Foreign Relations & International Law

Saving NATO

Scott R. Anderson
Wednesday, July 25, 2018, 9:00 AM

A recent series of statements by President Donald Trump has resurrected concerns about the United States’ long-standing commitment to the North Atlantic Treaty Organization. Ahead of the recent NATO summit, Trump wrote to several member states expressing “growing frustration” with their defense spending and warning that the U.S.

NATO Headquarters, Brussels (NATO Flickr,; CC BY-NC-ND 2.0 DEED,

Published by The Lawfare Institute
in Cooperation With

A recent series of statements by President Donald Trump has resurrected concerns about the United States’ long-standing commitment to the North Atlantic Treaty Organization. Ahead of the recent NATO summit, Trump wrote to several member states expressing “growing frustration” with their defense spending and warning that the U.S. commitment “is no longer sustainable.” He doubled down at the summit itself by threatening that, absent changes, the United States “would have to look to go its own way.” And while he closed the summit by signing a joint communique reaffirming the importance of the alliance and asserting that U.S. withdrawal was “unnecessary” due to NATO members’ commitments to increase their defense spending, Trump was unmistakably more amicable when he met with NATO rival Russian President Vladimir Putin shortly thereafter.

Last week, as he tried to contain the media fallout from his summit with Putin, Trump provided perhaps the most unvarnished picture to date of how he views the NATO alliance, stating the following in an interview with Tucker Carlson of Fox News:

CARLSON: Membership in NATO obligates the members to defend any other member who has been attacked. So let’s say Montenegro, which joined last year, is attacked: Why should my son go to Montenegro to defend it from attack? Why is that?

TRUMP: I understand what you’re saying. I’ve asked the same question. Montenegro is a tiny country with very strong people.

CARLSON: Yeah, I’m not against it—or Albania.

TRUMP: No, by the way, they have very strong people—they have very aggressive people. They may get aggressive, and congratulations, you’re in World War III. Now I understand that—but that’s the way it was set up. Don’t forget, I just got here a little more than a year and a half ago. But I took over the conversation three or four days ago and said, “You have to pay.”

President Trump’s comments about Montenegro—a country whose accession to NATO his administration approved in 2017, despite Trump’s recent suggestions to the contrary—reflect some fundamental misconceptions about NATO and the North Atlantic Treaty that established it. More importantly, they underscore how Trump sees the commitments that NATO membership entails: not as the lynchpin of a shared transatlantic strategy but as liabilities on the U.S. ledger sheet.

Commentators and policymakers of all political stripes have responded to Trump’s comments by roundly rejecting any suggestion that the U.S. commitment to NATO is wavering. Members of Trump’s own Cabinet reaffirmed the U.S. commitment to joint NATO activities. And Congress asserted its support through a pair of recent actions that passed unanimously in the House of Representatives and near-unanimously in the Senate.

But these efforts should not provide false comfort. Both international and U.S. domestic law give Trump an immense amount of control over the U.S. relationship with NATO. While some avenues may exist for constraining the president’s actions, they are limited and in many cases untested. This ability to act on his views despite widespread opposition is ultimately what makes Trump’s rhetoric so dangerous to the NATO alliance and its ability to deter external threats.

What Does NATO Membership Obligate the United States to Do?

The North Atlantic Treaty imposes several international legal obligations on NATO member states, which they have supplemented through follow-on agreements on topics as varied as information security and nuclear cooperation. The most fundamental obligation of all NATO members, however, is their common commitment to collective defense, which is famously codified in Article 5 of the North Atlantic Treaty. It states:

The Parties agree that an armed attack against one or more of them in Europe or North America shall be considered an attack against them all and consequently they agree that, if such an armed attack occurs, each of them, in exercise of the right of individual or collective self-defence recognised by Article 51 of the Charter of the United Nations, will assist the Party or Parties so attacked by taking forthwith, individually and in concert with the other Parties, such action as it deems necessary, including the use of armed force, to restore and maintain the security of the North Atlantic area.

This language does not require NATO members to go to war in defense of one another, as both Carlson and Trump implied. Instead, it imposes an international legal obligation—identifiable by the use of binding language such as “shall”—on member states to assist one another when one comes under armed attack while leaving it to each member state to determine what type of assistance is necessary and appropriate.

