What Congress Has Done—and What It Still Needs to Do—to Protect NATO

Former President Trump’s recent suggestion that he would “encourage [Russia] to do whatever the hell they want” to NATO allies who are “delinquent” in meeting long-standing defense spending targets has resurrected concerns about what a second Trump presidency might mean for the United States’s most important alliance. Trump’s views on NATO are nothing new: He has long been openly skeptical of the alliance, and his former advisers have warned that his repeated threats to withdraw from the alliance’s foundational North Atlantic Treaty while president were serious. And now that he has emerged as the presumptive Republican nominee for president once again in 2024, supporters of NATO on both sides of the Atlantic—and both sides of the aisle—are worried he may actually follow through if he returns to the White House.

Fortunately, supporters of continued U.S. participation in NATO can take some solace in knowing that Congress has recently taken steps to help protect the alliance. Late last year, Congress enacted a statutory provision—located at Section 1250A of the most recent National Defense Authorization Act (NDAA)—that expressly prohibits the president from withdrawing from NATO or using any appropriated funds for that purpose without permission from Congress. The provision, which is based on a proposal that I first outlined in Lawfare several years ago (and have advocated for since), not only puts the president’s legal authority to exit NATO on the narrowest possible grounds but also helps clear the way for a potential legal challenge if a future president chooses to proceed in spite of it.

But the legal protections installed by Section 1250A are incomplete. During congressional debates, a key provision was dropped that would have preauthorized litigation to challenge any presidential effort to exit NATO contrary to this provision. Absent such authorization, it is unclear who might have the standing necessary to challenge the legality of an attempted unilateral exit from NATO in federal court—even one that is ultimately unlawful. 

If Congress wants to strengthen the protections it has installed around U.S. participation in NATO, authorizing this litigation should be its first step. Doing so will make clear that no president will be able to exit NATO unilaterally without having to defend the constitutionality of their actions in federal court. Moreover, it is a step that either the House or the Senate could choose to take on its own—and one that the current Senate might be able to put in place through the next presidential term and beyond. 

What Section 1250A Does

The proposal that ultimately became Section 1250A was designed to put opponents of an exit from NATO—and specifically Congress—in the strongest possible position to mount a legal challenge against a president who attempts to withdraw from or otherwise exit NATO without Congress’s permission. While the exact language changed and evolved over the course of the legislative process, this basic strategy remained more or less in place. In its final form, the key operative language of Section 1250A states:

The President shall not suspend, terminate, denounce, or withdraw the United States from the North Atlantic Treaty, done at Washington, DC, April 4, 1949, except by and with the advice and consent of the Senate, provided that two-thirds of the Senators present concur, or pursuant to an Act of Congress.

These four verbs—suspend, terminate, denounce, and withdraw—are all terms of art describing different ways in which states may exit (or suspend) their treaty obligations under international law, the definition of which Section 1250A expressly ties back to the Vienna Convention on the Law of Treaties. A subsequent provision of Section 1250A similarly establishes that “[n]o funds authorized or appropriated by any Act may be used to support, directly or indirectly, any decision on the part of any United States Government official to” pursue any such actions, bringing Congress’s separate authority over appropriations to bear on the matter as well. Another directs the president to consult with relevant congressional committees and notify them in writing at least 180 days in advance of taking any of these steps.

To understand the significance of this relatively straightforward language, one has to understand the legal context in which it is operating. While the Constitution is clear that the president may enter the United States into treaties like the North Atlantic Treaty only with the advice and consent of two-thirds of the Senate, it’s silent on how the United States should exit them. (Modern presidents do enter into certain types of international agreements with authorization from Congress or on their own authority, but that was not the case with the North Atlantic Treaty.) The framers seemed split on the issue, with some treating it as a shared authority between the executive and legislative branches while others framed it as a strictly legislative responsibility. Early American practice was similarly diverse, with Congress or the Senate sometimes approving of treaty exits in advance and other times concurring to them after the fact. But as legal scholar Curtis Bradley has written, “Historical practice through at least the late nineteenth century suggests an understanding that congressional or senatorial approval was constitutionally required for the termination of U.S. treaties.” 

