Editor’s note: This article concludes the Just Security Symposium on Harold Hongju Koh’s “The National Security Constitution in the 21st Century”.
For those committed to the rule of law, almost palpable despair has set in as the world awaits President-elect Donald Trump’s second term. But rather than surrender to it, what short- and long-term strategies should rule of law advocates pursue?
This essay addresses in Part I the short-term challenge of the Trump 2.0 years: combining resistance to any anti-democratic actions with resilience to stem the likely influx of illegal initiatives, as I previously explored in The Trump Administration and International Law. But even while rule of law advocates are plugging the dam, they should not ignore longer-term structural issues. The second, longer-term challenge—addressed more recently in The National Security Constitution in the 21st Century—will be simultaneously rebuilding the constitutional dam by pursuing structural reform to restore institutional checks and balances between the branches of government. During the next few years of legislative deadlock and intense political polarization, comprehensive national security legislative reform will undeniably be difficult to obtain. Instead, my book—and the various responses to it in this Symposium and elsewhere—suggest a mosaic of reforms to be implemented in each branch over time (263-326). As Part II below suggests, the goal is gradually to retrofit all three branches to dampen the dysfunctional institutional interaction that keeps driving presidents to act unilaterally, Congress to do nothing, and the courts to rubber-stamp and defer. Even if adopted individually, over time, these reforms should start working collectively to counteract the institutional incentives that skew the current system toward executive unilateralism.
I. Stemming the Flood: Resilience and Resistance
Let’s start by recognizing that “it’s never as good as it looks or as bad as it seems.” Trump’s narrow congressional margins of victory limit his freedom and make it unlikely that his party’s two-chamber control will last beyond two years. Rather than being overawed by the looming challenge, rule of law advocates should approach those years by emulating the stranded astronaut in The Martian: “You solve one problem and solve the next one, and then the next. And if you solve enough problems, you” can get past the immediate crisis.
We should not forget that the rule of law problem the United States faces is bigger than Trump. Regardless of who occupies the White House, America faces a structural problem that transcends personalities: the Office of the President has become our most overlooked national security threat. With each successive presidency this century, the constitutional pendulum has swung further and further toward executive unilateralism, climaxing in Trump’s breathtaking assertion that Article II “gives me the right to do whatever I want.” Early announcements forecast that Trump’s second term will greatly exacerbate that threat. But we can already look past Trump to imagine even more dangerously unilateralist presidents: populist autocrats determined not just to disregard constitutional checks and balances, but more systematically and ruthlessly to invade foreign countries, shatter alliances, and undermine international law.
Unlike in 2017, some of Trump’s likely moves challenging the rule of law can be anticipated based on his track record. He is likely to favor impulse over strategy, transactions over relationships, hard power over diplomacy, and “resigning without leaving” over meaningful engagement with international institutions. He will most likely continue to denigrate the truth, lying relentlessly to distort reality. He is also likely to attack the press, diplomacy, and government bureaucracy in an effort to reverse Biden policies and confound longstanding bipartisan foreign policy commitments with which he disagrees. By targeting and demonizing opponents, he will discourage collective resistance to these actions by encouraging anticipatory capitulation. He will use “shock and awe”— what I once heard Lucas Guttentag call “public spectacle and hidden terror”—to encourage isolation, self-silencing, and self-exit of those committed to defending the rule of law, undocumented immigrants from the United States, and career civil servants from the U.S. government. We can count on him to “flood the zone” with numbing and distracting daily outrages. Reckless threats to impose massive North American tariffs, or to buy Greenland and the Panama Canal, will seek to shift the “Overton Window,” so that the previously unimaginable becomes the new normal: as he has already done by lowering the standard for Senate confirmation from “having a worthy Attorney General” to “better than Matt Gaetz.”
In the waning days before inauguration, Biden officials should take every available preventive measure to lock in rule-of-law-promoting measures, in a way that maximizes the difficulty of unilateral reversal. Examples include dramatically reducing the number of detainees remaining at Guantanamo, embedding civilian harm incident reduction (CHIRG) policies into Executive Orders, placing more human rights conditions on military aid to Israel, and various urgent Ukraine measures: sending last minute aid to Ukraine, publicly urging Ukraine’s admission to NATO, and transferring more income derived from frozen Russian assets. Of course, Trump can undo some of these measures, but it will take time to do so, running down his clock to do damage, while creating opportunities to generate opposing litigation. Meanwhile, Congress should rebuff imprudent calls, after the New Year’s New Orleans attack, to give Trump a “9/11-type blank check,” or to rush to confirm Trump’s national security team without careful process.
