Generating Support for a Hypothetical War: Presidential Cues and Justifications

26 Pages Posted: 10 Jan 2019

Date Written: January 8, 2019


Support for wars that involve the U.S. typically divides by party identification among American voters, leading many scholars to believe that voters need only a partisan cue to go along with war. However, presidents do not endorse war in isolation; instead, they justify it with a variety of reasons ranging from minimal to extensive. Causally identifying these factors --- partisan cues and justifications --- can be challenging because (1) both occur simultaneously and (2) measured opinion on war usually occurs after the public has been exposed to both. This study leverages experimental evidence that randomizes presidential cues (an actual sitting president) and justifications about a hypothetical (and unnecessary) war in which elites have not staked out positions. In this hypothetical situation, President Barack Obama proposes war with China in 2014 with minimal justifications. Results show that a presidential endorsement alone does not generate support for a hypothetical war, but the inclusion of a justification, even one that is minimal, can increase support for war and improve presidential approval. Overall support still remains low for a hypothetical war and is concentrated among in-partisans. These results imply that a segment of the American public will go along with war, but the reach is limited to their own party. Further research should investigate hypothetical wars proposed and justified by sitting presidents from the Republican party (the "hawks") as well.

Keywords: Public Opinion, Survey Experiment, Elite Cues, Presidency, War, China

Suggested Citation

Gooch, Andrew, Generating Support for a Hypothetical War: Presidential Cues and Justifications (January 8, 2019). Available at SSRN: or

Andrew Gooch (Contact Author)

La Salle University ( email )

Philadelphia, PA 19141
United States

Register to save articles to
your library


Paper statistics

Abstract Views
PlumX Metrics