This is a less stringent collective defense obligation than other treaties that were concluded around the same time, such as the (now-defunct) 1948 Brussels Treaty, which obligated its state parties to provide any other state party with “all the military and other aid and assistance in their power” if attacked. Indeed, the United States negotiated for this less demanding language precisely because of political and constitutional concerns with any international agreement that purported to automatically commit the United States to war—a general tack it has pursued in negotiating similar collective defense arrangements with other key allies.

Why, then, is Article 5 so important? Because of the way it interacts with the U.N. Charter. While Article 2(4) of the U.N. Charter makes it a violation of international law for states to use force against each other where not authorized by the U.N. Security Council, Article 51 of the same carves out an exception when a state is exercising “the inherent right of individual or collective self-defence” in response to an armed attack. Normally, for a state to use military force in defense of another state (i.e., in collective self-defense), the latter must consent to the former’s actions on its behalf. Through Article 5, NATO members provide each other this consent in advance, allowing them to respond quickly to armed attacks on any of them without violating international law. Further, by explicitly linking itself to Article 51 of the U.N. Charter, Article 5 suggests that the commitment to provide assistance is limited to genuine situations of self-defense and would not apply to acts of aggression by a NATO ally, as President Trump’s Montenegro hypothetical suggests.

In other words, the primary effect of Article 5 is to enable, not obligate, U.S. military action and expand operational flexibility while remaining within the confines of international law. But this means that, for NATO membership to deter external threats as intended, the United States and other NATO allies must credibly communicate their intent to act on this operational flexibility. This is why NATO members have pursued interoperability, engaged in joint exercises and deployed substantial forces on each others’ territories throughout the alliance’s existence—all activities that Trump has increasingly questioned.

Can Trump Withdraw From NATO?

The Constitution does not define a process for withdrawing from treaties, leaving the question of who has the authority to do so in doubt. While historical practice has varied, the mainstream position—which the executive branch espouses and has repeatedly acted upon in recent decades—is that the president may do so unilaterally, without congressional involvement. Others contend that because the Senate’s advice and consent is constitutionally required to make treaties, withdrawal should entail some congressional role. Adherents to this latter view have even legally challenged treaty withdrawals for their lack of congressional approval, though the courts have thus far declined to intervene. As a result, the mainstream position has remained the operational one.

For NATO allies concerned primarily with international law, it’s largely irrelevant that precisely who can withdraw from treaties is unsettled under U.S. law. To prevent states from having to audit each others’ legal and political systems when entering into treaty arrangements, international law recognizes certain officials—including heads of state—as having the ex officio authority to provide and revoke their state’s consent. The only relevant limit on this authority applies where an official’s grant of consent to a treaty is in “manifest” violation of relevant domestic law provisions “of fundamental importance,” a very high bar unlikely to be met here. So even if withdrawal from NATO were contested in the United States, other member states are likely to treat any decision by Trump to withdraw as effective and adjust their understanding of their respective international legal obligations accordingly.

That said, there is one notable limit on this capability: Under Article 13 of the North Atlantic Treaty, any member state that wishes to withdraw from NATO must provide notice at least one year in advance. So as a matter of international law, the United States would remain bound by the North Atlantic Treaty’s obligations—including Article 5—for a year after any withdrawal announced by Trump. This does not necessarily guarantee that President Trump would respect those obligations. But the United States would be as bound to do so during that one-year period as it is now.

Could Congress Stop Trump From Withdrawing From NATO?

Trump has made clear that he believes he can withdraw from NATO without seeking congressional approval, consistent with the mainstream position discussed above. This leaves Congress without any clear mechanism for blocking such action. And even if Congress passed a law opposing withdrawal, it seems unlikely that the executive branch would consider itself bound to respect the legislature’s wishes. Instead, Congress would most likely need to challenge the constitutionality of the president’s actions in the courts—an approach that has failed in the past, though there is reason to think that certain challenges may have more success in the future.