This changed around the turn of the last century, when the president began to assert the authority to suspend or withdraw from treaties on his own without any congressional involvement. This practice became more commonplace following World War II and was rarely challenged by Congress as a whole, leading it to gradually become more widely accepted. Today, the conventional wisdom is that the president generally has the legal authority to suspend, terminate, or withdraw from a treaty on behalf of the United States where authorized by the treaty’s terms or other applicable international law without having to secure any concurrence from Congress. This in turn means that the president may “denounce”—an antiquated term that essentially means to withdraw from—the North Atlantic Treaty one year after providing notice of its intent to do so, as set out in Article 13 of the treaty itself.

This authority is, however, subject to an important caveat. When it comes to the exercise of implied authorities not expressly provided for in the Constitution, the Supreme Court has suggested that “[p]residential powers are not fixed but fluctuate, depending upon their disjunction or conjunction with those of Congress.” Within this framework—first articulated by Justice Robert Jackson in his 1952 concurrence in Youngstown Sheet & Tube Co. v. Sawyer and since applied by the contemporary Court—“congressional inertia, indifference[,] or quiescence may sometimes, at least as a practical matter, enable, if not invite, measures on independent presidential responsibility.” But where a president seeks to “[take] measures incompatible with the expressed or implied will of Congress[,]” his power is at its “lowest ebb” and any such claim of “so conclusive and preclusive” an authority “must be scrutinized with caution[.]” In other words, there may be certain implied powers that a president can exercise where Congress is silent that he cannot exercise where Congress has expressed its will to the contrary. 

For their part, the federal courts have never clarified whether the president’s authority over treaty exit is this sort of implied power or whether it is something over which the president does indeed have “conclusive and preclusive” constitutional authority. And this is not for lack of trying on the part of plaintiffs, a diverse array of whom have challenged treaty withdrawals in the past only to encounter justiciability barriers. In Goldwater v. Carter, the only legal challenge to a treaty withdrawal to reach the Supreme Court, the justices declined to reach the merits because some believed it presented nonjusticiable political questions and because others felt that the dispute was not ripe for adjudication as Congress as a whole had not acted in opposition to the withdrawal in question. Lower courts have generally followed these models and dismissed subsequent challenges on similar grounds.

Section 1250A, however, improves the odds that a future legal challenge over a possible exit from NATO will be able to clear these justiciability barriers. By enacting Section 1250A and its express prohibition on withdrawal from NATO, Congress as a whole has undoubtedly “rejected the President’s claim” of authority and chosen to “confront [him]” in relation to any withdrawal from the North Atlantic Treaty, meeting the conditions for ripeness spelled out in Goldwater. And recent case law strongly suggests that the political question doctrine should not stand as a barrier “when an Act of Congress is alleged to conflict with the Constitution”; instead, the Supreme Court has emphasized that “it is emphatically the province and duty of the judicial department to say what the law is” in such cases, a responsibility the courts cannot shirk “merely because the issues have political implications.”

Perhaps more importantly, by expressly and unequivocally prohibiting the president from exiting NATO without Congress’s consent, Section 1250A also puts the president’s authority to exit NATO at its “lowest ebb.” Under the Youngstown framework, this means that, to uphold the president’s actions, a federal court would have to conclude that the Constitution gives the president the exclusive authority to exit treaties like the North Atlantic Treaty. 

For its part, the Trump administration—and perhaps the executive branch as a whole—has argued that this is exactly what the Constitution does. As it prepared to leave office in December 2020, the Trump-era Office of Legal Counsel (OLC) released a legal opinion asserting as much, specifically in defense of President Trump’s decision to withdraw from the Open Skies Treaty while disregarding a statutory provision requiring that the president provide advance notice to Congress before initiating such a step. While congressional leaders objected to the legality of Trump’s actions, no one challenged them in the courts, leaving this legal position untouched by judicial review. And because the Biden administration has not yet rescinded or modified this opinion, it appears to remain the stated position of the executive branch. 

Whether this view would be vindicated by the federal courts is unclear. But in my view, given the Constitution’s textual silence on the topic, the contrary views of several framers, the long history of congressional involvement in treaty exit up through the 20th century, and Congress’s central role in managing other aspects of U.S. involvement in international agreements, I believe it would be a much harder argument for the executive branch to win on than the 2020 OLC opinion acknowledges. When I convened a panel of leading legal scholars and practitioners to debate the constitutional effectiveness of an earlier version of Section 1250A in 2019, this is more or less where they came out as well.

