Strategic Rivalries, Protracted Conflict and Crisis Escalation

Point:  Action-reaction in crisis is radically different in rivalries than in regular conflict.  Rivalry not only makes escalation and crises more likely, but also significantly interacts with more traditional predictors of conflict such as capability ratios, the number of actors in a crisis, and the issues under contention, thus enforcing the notion that the study of rivalries is vital to the field.
They argue that 1% of interstate pairings are responsible for 60% of the 20th century crises.  This is because crisis in rivalries are much more violent and dangerous than in regular dyads. Authors present four mechanisms by which rivalry can interact with other variables to increase the probability of escalation in a crisis 1) Rivalry can trump or overwhelm constraints on conflict escalation such as objective signals of friendship or peace, norms, major power intervention, etc, 2) Rivalries lower the threshold for escalation, and raise the likelihood of military interaction so that nonviolent triggers,  non-gravity threats, and non-military issues will spark a conflict, 3) Increased uncertainty in rivalry dyads results in increased insecurity and possible overreaction especially as the system moves toward multipolarity, and 4) The ‘death watch’ hypothesis, where rival states look for opportunities to attack when other state is experience internal instability.
Authors find that rivalries increase the probability of war.  Of their four mechanisms, hey found that constraint had 2 of the 3 functions, Democracy has a stronger pacifying effect inside rather than outside rivalries, major power involvement is not a constraint both as a participant and as an intervening force (disproves their hypothesis), and crises move toward war in rivalries despite asymmetric distribution of capabilities.  On their second mechanism, they found that 2 of 3 the three issue categories increased the probability of conflict. For their third, uncertainty, they found mixed support for rivalry crises moving toward conflict when high salience issues are in play in a bipolar system.  In multipolar systems, rivalries don’t change the systemic workings, but there continues to be more conflict in rivalries than in non-rivalries.  And finally, they found no support for their concept that the ‘death watch’ hypothesis was a factor.
In conclusion, Colaresi and Thompson find strong support for claim that rivalries are more violent than non-rivalries.  Crises in rivalry dyads are more likely to involve war, mid-level threats, and militarized techniques leading to arms build up and an expectation of future conflict, in addition rivalry escalation involves higher levels of uncertainty and reduced constraints to make conflict more likely.