The decision to order the death of Iranian general Qassem Soleimani on January 3 sparked intense controversy from Democratic officials. Subsequent reporting by the New York Times has highlighted President Trump’s unique domestic context (the recent impeachment, the upcoming election) in his decision to order the strike. Even Trump himself has suggested that a President seeking reelection would benefit from starting a war with Iran. But is that the case?
It is widely assumed that wars increase presidential approval. However, political science research does not support this. Rather, the high visibility of a presidential election and the poor prospects for quick victory mean that a war in Iran would be far more likely to hurt the President’s approval rating than to help it.
The Guardian recently published an article citing Gallup Poll data from the Vietnam War, Operation Desert Storm, and the 2003 Iraq War. On its face, the article presents a pattern of behavior that links war to large spikes in the president’s approval rating. However, the correlation it presents is problematic. In many of these cases, brief increases in presidential approval are followed by steep declines.
Moreover, this analysis only shows the uses of force that have resulted in large increases in presidential approval, which tells us little about the expected effect of a single use of force. (This is referred to as ‘selecting on the dependent variable’, and it is a cardinal sin of social science.) Researchers that include all uses of force, rather than selecting those with the largest effect, find that after a use of force “the mean change in the president’s approval rating is 0%, even among the members of his party.”
This may explain why there is so little evidence of presidents using force as a diversionary tactic. Although this is a popular hypothesis, historical researchers have found “little evidence of any kind of link between domestic political conditions in the United States and uses of force or international crises.”
Even if war with Iran generated public support early on, there is significant evidence to suggest this would decline precipitously as casualties and military budgets mounted.
If we can’t expect that a war with Iran would necessarily lead to an increase in public support for the president, what can modern political science tell us about how public opinion might respond to a broader military conflict? Like most things in academia, this is far from a settled question. However, we can draw a few general lessons from the political science literature:
1. Public opinion is shaped by the opinions of political elites.
Political elites have substantial influence on how the public will view any conflict. In his book, “In Time of War,” Adam Berinsky suggests that partisanship, rather than the specific details of any particular action, has been the major driver of public opinion in conflicts from Pearl Harbor to Vietnam and beyond. Other authors have found similar results.
The strong influence of partisanship on public opinion and the already strident Democratic criticism of the prospects for war with Iran suggest little potential for a swell of national support. Even in the event of a substantial retaliation from Iran, Berinsky uses the case of Pearl Harbor to suggest that many who do not want war did not change their opinion after the attack. The strong public antipathy to the war in Iraq in 2003 provides a more contemporary example. A decision to go to war with Iran is unlikely to sway any voters outside of the Republican party. However, even voters within the Republican party may not support the war as new information emerges.
2. Public opinion responds to new information.
In addition to relying on party leaders, the American public responds to new information. This means that both Republicans and Democrats’ opinion towards any conflict will depend heavily on the progress of that conflict. Multiple historical studies show that as a conflict progresses, the public responds to information about its progress and costs. Comparatively, the size of this effect can be even larger than that of partisanship.
The likelihood that voters actually get the information that could change their mind depends on the media they consume. While it is a widely held belief that we live in increasingly siloed information spaces, this is not fully supported by research. The literature on partisanship, rationality, and perception is too vast for this brief article, and I won’t attempt to summarize it fully. However, there is evidence that people online are no more siloed than before, and will change their views in response to information from outside their party. Additionally, because this is an election year, the progress of any conflict is sure to be a major national talking point. Even if war with Iran generated public support early on, there is significant evidence to suggest this would decline precipitously as casualties and military budgets mounted.
The Trump Administration would be wrong to assume that a war with Iran would rally national support for the administration and increase the President’s chances of reelection. On the contrary, an initial spike in public approval is not only uncertain but likely to be short-lived (on the order of weeks). The role of the party in shaping public opinion suggests that even moderate Democratic voters will not likely change their opinion on President Trump because of a conflict. As any conflict progresses, intense public scrutiny on its progress is likely to highlight costs that will decrease support among voters across the political spectrum. Far from being a sure shot to reelection, a war with Iran could lose Trump his job.
Leah Matchett is a second-year PhD student at Stanford University, where she is also a member of the first cohort of Knight-Hennessy Scholars.
