Donald Trump’s first foreign trip as president has made clear that the only constant in his foreign policy is its lack of constancy.

Trump began last weekend with the Saudis, heralding a $350-billion arms deal and bowing ever-so-slightly before the Saudi king, despite previously having slammed President Obama for a similar head nod and having called the Saudis “people who push gays ... off buildings [and] kill women.” Trump then went to Israel, whose security establishment he’s said to have infuriated by reportedly sharing its extremely sensitive intelligence with top Russian officials—after previously declaring himself “Israel’s biggest friend.” Then came Brussels, home to NATO, which Trump had repeatedly said was “obsolete” before abruptly declaring it “no longer obsolete.” Yesterday, the pendulum swung back with a vengeance, with Trump scolding alliance counterparts for the “massive amounts of money” they purportedly owe the United States, and declining to reaffirm America’s commitment to collective self-defense.

Such reversals have become one of the few unifying features of Trump’s foreign policy. Trump was “going to get rid of NAFTA once and for all” before he agreed “not to terminate NAFTA at this time.” Similarly, Trump said the United States couldn’t take on Syrian President Bashar al-Assad in addition to ISIS, until Trump decided to launch 59 Tomahawk missiles at one of Assad’s airbases. In perhaps his quickest and most dramatic about-face, Trump anticipated a “major, major conflict” with North Korea until he said he’d be “honored” to meet with its leader.

It’s hard to keep up—not just for the American people, who want and deserve to know where their country stands on critical relationships with the rest of the world, but for foreign friends and foes alike. Trump thinks this kind of inconsistency is an asset. It’s actually a problem. And it’s the right time to scrutinize just why it’s such a problem as Trump and his team sift through the results of his first foreign trip and determine where to take his administration’s foreign policy from here.

Trump views unpredictability as a strategy unto itself. “We must, as a nation, be more unpredictable,” he said as a candidate. “We have to be unpredictable, and we have to be unpredictable starting now.” Thus, the Trump doctrine: Keep ‘em guessing. Push NATO to the brink, and it’ll spend more on counterterrorism. Push Canada and Mexico to the brink, and they’ll renegotiate NAFTA on terms more favorable to the United States. And so on.

But while it may lack the melodrama and ratings Trump seems to crave, the merits of predictability—at least as a baseline—in foreign policy should not be underestimated, as we learned firsthand as national security officials at the State Department and National Security Council, respectively.

Predictability keeps friends close. From Europeans who knew they could count on American protection from the Soviets throughout the Cold War, to Japanese and South Koreans who have long relied on American backing against China and North Korea, confidence in the United States has ensured steadfast American alliances, consistent American military basing, and enduring freedom of navigation for American commerce. Such confidence has also made it less likely that insecurity will drive dangerous decisions, like pursuing weapons of mass destruction or accommodating countries that may be hostile to the United States.

Likewise, when we worked on addressing diverse terrorist threats from Libya to Somalia to Yemen, we found that ensuring our partners understood exactly what actions we would be taking, and what support we would be providing to them, facilitated consent that provided the legal basis for critical U.S. strikes against terrorist targets. At the same time, it also boosted host countries’ own willingness to increase their investment in dealing with threats emanating from their soil. For example, Libyan Prime Minister Fayiz al-Sarraj has publicly encouraged and supported counterterrorism operations by groups aligned with his Government of National Accord. The operations have cost scarce lives and resources but have had significant payoff for the Libyan people in expelling ISIS from strategic locations in Libya thanks to steadfast, reliable U.S. air support that also avoided civilian casualties.

And predictability keeps foes at bay. If an adversary like North Korea understands how its actions will affect U.S. behavior, then that adversary, if rational, can be deterred, much as the Soviet Union was deterred from escalating in various proxy battlefields during the Cold War. But if a country like North Korea—or Iran—simply finds U.S. foreign policy erratic and unpredictable, then why make decisions based on inscrutable U.S. reactions, rather than simply pursuing one’s own interests as aggressively as possible?

