We presented 1,200 participants on Amazon’s Mechanical Turk platform with five pairs of scenarios in which an aggressor country attacked a target country. We randomly varied the values of 15 characteristics in the scenario. This let us simultaneously estimate the effect of a large handful of variables — including the presence and size of a tripwire force — on support for intervention.

Our experiment dealt with scenarios in which tripwire forces may have been present and could easily be drawn into a conflict, but had not been directly attacked. Understanding the effect of a direct attack on U.S. forces stationed abroad is an important next step, but it requires first exploring whether the presence of troops works as a tripwire.