At a minimum, exiting NATO in violation of Section 1250A would be far from a risk-free proposition for the executive branch. The resulting legal challenge would be unlike anything the executive branch has encountered before and would pose a much more substantial risk that courts will reach the merits. Perhaps the executive branch’s exceptionally broad view of presidential authority over treaty exit will be vindicated. But there is a substantial chance that federal courts will find that the president’s usual ability to exit treaties—unsupported by the constitutional text and inconsistent with interbranch practice for the first half of American history—is limited to circumstances where Congress is silent and does not apply here. In other words, by teeing up a serious and credible legal challenge along these lines, Section 1250A imposes meaningful costs and risks on the executive branch in the event it attempts to unilaterally exit NATO—something that may help deter the executive branch from taking such a step in the first place.

What’s Missing

There is, however, one major piece missing from Section 1250A. As originally passed by the Senate, Section 1250A didn’t just tee up a successful lawsuit: It also took the novel step of preauthorizing litigation on behalf of Congress in the event a president attempted to exit NATO without congressional consent. This provision was ultimately removed from Section 1250A during conference, at the apparent request of the House. But without it, the legal regime Section 1250A sets up—and the deterrent effect it is intended to have on the executive branch—is incomplete.

The omitted provision was intended to address a final justiciability barrier that has stopped prior legal challenges to treaty exits: a lack of standing. Federal courts require that plaintiffs suffer a “concrete and particularized” injury before they can challenge allegedly unlawful action. In the case of NATO, it’s not clear who might have the standing necessary to challenge an attempted exit, as a “generalized interest” such as that in broadly applicable foreign policy is generally considered to be insufficiently particularized for standing purposes. Some service members or U.S. nationals with relatives in Europe might be able to make a successful case for standing to challenge an exit from NATO, but it’s far from certain. Nor is it clear they would be interested in undertaking such a legal challenge, even if it were legally possible. And unless someone with the necessary standing is willing to challenge an attempted exit from NATO in the federal courts, then even an unlawful one may well prove effective simply by evading judicial review.

There is, however, one party that would have a strong case for a particularized injury: Congress itself. After all, by exiting NATO unilaterally, the president would have not only violated Section 1250A but also bypassed Congress’s own claimed constitutional role in treaty exit. 

The Supreme Court, however, has interpreted legislative standing narrowly. Individual legislators, it has held, generally lack standing to sue over injuries to the institutions of which they are a part, on the logic that their injury is “wholly abstract and widely dispersed” among that institution’s members, not particularized to those individuals. The one exception is in cases of “vote nullification,” where a given legislator or group of legislators allege to have had the legal effects of their votes effectively ignored. To make this showing, the plaintiffs must demonstrate that they had sufficient votes to change the outcome of the institutional decision in question. This has unique relevance in the treaty context. Because as some legal scholars maintain that the Constitution requires that two-thirds of the Senate provide their advice and consent to exit any treaty (just as they did to enter into that treaty), as few as 34 members of the Senate—the one-third plus one needed to withhold such advice and consent—would arguably have standing to challenge a unilateral treaty exit on this legal theory. That said, the legal challenges to treaty exit that legislators have pursued in the past have not fit this model or made this argument. 

The Supreme Court has also recognized that standing operates differently where legislative plaintiffs are “authorized to represent their respective House of Congress,” as the House and Senate themselves also suffer particularized injury where their constitutional roles are bypassed. The same is most likely true of Congress as a whole. In more recent cases, the Supreme Court has suggested that the logic of vote nullification applies to these institutions as well, meaning they can pursue legal challenges only where they can claim to have had the ability to change the outcome if their institutional role had been respected. But where these conditions are met, congressional institutions are generally understood to have the standing necessary to enforce their legal interests in the federal courts.