The main obstacle to previous challenges to treaty withdrawals has been federal courts’ unwillingness to reach the merits. In its 1979 decision in Goldwater v. Carter, the Supreme Court dismissed the only challenge to a treaty withdrawal that it has directly considered as non-justiciable, though a majority could not agree on why. And while Goldwater did not create any binding precedent, when subsequent courts have considered the question—specifically the district courts in Beacon Products Corp. v. Reagan in 1987 and Kucinich v. Bush in 2002—they have followed the Goldwater plurality’s lead in holding that treaty withdrawal presents a non-justiciable political question. Citing relevant post-Goldwater precedent, Kucinich also held that members of Congress lack the individual and particularized injury necessary to have standing to sue on Congress’s behalf, highlighting another potential obstacle. And it voiced agreement with Justice Lewis Powell’s Goldwater concurrence that challenges to treaty withdrawals are not ripe for adjudication where Congress—as an institution, not just individual members—has not clearly and unequivocally acted in opposition.

That said, these barriers may not be insurmountable. The Supreme Court has narrowed the political question doctrine substantially since 2002, going so far as to hold in 2012 that it did not apply in arguably similar circumstances in Zivotofsky v. Clinton. This alone seems likely to weaken the case for the Goldwater plurality’s conclusion that treaty withdrawal is non-justiciable. To ensure standing, meanwhile, Congress—or even just the House or Senate—could vote to authorize a lawsuit on its behalf, which has proven successful at establishing standing in other cases involving Congress’s institutional interests. And either Congress as a whole or the Senate—depending on whether the theory is that treaty withdrawal cannot override congressional opposition or actively requires Senate advice and consent—could enact legislation or adopt resolutions making its opposition clear to avoid ripeness concerns. Moreover, both of these measures could be taken before a given treaty withdrawal, establishing a credible threat of a legal challenge that might serve as a potential deterrent. (Alternatively, if operating on the theory that treaty withdrawal requires Senate advice and consent, there is reason to believe that a lawsuit joined by as few as one-third of Senate members plus one—enough votes to withhold the Senate’s advice and consent—can satisfy standing and ripeness requirements, despite lacking the numbers necessary to adopt resolutions or other formal measures.)

Of course, reaching the merits is no guarantee of success. Under the Youngstown framework through which courts often view such issues, a president seeking to withdraw from a treaty over Congress’s clear objections—particularly where the latter is codified in statute—would undoubtedly be operating at the “lowest ebb” of his powers, meaning his actions would have to stem directly from his own constitutional authority. But the executive branch has successfully defended similar foreign-affairs-related claims of authority before, and there are strong arguments in support of its constitutional position. Among these is the fact that the constitutional process for making treaties bears an undeniable resemblance to the process by which the president appoints principal officers in the government—and his ability to remove the latter unilaterally, regardless of congressional opposition and without seeking Senate approval, is well-established. In this sense, pursuing a legal challenge may be a double-edged sword.

Can Congress Make Trump Respect U.S. Commitments to NATO?

Of course, formally withdrawing from NATO may not be necessary. The executive branch maintains that the president has the constitutional authority to disregard treaties and other international legal obligations that do not automatically have the force of domestic law (i.e., are not “self-executing”), a category into which the North Atlantic Treaty most likely falls. By simply refusing to adequately respond to threats against NATO allies—a possibility that Trump has repeatedly raised—Trump could render NATO as ineffective as he could by formally withdrawing. And it is unclear what, precisely, Congress could do to prevent this.

This is not true of all treaties. Some deal with subject matters that are within the scope of Congress’s constitutional responsibilities. For example, if the president were to withdraw from a treaty guaranteeing a foreign state certain trade terms, Congress could exercise its constitutional authority over foreign commerce and enact legislation reinstating those terms. The president’s withdrawal from the treaty might still end the United States’ obligations under international law, but Congress’s domestic laws directing actions consistent with the terms of the treaty would remain in place.

