likelihood that an existing conflict will escalate to war.  Neither are these dynamics likely to lessen the potential for conflict between libertarian and non-libertarian states.  Third, this paper reviews some indirect empirical support for this argument.

          James Lee Ray addresses R. J. Rummel's hypothesis that “libertarian systems mutually preclude violence” early in his literature review of the democratic peace theory.  Like many other scholars, Rummel emphasizes the importance of democracies’ norms and institutions on their foreign policies.  From a cultural or normative perspective, he claims that “between libertarian states there is a fundamental sympathy of their peoples toward each other’s system, a compatibility of basic values.”[3]  This is a common view of the role that democratic norms plays, but it differs slightly from Bruce Russett’s more specific thesis.[4]  Russett contends that, because “democratic political culture encourages peaceful means of internal conflict resolution,” democratic leaders are willing to apply these norms “across national boundaries toward other democratic states,” and they expect their foreign counterparts to do the same.[5]  Yet, in both explanations, democratic dyads are peaceful because their members arrive at “positive perceptions of other democracies.”[6]  And it is this attitudinal outcome that Christopher Layne finds lacking in his case studies of “modern historical instances in which democratic great powers almost came to blows”[7] (this paper incorporates Layne’s observations below).

          Addressing democratic institutions, Rummel places a “Kantian stress on the role of pacifistic public opinion and interest groups.”[8]  Here he embraces another familiar causal mechanism: Democratic leaders are wary of entering wars, primarily because they can be voted out of office by the same citizens who “pay the price for war in blood and treasure.”  Other scholars highlight additional structural constraints on democratic leaders, including “political competition” and “the pluralism of the foreign policy decisionmaking process.”[9]  But in both variants of the institutional argument, “democratically elected leaders are unable to act quickly,” and this sluggishness has a dual impact.  On the one hand, “this cautious foreign policy behavior reduces the likelihood that a conflict will escalate to war.”  On the other hand, these institutions signal to other states that they should “anticipate a difficult and lengthy process before [a] democracy is likely to use significant military force against them.”  As a result, these other states will expect – and capitalize on – a greater “opportunity to reach a negotiated settlement.”[10]

          Yet while normative hypotheses explain why a dyad must consist of two democracies before it can achieve a democratic peace, structural arguments tend to posit that one state’s democratic institutions will lower the likelihood of war with a second state, to an extent that does not depend on whether the second state is democratic.  Accordingly, institutional arguments “cannot account for a democratic public’s willingness to fight wars against nondemocracies” more frequently than against democracies.[11]  Layne asserts simply that “[i]nstitutional constraints do not explain the democratic peace,” and he does not bother to test their influence in his case studies.[12]

          As these structural hypotheses limit their theoretical traction over the dynamics of democratic dyads, they also take us farther away from the approach outlined in this paper’s introduction: They treat democracy not as a source of regimes’ (shared) interests, but rather as a device for reconciling states’ (conflicting) interests.  Whereas cultural arguments speculate that democratic dyads produce a joint normative interest in peace, the aforementioned institutional theses claim – at most – that individual democratic leaders have an interest in avoiding war.

          Kenneth Schultz’s structural approach in Democracy and Coercive Diplomacy goes even farther.  He notes that “[t]he availability of information, and the nature of strategic interaction under conditions of uncertainty, intervene in the causal chain between interests and outcomes”; his book therefore concentrates “on information rather than on preferences.”[13]  Specifically, he argues that because a democracy experiences “[o]pen political competition,” it cannot help but reveal some information about its preferences or capabilities “to both domestic and foreign audiences.”  This reduction in asymmetric information helps it and its adversaries to “identify a settlement that both sides prefer to war.”[14]  In fact, Schultz implies that institutional arguments can be collapsed into informational theories: Dictators may run greater “political risks” in going to war than do elected officials, but “the political risks generated by their threats are [less] obvious to outsiders,” and the scarcity of information makes war more likely.[15]  Nevertheless, actors’ interests retain substantial causal weight – unless states have “conflicting preferences over the allocation of disputed goods,”[16] information is of little value to them.

          Institutional approaches to the democratic peace theory, then, seem to have surrendered their explanatory power – both to cultural and even neorealist arguments that build upon states’ preexisting preferences,[17] and to Schultz-like and neoliberal arguments that stress the importance of shared information for achieving cooperative outcomes.  Therefore it may come as a surprise that this paper advances an institutional explanation of a libertarian peace.  As foreshadowed in the introduction, it contends that libertarian institutions preclude conflicts of interest within libertarian dyads.

          Rummel, too, frames libertarianism in institutional terms.  “Freedom inhibits violence,” he writes; “libertarian states have natural inhibitions on involvement in violence,” and “economic institutional arrangements” assume a unique importance in his thinking.[18]  However, James Lee Ray suggests that Rummel’s causal mechanisms focus largely on the variables discussed in this paper’s previous section, including “the responsiveness of elected leaders to domestic interest groups or public opinion, which ordinarily will oppose violence, tax increases and conscription.”[19]  Therefore this section will develop a separate model of a libertarian peace, first examining the nature of a libertarian regime – including its “inhibitions on…violence” – and then deducing this regime’s inherent constraints on foreign policymaking.

          In ideal-typical form, a libertarian government has the sole purpose of protecting private property rights.  This fundamentally negative aim prevents the libertarian regime from adopting any positive objectives: It cannot interfere with its citizens’ properties; it cannot interfere with its citizens’ exchange of their properties.  It cannot spend tax monies on anything other than the physical defense of its citizens’ properties; if it did, it effectively would be confiscating its citizens’ monetary properties.  In Weberian terms, a libertarian regime cannot use its monopoly on the legitimate use of violence except to protect private property rights.  In even simpler terms, it may not initiate the use of violence – it may defend its citizens’ properties only.

          This principle parallels Rummel’s notion of libertarian “inhibitions on…violence.”  If we were to treat it as a norm – similar to the democratic norm of peacefully resolving disputes – then we could devise a cultural explanation for a libertarian peace.  But Rummel implies, and the preceding paragraph makes clear, that the non-initiation of violence is an institutional function of a libertarian regime.

          Therefore a dyad of libertarian states cannot experience a conflict of interests that could lead to war.  In an anarchical system where “each unit…is responsible for ensuring its own survival,” both states may seek to increase their military capabilities through internal balancing or external alliances.  But because anarchy also sets each state “free to define its own interests and to employ means of its own choice in pursuing them,”[20] each libertarian regime will advance only the “interests” of its citizens’ property rights.  Put differently, it will forswear any “means” that entail the initiation of violence.  And without the initiation of violence – without any contradictory interests – a libertarian peace will ensue.

          Of course, a libertarian regime will have no “inhibitions on…violence” against a non-libertarian state – even a democracy – if the latter violates its citizens’ property rights.  Like the norm-based models discussed in the preceding section, the libertarian peace can obtain only within a libertarian dyad.