As originally drafted, Section 1250A sought to maximize the strength of this argument for legislative standing by seeking express authorization for litigation on behalf of Congress as a whole. To my knowledge, this would have been the first time Congress has undertaken such a step—but if it wishes to vindicate its own constitutional authority in the federal courts, there is a good reason for Congress to pursue it. Even the most ardent judicial skeptics of congressional standing have expressed an openness to the idea that Congress has standing to vindicate its constitutional interests where doing so is statutorily authorized. And Congress as a whole would be in the best position to assert its claimed constitutional role in treaty exit (as well as in the legislative process that produced the limitations articulated in Section 1250A), alongside whatever interests the House and Senate might have individually.

Importantly, by preauthorizing litigation—instead of waiting to authorize litigation until a violation of Section 1250A takes place—Congress would also signal a strong commitment to follow through with a legal challenge if a president were to pursue a unilateral withdrawal. This is especially important as partisan political dynamics may complicate any future vote to authorize litigation after a president acts. Committing in advance to pursue a legal challenge if and when a president pursues a unilateral withdrawal not only avoids partisan political constraints but also increases the credibility of the threat of litigation and amplifies its potential deterrent effect.

By contrast, without this provision, it’s not even clear that presidents acting at the lowest ebb of their authority in unilaterally exiting from NATO in violation of Section 1259A would have to defend their actions in court. 

Next Steps

Fortunately, it is not too late for Congress—or the House and Senate individually—to remedy this omission. The simplest solution would be for Congress to enact the litigation authorization provision that was removed from Section 1250A. There is no reason that provision had to be adopted alongside the rest of Section 1250A, and it would have the same effect if enacted separately today. That said, whatever concerns led the House conferees to oppose including this provision in the 2024 NDAA most likely remain in play. Hence, this may not be a politically available option, even if it is the most legally straightforward.

But there are other ways Congress could proceed that may not raise these concerns. The House and Senate could put through a concurrent resolution approved by both chambers without having to submit it to the president for a possible veto—a measure that would simply be authorizing legal action on Congress’s behalf, which usually does not require legislation. This may cause issues if the House and Senate want the Office of Senate Legal Counsel to be involved in representing Congress in this regard, as its authorizing legislation—which is far more detailed than the comparably pithy statutory provisions governing the House’s Office of General Counsel—only allows it to pursue litigation authorized by “resolution of the Senate[.]” That said, it’s not clear whether this statute has to be read so narrowly as to exclude concurrent resolutions. Moreover, Congress could most likely authorize someone else to represent it in litigation if the Office of Senate Legal Counsel could not.

However it’s done, Congress as a whole is undoubtedly in the best position to assert that a unilateral withdrawal from NATO injures its legally protected interests in a manner that establishes standing. But if bicameral action proves unworkable, there is reason to think that either the House or the Senate could have the standing necessary to pursue their own legal challenge as well.

A panel of the U.S. Court of Appeals for the D.C. Circuit suggested as much in a 2020 opinion relating to former President Trump’s attempts to build a border wall in spite of restrictions in related appropriations legislation. Specifically, it held that—while “Congress does not have standing to litigate a claim that the President has exceeded his statutory authority” under long-standing D.C. Circuit precedent—both chambers have independent standing to sue over “institutional injury to [their] own institutional power to prevent the expenditure of funds not authorized,” as the Appropriations Clause requires affirmative action by both chambers to authorize any expenditure, giving each chamber the authority to halt or restrict such appropriations. Ignoring a statutory limitation on appropriations thus nullifies each chamber’s vote, “caus[ing] a concrete and particularized constitutional injury that [each chamber] experiences, and can seek redress for, independently.” And while this opinion was later vacated for unrelated reasons, this did not occur until after the broader D.C. Circuit denied a petition for rehearing en banc—suggesting that the panel’s views most likely aligned with those of the broader D.C. Circuit, whose composition has not substantially changed. 

The D.C. Circuit is the same appellate court that would be most likely to hear any House or Senate legal challenge to any attempted withdrawal from NATO in violation of Section 1250A. And as Section 1250A includes a provision barring the use of any appropriated or authorized funds in exiting from NATO without congressional support, any violation of Section 1250A would also constitute a violation of the Appropriations Clause—giving each chamber independent standing to pursue a legal challenge. 

