Claim 1: democracy reduces military conflict

Liberals predict that democratic states are better able to resolve international disputes without resorting to military force than non-democratic states. Realists disagree and predict that states will balance (e.g. increase defence spending or establish alliances) against all stronger states because these powerful agents represent a threat to a state residing in anarchy. For realists, democracies will behave just like autocracies: they will balance against the strong and use force if the situation calls for it.

Early empirical research on the behavior of democracies seemed to confirm the realist predictions. Wright (1942: 841) concluded that regime type has little impact on the frequency of war because democracies possess attributes that both encourage and discourage war. In an early statistical analysis of the relationship between war and regime type, Small and Singer (1976: 67) concluded that democracies had not been noticeably peaceful over the 1816–1965 period. In the following decade, Chan (1984) and Weede (1984) reached a similar conclusion using quantitative analysis techniques and large cross-national times-series data sets. Although some evidence supporting the democratic peace emerged (Babst 1972; Rummel 1983), the realist position reflected the general consensus in the early 1980s.

The realist consensus came under attack in a series of articles by Doyle (1983; 1986). Doyle reframed the debate by looking at the characteristics of both the initiator of conflict and the target of conflict. After compiling a list of liberal societies from 1700 to 1982 and a list of interstate wars from 1816 to 1980, Doyle found that no two democracies had engaged in a full scale war against one another. He concluded that ‘liberal states have created a separate peace, as Kant argued they would, and have also discovered liberal reasons for aggression, as he feared they might’ (Doyle 1986: 1151). Doyle’s path-breaking work triggered an avalanche of studies on the democratic peace. According to Levy (1988: 662), ‘the absence of war between democratic states comes as close as anything we have to an empirical law in international relations.’

Although there have been some critiques of the claim (e.g. Layne 1994; Gowa 1999; Oren 2003), none has challenged the finding that no two established democracies have gone to war. Most of the empirical analysis has centered on the causal mechanisms: why do democracies behave differently toward one another? When a dispute erupts between two democracies, each side knows that the other faces domestic constraints on the use of force. This expectation limits bluffing, dampens spirals of hostility, and slows the mobilization process. Extensive empirical analysis has been produced for the dyadic, inter-democratic peace (Babst 1972; Doyle 1983, 1986; Maoz and Russett 1993; Russett and Oneal 2001; Rousseau 2005).

A second explanation falls at the state level of analysis and claims that democracies are more peaceful regardless of the opposition (referred to as the ‘monadic’ democratic peace). Here the causal mechanism does not focus on expectations about the behaviour of the other party in the dispute. Rather, democracies are less likely to initiate disputes and escalate crises because they are constrained by domestic institutions and norms of conflict resolution. The existence of domestic political opposition makes democratic leaders more risk-averse because foreign policy failures (and even costly successes) can be politically costly (Morgan and Campbell 1991; Morgan and Schwebach 1992; de Mesquita and Siverson 1994). Although the early research did not provide much support for this monadic argument, a number of more recent studies have produced strong statistical evidence in support of the hypothesis (Schultz 2001; Huth and Allee 2002; Bueno de Mesquita et al. 2003; Bennett and Stam 2004; Rousseau 2005).

The most persistent critiques of the democratic peace come from those claiming capitalism provides a better explanation (Mousseau 2000; Gartzke 2007). Free market attitudes and integrated economies may explain the peace between democratic states better than the democracy variable. Mousseau (2009), for instance, argues that ‘contract-intensive’ economies will generate peaceful interaction with similar economic systems. These claims, however, have been challenged by Dafoe et al. (2013). Capitalist peace scholars also cannot easily explain the destructive wars fought between capitalist states engaged in high levels of trade, notably Germany and Britain in 1914. While capitalism is an element of peace, as liberals predict, it does not discredit the strong empirical evidence linking the expansion of political liberty to a reduction in interstate conflict.

Claim 2: economic interdependence reduces military conflict

Economic interdependence is traditionally defined as the degree to which two (or more) states are connected by flows of goods, services, capital, labor, and technology. Liberals predict that these flows will encourage peace. Realists predict that economic interdependence increases the probability of conflict by expanding the number of issue areas under competition (Waltz 1979: 138). For example, Gaddis (1986: 110) argues that economic isolation between the East and West contributed to the ‘long peace’ during the Cold War.

Liberals posit several causal mechanisms in this relationship. First, decision-makers will calculate the costs of escalating a dispute with a trading partner. If two states are highly interdependent, decision-makers will be less likely to use force. Second, as Kant emphasized, firms and workers benefiting from international trade and investment will pressure government representatives to de-escalate disputes that arise between trading partners. Although research has not decisively disentangled these distinct (but complementary) causal mechanisms, the empirical literature provides significant evidence supporting the interdependence claim (Wallensteen 1973; Gasiorowski 1986; Polachek and McDonald 1992; Mansfield 1994; Russett and Oneal 2001).

