Jonathan Ernst / Reuters

It’s easy to survey the last 10 days in America and be flabbergasted at President Trump’s tendency to inflict politically bruising crises on himself. I know this because I had just that reaction on Wednesday, trying to make sense of his abrupt signing of an executive order intended to end separations of unauthorized immigrant families at the border. The current moment is more acute than others, but it follows the template set by the aftermath of a white-supremacist march in Charlottesville, Virginia; brinksmanship with North Korea; the collapse of Obamacare repeal; the firing of James Comey; and any number of smaller, half-forgotten crises. Why doesn’t he learn?

But maybe the truth is that Trump keeps creating crises because he needs them.

The Trump candidacy was itself based on creating a sense of crisis. This was no small feat, given that Barack Obama was reasonably popular and the economy was growing. Yet Donald Trump adeptly manufactured a series of crises that helped convince slightly less than a majority of the country to vote for him, despite his manifold weaknesses as a candidate. He preached of a surge of immigrants, when border crossings were actually down; he painted a dark picture of widespread crime, even though crime rates are historically low; he warned of China stealing jobs and manipulating currency, when in fact both job loss to China and yuan devaluation had peaked years before. Crisis was the theme of both his nomination-acceptance speech at the Republican National Convention and his bleak inauguration speech.

“This American carnage stops right here and stops right now,” he said on January 20, 2017. Instead, his hyping of carnage has continued almost unabated, this week popping up as a counterpoint to wrenching stories about infants taken from their mothers at the border.

Conventional wisdom holds that at a time when the economy is booming, a president should be popular (Trump is not) and that a president would focus on that booming economy. (Trump isn’t doing that either.) He does fitfully boast about the economy, but it’s usually through the lens of complaining that the media won’t cover it. The media is usually too busy to cover it because they’re busy covering whatever crisis the White House has concocted that week. And the White House has concocted that crisis because it’s a way to bind the president’s supporters to him.

Trump is not a president for a time of smooth sailing and optimism; the only thing he shares with Dwight Eisenhower is a love of golf. Trump has styled himself a president for times of crisis. If things are good and boring, voters are apt to gravitate to a boring, competent administrator—a Jeb Bush or a Hillary Clinton, for example. Only in a moment of disaster would they gamble on Trump. Besides, Trump himself would be deadly bored if everything was going well.

So he keeps creating crises. His genius is focusing attention. He’s not an especially effective administrator of the federal government—it’s more shambolic than ever—and he’s not good at enacting laws, as demonstrated by his tiny list of legislative achievements. He can create a spectacle and draw eyeballs to it, though. Sometimes these crises work out well for him. His assault on NFL players over the national anthem seems to have riled up his supporters, and he successfully bullied NFL owners into a defensive position. In other cases, however, he overreaches. His statement that there were “very fine people on both sides” of the white-supremacist march in Charlottesville didn’t go over well. Neither, as it turned out, did separating infants from their parents.

Part of the reason is that even as Trump focuses attention, he’s not especially good at changing hearts and minds. This is why his approval rating remains low—though surprisingly stable, given the constant parade of crises.

Take immigration, for example. As Christopher Ingraham details in The Washington Post, the immigration crisis is not, in any sort of quantitative way, an actual crisis. There is no large upsurge in illegal immigration—apprehensions are near long-term lows; the numbers of immigrants entering the country are small in absolute terms; and despite the brutality of the gang MS-13, unauthorized immigrants appear to actually commit significantly less crime than the general population.

Yet Trump has managed to instill a sense of crisis around immigration. In 2016, just 7 percent of Americans told Gallup immigration was the country’s most important problem, placing it in the middle of the pack. In November 2017, it was down to 6 percent. By May, however, it was up to 10 percent, good for third place. (Looking at some of the other problems cited by Gallup respondents, it’s easy to see why Trump would rather focus on immigration: “dissatisfaction with government/poor leadership,” “race relations/racism,” “unifying the country,” “ethics/moral/religious/family decline,” “lack of respect for each other”—he doesn’t score well on any of these.)

Just because more people are worried about immigration doesn’t mean they agree with Trump. People feel strongly about family separations, but a majority are opposed. Opposition to the border wall remains strong. General support for immigration is soaring. The disconnect between making people pay attention to an issue and persuading people to agree with him is why Trump occasionally overreaches with his crisis creation.

Since the early days of the Trump administration, journalists, including me, have wondered what might happen when Trump faced a crisis that was not of his own creation. If he can bobble a self-created catastrophe, what would happen when he had to grapple with an unforeseen one? I’m starting to think I might have been wrong, though, in asking that question. Perhaps what’s really happening is that Trump is able to control what becomes a “real” crisis, too.

As Kombiz Lavasany pointed out when this idea came up again this week, Hurricane Maria’s impact on Puerto Rico was a genuine crisis. It’s possible as many as 4,600 Americans died because of the hurricane, and while storms are natural, the disasters that ensue are often the result of human decisions. The island remains devastated; water and electricity have been slow to return. Despite all this, Trump seems to have escaped the kind of opprobrium that George W. Bush received after Katrina. Some of the reasons for that are particular to Puerto Rico: The island is far away from the mainland U.S., it’s populated by minorities, and there seems to be some confusion about the fact that Puerto Ricans are American citizens.

Yet Trump’s de facto escape on Puerto Rico is also the result of decisions he made. When the storm hit, the president took widespread criticism for his detachment from the crisis, even as compared to his arms-length handling of hurricane landfalls in the Gulf Coast and Florida. When he finally did go to Puerto Rico, his visit was an embarrassment.

But Trump was playing a long game. By ignoring the storm for so long, he managed to shrink its perceived importance. When he finally did go, his appearance was painful—especially images of the president tossing paper-towel rolls to supplicating recipients like a baseline jumper—but that served to distract the focus from the squalor in Puerto Rico to Trump’s own gaucherie. That’s a success, after a fashion. By creating his own mini-crises, the president effectively marginalized a real one.

The family-separation crisis is not over. Thousands of children remain separated from their parents, with no clear plan for reunifying them, and the executive order that Trump signed seems vulnerable to judicial intervention. Even so, and despite the volatility of the Trump White House, it might even be possible to spot the next crisis coming over the horizon. Trump has started a multi-front trade war, which escalated this week when the president announced plans for a second round of $200 billion in punitive tariffs against China, which promised to respond in kind. Farmers, some of whom were Trump’s strongest supporters, are starting to get upset about the damage the trade war could inflict on them.

For the White House, the game plan seems simple, if not easy to execute: Create a trade war, solve it (either by cowing China or by somehow compensating farmers), and then declare victory and take credit. If it turns out that winning trade wars is not as easy as he has said, however, the president can always just concoct a new crisis.

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.