Pursuant to its rules, the House has delegated the authority to authorize litigation on its behalf to a Bipartisan Legal Advisory Group (BLAG). While there may be some limits to this delegation, the BLAG has successfully authorized a range of litigation on the House’s behalf in recent years. In the event of a unilateral exit from NATO in violation of Section 1250A, the BLAG could most likely authorize litigation challenging this move on the House’s behalf. Alternatively, the House as a whole could provide this authorization through a House resolution, a move that would arguably put the legal challenge on firmer legal footing by avoiding any questions regarding the scope of delegation to the BLAG.

Preauthorizing litigation is more difficult, however, as the House traditionally reconstitutes itself with each new Congress every two years. Not only does each House adopt new internal rules—though they are often quite consistent across congresses—but the actions of a prior House do not carry forward into the new House. This includes authorizations to engage in litigation. In recent years, the House rules have included provisions designed to allow for a new House to continue litigation initiated by its predecessor without any major disruption in litigation, but these provisions apply only after the litigation has begun. Any preauthorization for the House to engage in litigation over Section 1250A, by contrast, would almost certainly have to be readopted by each Congress. That said, if incorporated into the House rules and carried forward, this could be done alongside other standing measures.

For its part, the Senate does not broadly delegate litigation decisions to a BLAG-type body. By statute, the Office of Senate Legal Counsel can pursue certain types of litigation on the Senate’s behalf at the direction of a supermajority of the Senate’s Joint Leadership Group (JLG) and others pursuant to a Senate resolution, though neither category clearly encompasses affirmative litigation in which the Senate is the plaintiff for purposes other than subpoena enforcement. That said, if there were concerns as to the Office of Senate Legal Counsel’s ability to represent the Senate in such a legal challenge, then the Senate could presumably authorize a third party to represent it instead, as it can do in certain other litigation contexts. In either context, this will most likely require a Senate resolution to this effect.

While it does not delegate litigation decisions as the House does, the Senate is likely to have an easier time preauthorizing litigation. Unlike the House, the Senate is a continuing body, meaning that Senate resolutions adopted in one Congress can carry over indefinitely into future congresses, at least where this is the Senate’s stated intent. Indeed, this is common practice for the Senate, which maintains a number of these “standing orders” governing everything from the names of Senate office buildings to rules governing television and radio broadcasts of Senate proceedings. At least one of these standing orders authorizes various types of litigation on the Senate’s behalf, providing a clear precedent for a standing order preauthorizing litigation over an exit from NATO. This means that the current Senate could most likely adopt a Senate resolution making it a standing order to preauthorize litigation on the Senate’s behalf over a violation of Section 1250A—and that this authorization would in turn carry forward into future congresses, until countermanded by some other Senate action. For this reason, barring action by Congress as a whole, preauthorization by the Senate is likely the next best option.

***

No one should be under the illusion that Section 1250A is some sort of panacea. Even if he cannot exit NATO, a future President Trump will still be able to undermine it substantially. His authority as commander in chief almost certainly means that he will be able to prevent or limit the use of the U.S. armed forces to defend NATO allies if he wishes, even if doing so puts the United States in violation of its treaty obligations (and it’s not clear it would). And while there may be steps that Congress should be able to compel the executive branch to take—including the provision of certain types of foreign and security assistance—a president determined to be an obstacle could almost certainly interfere with and delay such efforts, at least until the federal courts resolve the array of constitutional questions such an unprecedented scenario would likely raise. As I wrote in 2018, “The law does not provide any easy solutions” to a scenario in which the president is eager to undermine NATO; instead, “any effective response … will need to be primarily political.”

But this does not make Section 1250A of no consequence. So long as the United States remains a part of NATO, there will still be working-level mechanisms of cooperation and coordination that will help to advance its members’ collective security. And even if a future President Trump proves unwilling to live up to U.S. commitments to our NATO allies, his successors may not. Preventing an untimely U.S. exit from NATO is the best way to preserve the alliance through the medium and long terms, even if it suffers disruptions in the short term. This makes the legal fight over President Trump’s authority to unilaterally exit treaties like NATO one worth fighting.

Congress has already set the stage for this struggle by enacting Section 1250A. Now the House and Senate need to make clear that they are ready and willing to bring that fight all the way to the federal courts, even (or especially) if no one else can. Absent such a measure, Section 1250A may ultimately prove ineffective. But with it, Congress—or even just the Senate—may be able to safeguard continued U.S. participation in NATO for another generation, no matter who is elected into the White House in 2024.