Some scholars have qualified the liberal interdependence claim by specifying conditions that restrict the scope of the claim. For example, Keohane and Nye (1977: 10) make a clear distinction between symmetrical interdependence (i.e. both states are equally dependent on each other) and asymmetrical interdependence (i.e., state A is very dependent on state B, but state B is not very dependent on state A). While symmetrical interdependence creates a mutual desire for continued trade and investment, asymmetrical interdependence invites attempts to exploit weakness and manipulate behaviour (also see Hirschman 1945). In a similar vein, Copeland (1996) argues that the expectation of continuing trade is the key conditional variable. Only if state leaders expect trade to continue (or increase) are they less likely to use force. Finally, Ripsman and Blanchard (1996/7) contend that interdependence should only inhibit conflict if the trade involves strategic goods (e.g., oil or nitrates) which cannot be supplied from alternative sources. If substitute goods or markets are readily available, the cost of disrupting the relationship can fall dramatically.

Some argue that the empirical relationship between trade and conflict may be spurious. For example, Gartzke (2007) provides statistical evidence that market openness rather than trade interdependence (or democracy) reduces violence in the post-Second World War era. Similarly, Kim and Rousseau (2005) find that the pacifying impact of trade evaporates when using several different measures of interdependence and a model of reciprocal causation (i.e., simultaneously testing two claims: military conflict decreases trade AND trade decreases military conflict). It should also be noted that the findings from qualitative case studies are often inconsistent with the theoretical expectations. For example, Ripsman and Blanchard (1996/7) find little concern for the costs of interdependence in their analysis of historical crises among great powers during the July Crisis in 1914 and the Remilitarization of the Rhineland Crisis in 1936. Combined with the findings of the capitalist peace, significant empirical evidence supports the liberal claim that trade and interdependence are associated with peace. These findings, however, are less robust than the democratic peace claim.

Claim 3: international institutions reduce military conflict

The third pillar of the liberal peace claims that international institutions decrease the probability of conflict. In contrast, realists tend to view international institutions as either generally ineffective or the instruments of powerful states (i.e., international institutions have no independent causal impact, Mearsheimer 1994/5; Organski 1968). Although most realists and liberals would agree that the number of international institutions has grown exponentially over the last 100 years (e.g., Shanks et al. 1996), they disagree over their impact.

Early studies focused on formal ‘international organizations’ such as the UN. However, over time the research programme expanded beyond analysis of rules, procedures, and outcomes within formal institutions (e.g., UN voting patterns) and toward ‘international institutions’ more generally, including broad conceptualizations of ‘global governance.’ For example, Lipson’s analysis of the banking sector’s response to the debt crisis in the 1980s emphasizes the ‘informal’ regime created by banks seeking cooperation with each other (Lipson 1986). Following the lead of Mearsheimer (1994/95: 9), we can define international institutions broadly as a set of rules that govern how actors cooperate and compete with each other within an issue area (see Simmons and Martin 2002). These rules govern behavior in formal institutions such as the World Trade Organization (WTO) and informal institutions such as the debt-crisis banking regime.

These institutions promote cooperation in a wide range of issue areas, from trade and the environment to human rights and gender equality. In the more restricted domain of peace and security, how do international institutions promote peace? First, collective security organizations and alliances can promote peace by deterring aggression or intervening to halt a conflict (Huth and Allee 2002: 278). Second, international and regional institutions can mediate disputes (e.g., the good offices of the UN secretary general) or provide for arbitration (e.g. International Court of Justice). Third, international institutions can monitor compliance with agreements and reduce transaction costs for follow-up accords (Keohane 1984). Fourth, international institutions can promote conflict-reducing norms and alter identities and related interests (Wendt and Duvall 1989; Barnett and Finnemore 1999). Many international institutions reduce conflict through several of these mechanisms simultaneously. For example, the WTO and the Organization for the Prohibition of Chemical Weapons (OPCW) adjudicate disputes, monitor compliance, encourage norms (e.g., promoting economic liberalism in the WTO and banning weapons of mass destruction in the OPCW), and alter cost-benefit calculations (e.g., increasing trade ties in the WTO and collectively punishing of defectors in the OPCW).

Although this pillar of the liberal peace has produced mixed results, extensive evidence supports predictions that international institutions reduce military conflict (Oneal et al. 2003). Despite the intensity of the superpower conflict during the Cold War, new international institutions helped foster a degree of cooperation. In some cases, the link between international institutions and conflict was quite direct. For example, studies have highlighted the role of the International Atomic Energy Agency (IAEA) in curtailing proliferation among rivals such as Brazil and Argentina (Cirincione et al. 2005). In other cases, the role of international institutions was indirect (Dorussen and Ward 2008). For example, most observers credit the General Agreement on Tariffs and Trade (GATT) and the WTO with contributing to the explosive growth in trade during the post-Second World War period (Held et al. 1999: 175). In more general terms, Russett and Oneal (2001: 170–1) find that as two states increase their membership in international organizations, the probability of military conflict declines. Boehmer et al. (2004) argue that international organizations vary with respect to reducing conflict. They demonstrate that organizations with greater institutional structure, particularly those with a security mandate and low levels of member contentiousness, are more likely to reduce interstate conflict. Shifting to the duration of conflict, Shannon et al. (2010) find that international organizations decrease the length of international conflicts. Using network analysis, Dorussen and Ward (2008) demonstrate that membership in international organizations has both a direct and indirect pacifying effect. Although no liberal would claim that international institutions are sufficient for peace, these institutions appear to reduce a wide variety of conflicts.