Political polarization in modern American politics likely can't be reduced by foreign threats and crisis events, according to new research, challenging notions held by political scientists that threats from abroad generate greater domestic unity across partisan divides.

Using three separate studies, Rachel Myrick, assistant research professor at Duke University and a Ph.D. candidate at Stanford University, found that it is unlikely that partisan polarization over U.S. foreign policy or effective polarization among the American public has been substantially shaped by America's threat environment. Building on prior literature, Myrick had hypothesized that foreign threats might create partisan unity through two mechanisms: the information mechanism, which involves threats that credibly signal the gravity of a foreign policy situation and create a shared objective among policymakers; and the identity mechanism, which involves foreign threats that encourage social cohesion by emphasizing a sense of national unity over partisan identity.

Myrick, who published her findings April 20 in International Organization, said she was interested in this topic because the idea that external threats reduce domestic polarization is pervasive in the national security community.

"It is often discussed in the context of a U.S.-China rivalry, with the expectation that a rising China will be good for domestic politics in the United States," she said. "I see this argument asserted a lot but wondered: what about the kind of evidence we actually have from the historical record? So this paper is an attempt to look systematically at whether crises and rivalries historically reduced domestic polarization in the American context."

The two mechanisms were tested in three separate analyses. The first implemented a computational text analysis of the Congressional Record to determine the extent to which Republican and Democratic legislators' rhetoric might differ when faced with various crises triggered by foreign adversaries. The second study correlated security crisis events with biweekly Gallup polls from 1953 to 2012 to measure changes in the partisan gap in presidential approval ratings and determine approximate levels of polarization. Finally, researchers fielded a national survey to a representative sample of 2,500 American adults that fabricated a threat posed by China to the U.S. and analyzed the scenario's impact on expressed "affective polarization," or the tendency of people to favor their own party and be unfavorable toward the opposing party.

The supervised machine-learning process found no association between heightened threats and polarization of legislators' rhetoric, which appeared to instead reflect the preexisting partisanship of the political environments. Using the second study, Myrick then analyzed how security crises might influence affective polarization among the American public and found that crises were associated with a smaller partisan difference in presidential approval ratings — although such effects were typically small and short-lived.

The survey utilized by Myrick failed to support the identity-mechanism hypothesis and offered mixed support for the information mechanism. When the report of a Chinese threat was attributed to President Trump, for example, attitudes among survey takers grew more polarized. If nonpartisan attribution was used, though, the external threat hypothesis was more likely to hold up.

Looking at the studies collectively, Myrick found that partisan polarization is likely to be influenced by prior demographic and institutional changes, which in turn are likely to affect foreign affairs. Thus, claims that crises or emergent threats from rival powers will automatically unify various bickering factions within the U.S. should be met with a healthy dose of skepticism. 

In the short term, for example, some have argued that the "rally 'round the flag effect" can help to generate domestic unity in the face of foreign threats, but Myrick's research suggests that threat environments are unlikely to calm partisan polarization over U.S. foreign policy or affective polarization among the American public.

Myrick said this might be because politicians find it more difficult to credibly communicate threat information without partisan language when an environment is highly polarized. 

"In polarized times, policymakers and the public are likely to see the president as an inherently political actor. And rather than simply defer to the president on matters of national security, the political opposition has incentives to obstruct or criticize the executive," Myrick said. "Importantly, this is not to suggest that external threats will never reduce polarization. Instead, we should expect large, persistent reductions in polarization resulting from foreign threat to be exceedingly rare."

Myrick believes that a natural extension of this project would be to consider which threats are more likely or less likely to be polarizing and in what particular contexts. 

"This research only considers security threats in the American context," she said. "But serious threats to countries that are smaller or are bordered by adversarial powers might be more likely to generate partisan unity. It would also be interesting in further research to distinguish between different types of threats, including economic threats and public health threats."

The study, "Do external threats unite or divide? Security crises, rivalries, and polarization in American foreign policy," published April 20 in International Organization, was authored by Rachel Myrick, Duke University and Stanford University.