Individual Submission Summary
Share...

Direct link:

Do Alliances Deter? An Experimental Design in the U.S.

Fri, October 1, 8:30 to 9:00am PDT (8:30 to 9:00am PDT), TBA

Abstract

Do alliances deter, provoke, or simply have no effect on conflict between states? This question is the subject of a recent debate in the Journal of Politics (Kenwick et al. 2015; Kenwick and Vasquez 2016; Leeds and Johnson 2017; Morrow 2017). While alliances are increasingly the focus of experimental research (Tomz and Weeks 2019; Fjelstul et al. 2015), we conducted the first experiment to address the debate over the effectiveness of alliances to deter.
We conducted a survey experiment via the Cooperative Congressional Election Study (CCES) in 2020 to test key propositions from competing theories of alliances and address a prominent debate on the relationship between alliances and deterrence success. Our experiment builds on recent experimental work in international relations (Fjelstul et al 2015; Tomz and Weeks 2019). Our motivating vignette, how we describe alliances to the respondents, and the phrasing of the dependent variable imitate prior research. The primary mechanism motivating institutionalist theory is costly signals. By signing formal written alliance agreements, allies signal their preferences to adversaries. The alliance, at least for leaders in democracies, is a costly signal. Recent experimental work finds support for this mechanism in the context of alliances (Fjelstul et al 2015; Tomz and Weeks 2019). Our first treatment, then, is the presence of an alliance, whose effect we will compare to a condition of no alliance. In addition to tying hands signals, institutionalists contend that sunk cost actions, such as troop mobilization or joint military exercises, can signal credibility(Fearon 1997; Fuhrmann and Sechser 2014). Research in this area motivates our second treatment, the presence or absence of sunk cost signals. Our third treatment more directly tests the causal mechanism in the Steps-to-War argument. Kenwick et al (2015) argue that recent alliances provoke conflict. Morrow (2017) concurs and adds that older alliances deter. The third treatment manipulates the age of the alliance. In our experiment, deterrence success means more respondents should oppose the use of force to stop the development of nuclear weapons in the target country.
This experiment sheds new light on the effect of alliances on international security and the mechanism of alliances (de)escalating interstate conflicts.

Authors