Our results show that the presence of larger tripwire forces modestly increases support for intervention. Respondents were only about 3 percentage points likelier to support an intervention when they read that tens of thousands of U.S. troops were stationed inside the victim country compared to when there were no troops present. A smaller force of a few thousand also increased support by about 3 percentage points compared to the no-troops condition.

To put that into perspective, the presence of an alliance treaty between the United States and the victim also increased support by about 5 percentage points, as did U.N. Security Council approval. A large tripwire force, in other words, is slightly less important (or, given statistical uncertainty, about as important) as these diplomatic measures.

All of those factors, however, were far less meaningful than the likelihood of success and the expected levels of casualties the United States would sustain. The probability that an individual would support intervention was roughly 15 percentage points lower when respondents read that the operation would probably not succeed, compared to when they read that an intervention would almost certainly succeed. Even giving an intervention roughly even odds of succeeding decreased support by about 6 percentage points relative to almost certain success.