President Trump is now suggesting that the country be “opened up” by Easter, despite warnings that the coronavirus pandemic has yet to peak in the United States. If that remains his suggestion, would U.S. citizens in a deeply divided society go along with his directive?

To answer that, let’s look at what social science research finds about when and how leaders can successfully persuade large groups to cooperate with a particular course of action. That won’t be simple in a highly polarized society in which some citizens passionately support and trust Trump while others find him deeply untrustworthy.

When it’s critical to get citizens to agree on a course of action

As the epidemic unfolds, interactions among citizens are an example of what game theorists call a “global game,” a situation in which people could benefit from coordinating their choices — but are either uncertain about or disagree on what choice is best.

For instance, if everyone in the office is working remotely, the downside to doing so oneself (instead of showing up) is low. But if everyone is going into the office, staying home may get you fired. Even still, you may want to do so — and wish others would as well — if, for example, you are concerned for the health of your elderly relatives.

Let’s think about this at a national level. If the president instructed almost all citizens to stay home, it could significantly “flatten the curve” of disease transmission but devastate the economy. If the president continues supporting the idea of everyone returning to work mid-April, it could prevent still greater economic collapse but would risk overwhelming hospitals and boosting mortality rates. Political leaders have to consider the trade-offs between these different sources of human misery, all in a time of tremendous uncertainty. Perhaps more important, after choosing a direction, leaders will want to communicate that to a highly divided nation in a way that will encourage them to cooperate.

How can a democracy get its citizens to follow the leader?

So when can public signals like a leader’s statements get citizens to coordinate in a mutually beneficial course of action? Here’s one critical factor: how trusted the leader is. If all citizens regard the leader as trustworthy, they will willingly follow the leader’s directives.

But that might not work, according to global games models, if a significant group of citizens believes the leader to be self-serving, uninformed or deceitful. Those who distrust the leader will be tempted to rely on other sources of guidance. This makes it harder for everyone to coordinate in a shared course of action, no matter how much it would benefit them all.

And we can expect that at least some group of citizens will distrust or dislike the leader in any well-functioning democracy. Those who support the opposition will be likely to second-guess and disregard the leader’s directions — and ignore them, going their own way. Under ordinary circumstances, that’s just what a free society looks like.

But we’re living in extraordinary circumstances. In the coronavirus pandemic, everyone’s physical and economic well-being depends on how our fellow citizens behave, just as they depend on us. Having a president who can effectively use the “bully pulpit” is critical in coordinating the nation’s behavior.

Many citizens believe anything Trump says and would dismiss experts who advise otherwise — alongside many who believe Trump is uniquely untrustworthy and will listen only to experts.

Here’s how the president could reach both his supporters and opponents

Trump has a particular communication style. In recent days he appears to want to project optimism and to denounce reporters who convey bad news — perhaps believing they undermine his authority, But in a crisis that requires broad cooperation among all citizens, publicly lambasting journalists undermines the president’s credibility as an information source, at least in his opponents’ eyes — and makes it less likely his approach will be followed.

Second, politically neutral experts’ credibility can be a precious resource for leaders trying to persuade citizens to follow a course of actions. Some observers suggest that Trump should stop speaking at the White House’s daily pandemic briefings, instead letting experts such as Anthony S. Fauci hold the floor. But that could backfire. While Trump opponents might be more willing to listen to and follow Fauci’s suggestions, Trump supporters are likely to ignore him. Similarly, removing Fauci from the briefings and replacing him with experts willing to praise the president could encourage Trump supporters to listen — but risks alienating Trump opponents entirely, decreasing the likelihood that they’ll follow the recommended course of action.

Third, Trump’s reelection argument was that he’d built a booming economy. When he urges Americans to “stay home,” it’s like anti-communist hawk Richard Nixon opening relations with China; even his detractors will take it seriously.

That may not be true if he announces that it’s time to go back to work. His supporters may comply, but his opponents may ignore the announcement, even if it’s the right policy.

So what could reach both groups of Americans? Matching, politically independent signals from both politically neutral experts and the president.

Here’s what that could look like. If the president speaks first, followed by experts who agree with him, his opponents might doubt that the experts are neutral. After all, the president can remove experts, but not the other way around. But if the president spoke second and endorsed what the experts said, without adding, subtracting or editorializing, both his opponents and his supporters would be likely to follow the leader.

Such an approach might be challenging for a president who rarely sticks to a script. But it may be the most effective way to get the country to agree to listen.

Sanford C. Gordon (@sanford_gordon) is professor of politics at New York University and conducts research on legitimate authority, executive branch politicization and electoral accountability.

Dimitri Landa is professor of politics at New York University and conducts research on political accountability, communication, and leadership.