Whatever one’s view of Trump’s decision to conduct strikes against the Assad regime after its appalling chemical weapons attack, the administration’s initially contradictory and ultimately incomplete explanation for that attack left no clear sense of whether it was intended to dissuade Assad from using chemical weapons again, to punish any attacks against civilians, or to provide a prelude to pushing him from power. Whether the strikes should, in the end, be considered a success will depend on Assad’s behavior going forward, which in turn will depend on which of those messages he heard.

Unpredictability can also be downright dangerous. Consider how North Korean leader Kim Jong Un might interpret being implicitly threatened with war one day and all but offered a state visit the next. If he believes the former, he might opt for a preemptive strike against American allies or military facilities in the region. If he believes the latter, he has no real incentive to abandon his nuclear weapons.

Unpredictability can be costly, too. The continuing uncertainty regarding NAFTA’s fate appears to have contributed to the double-digit percentage drop in the value of the Mexican peso after Trump’s election, a sharp plummeting of the foreign investment critical to the Mexican economy, and the anticipation of slowed economic growth and significant job losses in the country. These aren’t just problems for Mexico—they’re problems for the United States, too, and in particular for the many U.S. companies that do business with America’s third-largest trading partner.

To be clear, unpredictability (sometimes dubbed “strategic ambiguity”) can have an important role to play in astute foreign policy. Any negotiator finds it helpful, at the right moment, to sow doubt in the mind of whoever is sitting across the table as to whether one might simply walk away from negotiations. (Even the bumbling President Selina Meyer of Veep figured that out when she needed to bargain down the cost for the Chinese to free Tibet.) On more than one occasion during the Iran nuclear talks, Secretary of State John Kerry conveyed with great effect that the United States was ready to head home if Iran did not adjust its position. And, at a key juncture, it can of course prove useful to lead an ally to think that an alliance might fray unless that ally shoulders more of the load, or to cause an adversary to back down lest escalation lead to undesired conflict.

But unpredictability is a tactic, not a strategy. And it works only against a baseline of reliability. That’s why Veep’s President Meyer has her chief of staff exclaim in front of the Chinese, “She never does this!” when she wants to feign a willingness to break off negotiations—it’s precisely the exceptional nature of unexpected escalation that makes unpredictability effective in the right circumstances. But that rare moment must be a silhouette against a backdrop of reliability and clarity. Bluffs can only be called so many times. Repeated exposure as a bluffer erodes the value of the tactic. And unpredictability without a strategy to take advantage of it is just another name for confusion.

On the other hand, humdrum reliability and clarity offer far more sustainable foundations for successful foreign policy. If allies can rely on the United States to support them diplomatically, economically, and in appropriate circumstances militarily, they are more likely to take on greater responsibility for, say, countering terrorism within their own borders and for pursuing stability beyond them. In turn, America can then limit its own foreign adventures by relying on allies’ efforts rather than acting as a global enforcer itself—a burden-sharing objective Trump has rightly adopted.

If adversaries can rely on the United States to respond to aggression in predictable ways, its military and economic power can deter them from undertaking such aggression in the first place. Uncertainty, conversely, leads to perceived opportunities for those who would unsettle existing power structures and create instability—and stability tends to be good for U.S. security and good for U.S. investment.

Reports indicate that the Trump White House is hard at work on a new national security strategy, a project likely to become the administration’s foreign-policy focus now that the president’s first trip abroad is concluding. That document, for any administration, is an exercise in articulating how a president intends to handle a whole world of challenges. For a country as large as the United States, with myriad complex relationships and interests, it is inherently challenging to formulate a coherent approach that doesn’t sound like a laundry list.  

But, at a minimum, drafting such a document presents the opportunity to show allies and adversaries alike the ways, and circumstances, in which the United States can be relied on to support its friends and oppose its foes. Rolling out such a document to the public provides a much-needed chance for the president or secretary of state to make a first major address on the administration’s approach to foreign policy as a whole.

To provide that backdrop and back it up with actions, then occasionally sow doubt, is to be crazy like a fox. To confuse unpredictability with an actual national security strategy is to be just plain crazy.