Earlier this summer, after Iran downed an unmanned U.S. military aircraft, President Trump said he had approved — then put a stop to — a U.S. missile strike on Iran. Polling showed most Americans supported aborting the mission. One Harvard CAPS/Harris Poll survey found that 57 percent of voters opposed “military confrontation with Iran” unless Iran had attacked the United States. Only five percent of respondents wanted the United States to “declare war on Iran.”

This unpopularity of intervening in Iran fits with larger public opinion trends. In a forthcoming article in the Journal of Conflict Resolution, I analyzed the results of more than 1,000 nationally representative survey questions about the use of force overseas. Collected from 1981 to 2016, these data reveal steady patterns in public attitudes that continue today, showing that Americans would hesitate to pick fights with governments not known for harming the United States or its allies. Here’s what I found.

Overall approval for intervention has not risen significantly since 9/11.

The Americans polled about military intervention tend to weigh the risks of action against the perceived threat, according to the aggregate data. If the proposed intervention appears to increase risk — such as by deploying servicemembers on a dangerous, nonessential mission — respondents tend to balk.

By contrast, if the intervention looks set to reduce a danger posed to Americans — or does not appear innately risky for the servicemembers deployed — then a larger share of respondents will support it. For example, mentioning other countries’ backing through a United Nations mandate increases support by 10 points. Items that omit any talk of potential American casualties are roughly 15 percent more popular than those that do mention U.S. losses.

These patterns hold across the 26 years of data, including the “post-9/11” era, when many Americans were concerned about national security. From 2002 to 2015, support for military intervention in Iran averaged around 50 percent. The data also reveals that Americans had been unwilling to fight with North Korea in 2003-2014, when only 37 percent supported military intervention; with the Syrian government in 2012-2014, when only 33 percent supported use of force; and when pro-Russian forces invaded eastern Ukraine in 2014, when only 18 percent supported U.S. involvement.

Of course, public opinion does not predict presidential policy. Though only 30 percent of Americans endorsed intervening in Libya in 2011, President Barack Obama deferred to the United States’ European partners and launched the operation anyway. Public resistance to intervention would also likely not stop Trump from ordering an attack, but it suggests he would face political backlash.

Iran’s cooperation with the Taliban could affect talks on U.S. withdrawal from Afghanistan

Why were Americans so unusually willing to fight in Afghanistan and Iraq?

Notably, Americans were much more willing to fight in Afghanistan and Iraq. In prewar surveys that did not mention possible casualties, nearly three out of four respondents (74 percent) approved attacking the Taliban in Afghanistan, and 62 percent supported military action against Iraq. Across all the surveys from 1981 to 2016, Americans were 28 percentage points more likely to support intervention against Saddam Hussein and Iraq than any other target, and 37 percentage points more likely to support attacking al-Qaeda or Osama bin Laden in Afghanistan. Of course, that’s because Americans were outraged by al-Qaeda’s 9/11 terrorist attacks on New York City and Washington, D.C.

This sort of response can be amplified by presidential messaging, as when the George W. Bush administration worked for months to persuade voters, Congress and the U.N. Security Council to support a U.S.-led invasion of Iraq. The second Bush administration was able to use widespread hostility toward Hussein, left over from President George H.W. Bush’s 1990-1991 Gulf War, well before the second Bush’s public relations campaign got underway.

At times, American antagonism toward certain opponents is in conflict with a commander in chief who favors restraint. After Islamic State attackers killed 130 people in Paris in November 2015, 68 percent of Americans surveyed thought Obama — whose administration had been striking ISIS targets in Iraq and Syria for over a year — needed to be more aggressive. Fifty-three percent favored “sending [U.S.] ground troops into combat operations against ISIS,” an idea that Obama had called reckless and ineffective. In general, over all these surveys, Americans were nine percent more likely to support using military force against the Islamic State than against other possible targets.

The public backs intervention against vivid threats.

A cautious public might be willing to support armed confrontation with Iran in either of two possible situations. First, the Trump administration could rely on the second Bush administration’s strategy to build support for invading Iraq and try to scare Americans into an “Iran War.” So far we’ve seen no evidence of such a coordinated White House push. Nor would that be an easy sell: Americans don’t have the kind of acute hostility toward Iran that they had toward Saddam Hussein; the inconclusive wars in Afghanistan and Iraq left them cautious about regime change; and there’s been no recent attack on Americans that might prompt a desire for attack.

Second, Iran might escalate its attacks. So far it has downed U.S. drones and seized foreign tankers. American public opinion might change if Iran began endangering U.S. aircrews and sailors, abandoning its current tit-for-tat strategy.

So far, the skirmishing with Iran hasn’t come close to matching the dramatic concern Americans felt after 9/11, making them willing to go after al-Qaeda, Iraq and the Islamic State. Most Americans feel cautious about armed intervention. Those attitudes may help keep the United States negotiating rather than attacking.

Don’t miss anything! Sign up to get TMC’s smart analysis in your inbox, three days a week.

Jason Brownlee (@jasonbrownlee) is a professor of government at the University of Texas at Austin, where he researches and teaches about authoritarianism, war and diplomacy.