Studies Quarterly, 6, 1997, p. 288.

^[22] See B. Bueno de Mesquita & D. Lalman, War and Reason (New Haven: Yale University Press, 1992).

^[23] See K. Benoit, “Democracies Really Are More Pacific (in General)“.

^[24] See J. Ray, “Democracy. On the Level(s), Does Democracy Correlate with Peace?”.

^[25] D. Rousseau, Domestic Institutions and the Evolution of International Conflict, Ph.D. dissertation, University of Michigan, 1996.

^[26] See J. Levy, “The Study of War and Peace”, in W. Carlsneas & T. Risse, B. Simons, Handbook for International Relations (London: Sage Publications, 2002) pp. 350-368.

^[27] In a rather creative attempt, Russett analysed pre-industrial societies for their involvement in conflict. He found that the higher the level of participation and public discussion in the political process was, the less likely the society was involved in armed conflict. See B. Russett, Grasping the Democratic Peace, Principles for a Post-Cold War World, chapter three.

^[28] E. Mansfield & J. Snyder, “The Dangers of Democratization”, International Security, 20, 1995 p. 5­38 and E. Mansfield & J. Snyder, “ The Effects of Democratization on War: The Authors Reply”, International Security, 20, 1996, p. 196-207 offer analysis about confounding variables, finding that stability is a major but underrated aspect of the monadic analysis.

^[29] B. Russett & J. Oneal, Triangulating Peace: Democracy, Interdependence, and International Organizations, p. 45.

^[30] See B. Russett & J. Oneal, Triangulating Peace: Democracy, Interdependence, and International Organizations, pp. 44-45, empirical findings of dyadic based analysis are considerably more robust.

^[31] See H. Müller, “The Antinomy of Democratic Peace“, paper prepared for the third ECPR Meeting in Marburg, Germany, 2003.

^[32] See I. Kant, Zum Ewigen Frieden: Ein Philosphischer Entwurf (Stuttgart: Reclam, 1984, 1795).

^[33] The assumption that a homo economicus is likely to avoid war is taken up by E. Czempiel, “Kants Theorem. Oder: Warum sind die Demokratien (noch immer) nicht friedlich?“, Zeitschrift für internationale Beziehungen, 3, 1, 1996, p. 80. and B. Russett & J. Oneal, Triangulating Peace: Democracy, Interdependence, and International Organizations, pp. 272-273.

^[34] H. Müller, “The Antinomy of Democratic Peace”, p. 5

^[35] Müller stresses the importance of definitions of costs and gains this in a recent article. See H. Müller & J. Wolff, “Dyadic Democratic Peace Strikes Back”, p. 8.