Among his many recent snafus, Donald Trump’s comments on foreign policy may be most revealing. His insistence that Russia is “not going into Ukraine” and his apparent belief that nuclear weapons exist to be used suggest he is at the very least untutored in foreign policy. Russia has been “in” Ukraine for more than two years, and the architecture of deterrence, which for decades has prevented wars destructive beyond imagining, demands that nuclear weapons remain grounded.

But even from the mouths of the plainly ignorant may arrive occasional nuggets of wisdom. Trump came close to one in a widely read July 21 interview with the New York Times. Of U.S. security commitments to NATO allies, the Republican nominee said, “I would prefer that we be able to continue, but if we are not going to be reasonably reimbursed for the tremendous cost of protecting these massive nations with tremendous wealth . . . then yes, I would be absolutely prepared to tell those countries, ‘Congratulations, you will be defending yourself.’”

To many politicians and pundits, this, too, appeared outlandish, a “dangerous,” “deadly,” “invitation to war” with Russia. Demonstrating, as close observers of foreign affairs have long known, that ideological competition is no obstacle to common cause against mutual foes, National Review joined Vox in vigorous condemnation. Republican senators Mitch McConnell and Lindsey Graham were no less opposed than nonpartisan diplomat Nicholas Burns. In her speech accepting the Democratic nomination, Hillary Clinton declared herself “proud to stand by our allies in NATO against any threat they face, including from Russia.”

The source of concern is easily recognized, which might explain the near unanimity of Trump’s rebuke. For decades, NATO has been the West’s force and fortress against first the Soviet Union and later Russia. This is thanks mainly to Article V of the treaty, which commits members to collective defense. The alliance rests on trust: the signatories’ firm understanding that this commitment will be honored.

Trump’s comments undermine this trust just as Russia appears especially threatening. Russians have annexed Crimea. Trump evidently does not know this, but they also have invaded Ukraine, where they directfinance, and fight alongside separatists. Russia operates unannounced combat-training exercises on the borders of Baltic NATO allies already shaken by these and other aggressive moves. Who could think of unraveling NATO at a time like this?

Sober foreign policy minds can see the logic in Trump’s approach, even if they don’t find it wholly agreeable.

But perhaps Trump’s proposal is not so outrageous. For one thing, his statements don’t actually renege on any commitments; Article V doesn’t in fact require that NATO members come to each other’s aid but instead allows each member to behave as it “deems necessary.” This is not a slip of words. The NATO charter explicitly diverges from its predecessor, the Treaty of Brussels, which obligated allies to respond with military force.

And sober foreign policy minds can see the logic in Trump’s approach, even if they don’t find it wholly agreeable.

“The generous interpretation of Trump’s remarks is that they are a bargaining tactic to frighten NATO allies into pulling their weight,” explains Alexander Downes, a political scientist and foreign policy expert at George Washington University’s Elliott School of International Affairs. “By confronting them with the specter of abandonment—leaving them to Putin’s tender mercies—the hope is that they will start paying their way and this will reduce the U.S. share of NATO expenses.”

It is a tactic with the potential to restore NATO to its intended functioning. When NATO was founded in the 1940s, the central goal was “to get the European states to pool their resources together to oppose the Soviet Union,” says Joshua Shifrinson, a political scientist at the Texas A&M Bush School of Government and Public Service. From the beginning, the United States supplied outsized military contributions, mobilizing fleets of ships, air battalions, and tens of thousands of personnel for exercises in Europe. Americans led NATO’s planning and movements. That was the arrangement: the powerful United States would, in its own and Western European interests, secure peace in Europe and check the Soviets. Heedful of the massive benefits, the allies would ease the burden of costs. Today, NATO continues to provide security against threats from the east, but now Europe relies on what Shifrinson calls “American largesse.”

As recently as the 2006 Wales Summit, NATO members committed to changing this. Each agreed to spend at least 2 percent of their respective national GDPs on defense, so that each would, in turn, be able to commit significant military capacity to the collective need. But, at present, only five of the twenty-eight NATO member states meet that condition, and the United States spends orders of magnitude more than most of the NATO allies. “European nations have cut their military spending by roughly 15 percent in real terms since the end of the Cold War,” MIT’s Barry Posen writes in a 2013 article in the journal Foreign Affairs. Japan, site of some of the largest U.S. bases abroad and one of the major beneficiaries of U.S. security guarantees, has consistently spent about 1 percent of GDP on defense per year, for decades. Posen has called security subsidies for Germany and Japan “welfare for the rich.”

NATO members do make direct financial contributions to the alliance, and some of our allies—within NATO and without—regularly pay a portion of U.S. expenses. But these payments are in the millions and low billions. The true bill is far larger, reflecting the costs of deploying abroad thousands of personnel and their equipment to train with allies and stand ready to fight for them. Taking the lead in NATO means the United States must also sustain a military force far larger than is needed to protect itself. This force must be able to deploy to an ally’s aid on short notice, necessitating a hair-trigger stance that also doesn’t come cheaply.

None of this considers lives lost, of course. Nor do cost-benefit analyses account for strategic blunders that might not have occurred without the blessing of multilateralism—such as the NATO wars in Afghanistan and Libya, the wisdom of which are still widely debated.