The problem with the North Atlantic Treaty—and with Article 5 in particular—is that it is primarily concerned with the use of military force. And in the United States, many decisions regarding how the military is used are seen as falling within the president’s constitutional authority as commander-in-chief. Indeed, Congress has embedded this view into the U.S. military’s chain of command, which places the president at the top and makes it a violation of the military justice system for U.S. service members to disregard any lawful order that is properly passed through the system. And while Congress has its own war powers, they are often better suited to obstructing military action than compelling it. For example, Congress can withhold funds to stop disfavored military activities, but its ability to use this authority to condition the provision of funds and direct executive-branch actions is limited to its own constitutional authority. Similarly, Congress can refuse to enact a declaration of war and thereby stop the president from engaging in at least certain types of major armed conflicts. But the extent to which such a declaration or similar legislation can actually compel the executive branch to undertake military action is at best unclear. At a minimum, long-standing practice would supports the view that the president has exclusive constitutional authority over much of the operational command and control of the military—if within certain bounds set by Congress—and the executive branch has not been shy about using this argument to push back when it believes Congress is out of its lane, such as in the regulation of military operations. And while the courts may not ultimately sustain this executive-branch position if challenged, it seems unlikely that the judiciary will resolve such questions in a timely and definitive manner. Thus the president is likely to retain substantial control over any immediate military response to a threat or armed attack against a NATO ally, even if Congress’s efforts to legislate an alternative approach are vindicated in the long run.

Congress also plays a role in structuring the responsibilities and authorities of members of the Defense Department and U.S. armed forces. At times, Congress has used this authority to make other officials, such as the secretary of defense, responsible for certain determinations or policy areas, thereby diminishing the exclusivity of the president’s decision-making authority. In theory, Congress may be able to use this authority to empower officials who are more likely to support fulfilling U.S. commitments to NATO; in practice, however, such efforts are likely to encounter substantial limitations. Any such structure remains subject to the uncertain limits of Congress’s constitutional authority, and the executive branch has frequently asserted its intent to interpret such provisions narrowly where it believes Congress has overstepped its bounds. Further, using his appointment power, the president may remove and replace officials who refuse to execute his chosen policy. And as mentioned above, U.S. service members are legally obligated to obey any lawful order given by the president that is communicated down the chain of command while civilian officials may feel duty-bound to resign when they cannot do so. As a result, congressional efforts to structure policymaking may succeed at empowering supportive policymakers in interagency discussions or raising some political costs of not fulfilling U.S. NATO obligations—for example, by requiring controversial public determinations or triggering high-profile resignations—but they are unlikely to reverse the president if he is strongly committed to a different course of action.


What do the answers to these questions mean for the future of NATO? At a minimum, that President Trump’s words should not be taken lightly, even if they are extreme outliers on the U.S. political stage. To the contrary, Trump’s skepticism of NATO should be deeply concerning, as his support is all-but necessary for any coordinated NATO military response involving the United States—the outcome that NATO is intended to facilitate. Moreover, Trump’s willingness to express this skepticism so openly—to say nothing of his courting of NATO rivals—weakens the political commitments on which NATO’s deterrent effect is based, inviting the risk that a rival will test its limits. And this risk stands to increase as Trump’s time in office grows shorter and NATO rivals see this unprecedented window of opportunity narrow.

Those wishing to preserve the United States’ relationship with NATO may draw a second lesson: The law does not provide any easy solutions. While some legal responses may be worth pursuing, the bulk of domestic and international legal authority—and the political and institutional context in which it operates—reinforces the president’s authority over how the United States engages with NATO. For this reason, any effective response to Trump’s posture on NATO will need to be primarily political. Supporters of continued U.S. engagement with NATO—particularly those in Trump’s own party, who are in a better position to influence him—will need to undertake the hard work of persuading Trump to change his behavior and punishing him if he does not. Demurring in the hope that other actors and institutions will somehow restrain the president’s worst instincts ignores how the relevant legal authorities are structured—and those who do so will rightfully share responsibility with him for the eventual outcome.

Scott R. Anderson is a fellow in Governance Studies at the Brookings Institution and a Senior Fellow in the National Security Law Program at Columbia Law School. He previously served as an Attorney-Adviser in the Office of the Legal Adviser at the U.S. Department of State and as the legal advisor for the U.S. Embassy in Baghdad, Iraq.

Subscribe to Lawfare