On Saturday, Syrian President Bashar Al-Assad reportedly launched another chemical weapons attack, this time against civilians in the town of Douma. President Trump tweeted that those responsible would have a “big price to pay” and called the incident “another humanitarian disaster for no reason whatsoever.”

That’s much like the White House’s reaction to Assad’s use of chemical weapons in April 2017. Then, Trump condemned the attack as a “reprehensible” act that “cannot be ignored by the civilized world” and retaliated with a targeted military strike on the launch site.

But Sunday’s statement is more complicated. Trump had just pushed the Pentagon to withdraw U.S. troops fighting the Islamic State in Syria. He argued, in keeping with his “America First” approach, that the United States had spent too much time and money in Syria and now had other priorities.

Which of these approaches will Americans prefer? Scholars are divided on this question. If Daniel Drezner’s research on what he calls the “realist tradition in American public opinion” is right, Americans will applaud Trump’s push to put American interests and people first – but not his more recent humanitarian emphasis.

By contrast, our research joins other recent evidence suggesting that Trump’s humanitarian rhetoric about the attacks can persuade the public to support military action – even when partisanship and polarization are high.

How we did our research

To see whether and why Americans support humanitarian interventions, we conducted three survey experiments between 2014 and 2015. The first and third experiments were conducted through Amazon’s Mechanical Turk with convenience samples of 330 and 408 U.S. adults, respectively. The second experiment was fielded though YouGov and included a nationally representative sample of 1,500 U.S. adults.

We wanted to see whether Americans were more likely to support military action for humanitarian reasons or military action to defend a foreign country against invaders. In the humanitarian scenarios, we told respondents that a foreign country’s military was being used to “massacre civilians, including innocent women and children” – and that the country being attacked was “suffering a humanitarian crisis.” In the other set of scenarios, we told them that the foreign country’s military was attacking a country “unable to defend itself” to “gain power and resources.”

The humanitarian scenarios were framed much the same way as Trump’s statement on Syria. The other scenario was more like the U.S. intervention in the 1991 Persian Gulf War under President George H.W. Bush, when Iraq invaded Kuwait – and a U.S-led military coalition drove it out.

We then asked all respondents whether they would support military action, and followed up with questions about why.

Americans feel a moral obligation to intervene for humanitarian reasons

We found that the public is more likely to support the use of force for humanitarian purposes than for defending another country — by margins of up to 27 percent.

But why? The answer: moral obligation. When we asked followup questions, we found that respondents didn’t believe humanitarian interventions would involve lower costs or casualties or that inaction would present the United States with adverse strategic consequences.

Rather, in the first experiment, 20 percent more respondents thought the United States had a moral obligation to act in the humanitarian scenario than thought that for the defensive scenario.

In the second experiment, in which we told respondents that the United States would cover most of the costs of the mission, moral obligation continued to be the main reason people supported intervention for humanitarian purposes.

In the third experiment, we asked respondents why they felt the United States was obligated to intervene, and allowed them to pick all that apply from a list of possible reasons. We learned that individuals were especially concerned about people being harmed — this explained over 20 percent of the increased support for humanitarian interventions.

Individuals were less concerned about whether leaders could get away with human rights abuses, if allowing such behavior would create disorder in international society, whether inaction would betray U.S. values, or if the foreign country was violating expectations for how civilized countries should behave. Each of these reasons explained less than 5 percent of the increased support for humanitarian action.

That sense of moral commitment holds true for those in both parties

That sense of moral obligation for humanitarian interventions crosses the aisle – but there are partisan differences when it comes to support for the security intervention. Among our Republican respondents, about three-fourths felt a moral obligation to intervene for humanitarian reasons – and about the same felt an obligation to intervene for security reasons. Among Democrats, about three-quarters said they felt a moral obligation to support humanitarian interventions, but this sense of obligation dropped by between 7 and 33 percent, depending on the survey, for the use of force to defend another nation. Emphasizing humanitarian violations is thus a good way to rally bipartisan support for military action.

And in a series of 10 survey experiments conducted from May 2016 to May 2017, we found that support for humanitarian interventions stayed steady throughout the 2016 election campaign and afterward.

In other words, if Trump decides to further involve the U.S. military in Syria, emphasizing humanitarian needs can build a sense of moral obligation and support from both Democrats and Republicans. Focusing narrowly on U.S. national interests won’t do the trick.

Sarah Kreps (@sekreps) is an associate professor of government at Cornell University and an adjunct scholar at West Point’s Modern War Institute. She is the author, most recently, of Taxing Wars: The American Way of War Finance and the Decline of Democracy (Oxford University Press, 2018).

Sarah Maxey (smaxey265) is a postdoctoral fellow at the University of Pennsylvania’s Perry World House.