Even restricting ourselves to dollar terms, though, the price is likely enormous. “The bottom line on this question has always depended on what you count,” says Cindy Williams, a military budgets researcher at MIT’s Security Studies Program. But Williams estimates that the total cost imposed by alliance obligations accounts for roughly half of the U.S. defense budget, or about $300 billion per year in recent years. She points out, though, that if alliances were cheap or nonexistent, this money might be spent for other military purposes.

For Posen, a leading proponent of foreign policy restraint, NATO and other permanent alliances are not just a financial drain; they also arguably make Americans less safe. He points out that U.S. military commitments abroad have inspired defensive responses and alliances against the United States, such as Chinese and Russian cooperation in the UN Security Council. Both countries are spending on military improvements, and Russia sells arms abroad—sophisticated ones, in abundance—fueling terrorist threats to U.S. security.

Our friends also put the United States—and themselves—in difficult positions by provoking strong powers. “The Philippines and Vietnam (the latter of which has no formal defense treaty with Washington) . . . seem to have figured out that they can needle China over maritime boundary disputes and then seek shelter under the U.S. umbrella when China inevitably reacts,” Posen writes. He argues that the Republic of Georgia acted brazenly in its 2008 conflict with Russia over the disputed territory of South Ossetia because Georgia was “overly confident of Washington’s affection for it.” Russia’s victory provided a foundation of confidence on which President Putin has been building ever since.

“Restrainers are at best unenthusiastic about U.S. alliances, including NATO,” Downes says. “They’re not inherently anti-alliance, but [they are] against permanent alliances that are not a response to a real threat to interests.”

Elements of restraint have come to Washington in President Obama’s less intervention-prone second term. In some ways it looks like Trump is ready to continue what Obama started, even to plant his flag deeper in the restraint camp by cutting off costly, entangling alliances.

But there are reasons to believe otherwise.

Trump is inconsistent in his position on alliances. On the one hand, he claims “we have no choice” but to discard allies who won’t pay their way, and he calls out free-riding Japan, Germany, and Saudi Arabia. On the other hand, he has been nothing if not supportive of U.S. alliance with Israel, a wealthy and powerful nuclear-armed state whose defense costs the United States billions of dollars each year.

Whether this alliance benefits the United States, and in what ways, and whether it is a moral necessity even if not a political one, is as hotly contested as any other matter concerning Israel’s past, present, and future. But there is no doubt that exceptionally close ties with Israel strain other relationships, particularly in the Middle East and other heavily Muslim regions, that matter for the security of the United States and its allies. U.S. policy toward Israel is also a source of grievance among Islamist terrorists who have exacted a significant toll. Even so, Trump, whose rhetoric on terrorism is as pugnacious as anyone’s, has called Israel a “strategic ally.” Speaking to the American Israel Public Affairs Committee, he implied that the United States needed to place less pressure on Israel and said, “When the United States stands with Israel, the chances of peace really rise.” These statements play well in American elections, but a proponent of foreign policy restraint wouldn’t likely agree.

Nuclear weapons proliferation is another area in which Trump departs from the restraint position. While he has argued that Japan and South Korea should build nuclear weapons for their own defense, Posen considers “limiting nuclear proliferation” a key goal of restraint, responsive to one of the few significant threats to American power and territorial safety.

And, whereas Posen argues for decreased military spending, Trump wants to raise it, while slashing elsewhere.

Not every student of foreign affairs promotes restraint or its skepticism toward alliances. But even fellow critics of deadbeat allies don’t necessarily think Trump is taking the right approach to the problem.

Rajan Menon, a Russia specialist and author of The End of Alliances (2007), is no fan of the status quo, describing the NATO funding situation as “a bad deal for us” that “infantilizes Europe.” He wants to see “hard-nosed negotiations aimed at reforming NATO.” But, he adds, “Sensitive diplomatic issues require that the bullhorn be left at home.”

“You don’t address the burden-sharing issue in NATO by ultimata and proclamations,” Menon says. “Calling Article V into question will not reform the alliance and could destroy it.” Shifrinson agrees: “Trump is gambling that NATO will keep going in perpetuity and that his poker mentality is viable. But he misses that the other players might walk away from the table.”

Downes worries that some allies will proactively defect, emboldening aggressors. Undermining NATO solidarity “could cause Putin to behave more aggressively,” he says, drawing the U.S. into conflict with a capable enemy. “In the event of Russian aggression, President Trump—no matter what he believes—may not be able to resist the pressures to honor U.S. alliance obligations. Thus, publicly questioning the alliances could lead to preventable war.”

As Shifrinson sees it, the challenge is that you “need to scare the allies out of complacency if you want to get them to spend more on defense,” but “you can’t scare them so much that they won’t play along.” Trump and his supporters might contend that this is just what he’s doing, opening those hard-nosed negotiations on his terms.

But there is another possible explanation, one that fits snugly a reputation Trump has earned throughout the campaign. In whipping up discontent over the costs of NATO, he has simply isolated another problem that he can cast in his favorite terms: winning and losing. He may not be able to fix it, but he can offer a crude diagnosis. America is losing because it has to pay. “Our country doesn’t win anymore,” Trump said in a Republican debate last October. “We used to win, we don’t win anymore. We lose on trade. We lose with ISIL.” We lose on NATO, too, until someone else ponies up.

That is the sort of narrow and simplistic mindset—paying equals losing—that more calculating strategic minds will be only too happy to exploit.