[Paper Review] Triangulating War: Network Structure and the Democratic Peace

Benjamin Campbell, Skyler Cranmer+1 more|arXiv (Cornell University)|Sep 11, 2018

Political Conflict and Governance31 references4 citations

TL;DR

This paper challenges the democratic peace theory by showing that the observed peace between democratic dyads is not due to democracy itself, but rather a network effect: democracies avoid conflict because they often coordinate against common enemies. Using inferential network analysis, the authors find that once common enemies are accounted for, the democratic peace vanishes and joint democracies are actually more conflict-prone than mixed-regime dyads.

ABSTRACT

Decades of research has found that democratic dyads rarely exhibit violent tendencies, making the democratic peace arguably the principal finding of Peace Science. However, the democratic peace rests upon a dyadic understanding of conflict. Conflict rarely reflects a purely dyadic phenomena---even if a conflict is not multi-party, multiple states may be engaged in distinct disputes with the same enemy. We postulate a network theory of conflict that treats the democratic peace as a function of the competing interests of mixed-regime dyads and the strategic inefficiencies of fighting with enemies' enemies. Specifically, we find that a state's decision to engage in conflict with a target state is conditioned by the other states in which the target state is in conflict. When accounting for this network effect, we are unable to find support for the democratic peace. This suggests that the major finding of three decades worth of conflict research is spurious.

Motivation & Objective

• To challenge the foundational assumption of the democratic peace theory that democracies inherently avoid war with one another.
• To address the theoretical and methodological flaw in dyadic conflict analysis that ignores network dependencies beyond the focal pair.
• To investigate whether the democratic peace is an artifact of unaccounted-for network structures, particularly shared enemies.
• To test whether joint democracies are less conflict-prone due to strategic coordination against common adversaries rather than democratic norms or institutions.
• To demonstrate that the democratic peace effect disappears when meso-level network dynamics—especially triadic closures involving common enemies—are properly modeled.

Proposed method

• Employing inferential network analysis to model state conflict as a relational, networked phenomenon rather than isolated dyadic events.
• Using statistical models that account for network dependencies, including triadic closure and common enemy structures, to test the robustness of the democratic peace.
• Introducing a network-based predictor for conflict likelihood based on shared enemies and regime alignment in triadic configurations.
• Applying regression models with network-corrected standard errors to control for non-independence in dyadic conflict data.
• Comparing results across multiple operationalizations of the outcome variable (e.g., militarized interstate disputes) and different measures of joint democracy.
• Conducting sensitivity analyses by removing high-influence dyads to test the stability of the democratic peace effect.

Experimental results

Research questions

• RQ1 Does the democratic peace effect persist when network dependencies—particularly shared enemies—are accounted for in conflict modeling?
• RQ2 To what extent do joint democracies avoid conflict due to coordination against common enemies rather than intrinsic democratic norms or institutions?
• RQ3 How do the structural properties of the conflict network, such as triadic closure and regime homophily, influence the likelihood of militarized interstate disputes?
• RQ4 Is the democratic peace a spurious finding resulting from the failure to model network interdependencies in dyadic conflict data?
• RQ5 Do mixed-regime dyads exhibit lower conflict propensity than joint democracies once network effects are controlled for?

Key findings

• After accounting for shared enemies and network structure, the statistical support for the democratic peace completely vanishes.
• Jointly democratic dyads are found to be more conflict-prone than mixed-regime dyads when network effects are controlled for.
• The most influential dyads for the democratic peace effect are not joint democracies, but mixed-regime dyads and jointly autocratic pairs, particularly those with enduring rivalries or common enemies.
• The U.S. and Canada, and the U.K. and France—two of the few consistently joint democratic dyads—maintain peace largely due to coordinated action against common enemies like Nazi Germany, Saddam Hussein, and Kim Il-Sung.
• The top ten dyads most responsible for the democratic peace effect include only two that are consistently democratic, while four are mixed dyads and four are jointly autocratic, indicating that shared enemies, not regime similarity, drive the observed peace.
• The mechanism underlying the democratic peace is not institutional or normative, but strategic: it is inefficient for two states fighting a common enemy to also fight each other, making conflict between them unlikely.