If the U.S. economy tanks, should we expect Donald Trump to engage in a diversionary war? Since the age of Machiavelli, analysts have expected world leaders to launch international conflicts to deflect popular attention away from problems at home. By stirring up feelings of patriotism, leaders might escape the political costs of scandal, unpopularity — or a poorly performing economy.

One often-cited example of diversionary war in modern times is Argentina’s 1982 invasion of the Falklands, which several (though not all) political scientists attribute to the junta’s desire to divert the people’s attention from a disastrous economy.

In a 2014 article, Jonathan Keller and I argued that whether U.S. presidents engage in diversionary conflicts depends in part on their psychological traits — how they frame the world, process information and develop plans of action. Certain traits predispose leaders to more belligerent behavior.

Do words translate into foreign policy action?

One way to identify these traits is content analyses of leaders’ rhetoric. The more leaders use certain types of verbal constructs, the more likely they are to possess traits that lead them to use military force.

For one, conceptually simplistic leaders view the world in “black and white” terms; they develop unsophisticated solutions to problems and are largely insensitive to risks. Similarly, distrustful leaders tend to exaggerate threats and rely on aggression to deal with threats. Distrustful leaders typically favor military action and are confident in their ability to wield it effectively.

Thus, when faced with politically damaging problems that are hard to solve — such as a faltering economy — leaders who are both distrustful and simplistic are less likely to put together complex, direct responses. Instead, they develop simplistic but risky “solutions” that divert popular attention from the problem, utilizing the tools with which they are most comfortable and confident (military force).

Based on our analysis of the rhetoric of previous U.S. presidents, we found that presidents whose language appeared more simplistic and distrustful, such as Harry Truman, Dwight Eisenhower and George W. Bush, were more likely to use force abroad in times of rising inflation and unemployment. By contrast, John F. Kennedy and Bill Clinton, whose rhetoric pegged them as more complex and trusting, were less likely to do so.

What about Donald Trump?

Since Donald Trump’s election, many commentators have expressed concern about how he will react to new challenges and whether he might make quick recourse to military action. For example, the Guardian’s George Monbiot has argued that political realities will stymie Trump’s agenda, especially his promises regarding the economy. Then, rather than risk disappointing his base, Trump might try to rally public opinion to his side via military action.

I sampled Trump’s campaign rhetoric, analyzing 71,446 words across 24 events from January 2015 to December 2016. Using a program for measuring leadership traits in rhetoric, I estimated what Trump’s words may tell us about his level of  distrust and conceptual complexity. The graph below shows Trump’s level of distrust compared to previous presidents.

These results are startling. Nearly 35 percent of Trump’s references to outside groups paint them as harmful to himself, his allies and friends, and causes that are important to him — a percentage almost twice the previous high. The data suggest that Americans have elected a leader who, if his campaign rhetoric is any indication, will be historically unparalleled among modern presidents in his active suspicion of those unlike himself and his inner circle, and those who disagree with his goals.

As a candidate, Trump also scored second-lowest among presidents in conceptual complexity. Compared to earlier presidents, he used more words and phrases that indicate less willingness to see multiple dimensions or ambiguities in the decision-making environment. These include words and phrases like “absolutely,” “greatest” and “without a doubt.”

A possible implication for military action

I took these data on Trump and plugged them into the statistical model that we developed to predict major uses of force by the United States from 1953 to 2000. For a president of average distrust and conceptual complexity, an economic downturn only weakly predicts an increase in the use of force.

But the model would predict that a president with Trump’s numbers would respond to even a minor economic downturn with an increase in the use of force. For example, were the misery index (aggregate inflation and unemployment) equal to 12 — about where it stood in October 2011 — the model predicts a president with Trump’s psychological traits would initiate more than one major conflict per quarter.

Of course, predictions from such a model come with a lot of uncertainty. By necessity, any measures of a president’s traits are imperfect. And we do not know whether there will be an economic downturn. Moreover, campaigning is not governing, and the responsibilities of the Oval Office might moderate Donald Trump. The psychologist Philip Tetlock has found that presidents often become more conceptually complex once they enter office.

Nevertheless, this analysis suggests some cause for concern about the international ramifications of an economic downturn with a President Trump in the White House.

Dennis M. Foster is professor of international studies and political science at the Virginia Military Institute.