Once Trump assumes office, the best strategy for rule of law advocates remains using the law and bureaucratic inertia to check anti-rule of law initiatives. Trump aspires to two years of “frictionless government,” contrary to Justice Louis Brandeis’s observation in Myers v. United States that “the purpose of separation of powers is …‘by means of the inevitable friction incident to the distribution of the governmental powers among three departments, to save the people from autocracy.’” As Ashley Deeks and Kristen Eichensehr have astutely warned, when one party dominates all branches of government, or when no partisan disagreement emerges on a particular issue (e.g. getting tough on China),
[t]he disappearance of partisan, interbranch, and interagency checks can amplify cognitive biases that often arise in decision making, including groupthink, and result in governmental actions that spark or escalate conflict, trigger actions by U.S. adversaries that undercut U.S. security goals, and unlawfully target domestic constituencies perceived to be linked to foreign adversaries.
The president can be restrained, they urge, by building into standing institutions the “friction imposed by the political branches on themselves (or on each other) and sources of friction external to the U.S. government.”
But as Trump 2.0 unfolds, both kinds of friction can and should arise. Trump’s supporters have already announced disturbing plans to reduce law enforcement independence, relax restraints on conflicts of interest, and restrain domestic military adventurism. Any abuse of the Insurrection Act, the Alien Enemies Act, and the National Guard will surely attract intensive litigation. His high-profile pledge to improve government efficiency could create exploitable opportunities for bureaucratic reform that could refashion executive mechanisms to promote better national security legal advice. A Congress with key members skeptical of Trump’s national security excesses could adopt the core legislative proposals that I suggest and Deeks and Eichensehr support: a Joint Committee for National Security, a Congressional Legal Adviser, and better congressional tools to create the frictions that can restrain executive unilateralism.
Over time, as Mike Johnson’s narrow re-election as House Speaker has already exposed, other frictions will arise: some naturally, and others caused by resistance, not just from across the political aisle, but among the factions who make up Trump’s variegated support. Trump’s overconfidence will lead his team to take extreme first moves on such issues as immigration, tariffs, and tax cuts, that will hurt middle-class Americans on pocketbook issues. As Trump overplays his hand—as he has already done with his preposterous intended nominations of Gaetz, Pete Hegseth, Kash Patel, Robert F. Kennedy, Jr., and Tulsi Gabbard, among others—alienation may set in among his most marginal backers. In time, Albert O. Hirschman’s notion of “countervailing passions” will likely kick in, so that Trump’s extreme “movement supporters” will increasingly disagree with those narrowly interested in seeking personal gain from his reascendancy.
As in Trump’s first term, even more frictions will arise as institutional counterweights coalesce and mobilize on a repeat-player basis: states and localities (especially on climate change and immigration), uncowed independent media, rule of law defenders within international institutions and regimes, powerful private interest groups (including some technology companies, concerned former military, and democracy and civil liberties organizations), and nongovernmental rule of law-promoting institutions. Bureaucratic resistance, leaking, and whistleblowers—such as those who reported Trump’s coercive 2019 conversation with Ukraine’s President Volodymyr Zelenskyy—will help make executive branch actions more transparent.
Different sources of friction will arise on the international plane, where Trump’s anti-rule of law policies will face frequent, forceful pushback in both diplomatic and dispute-resolution fora. As I have previously recounted, Trump’s first term reminded us that the rule of law is sticky and bends more easily than it breaks. Today’s global norm is not autarky, but interdependence. Domestic and international law are so intertwined in what I have called a “transnational legal process” that they create default patterns of law-observant behavior that even a willful and lawless president cannot easily discard. Transnational legal process is bigger than Trump and so are such substantive challenges as global climate change and pandemic prevention. International bureaucracies and alliances assume legality and resist repeated insults to the rule of law. Violations are costly and trigger cascades of lawbreaking. Illegalities abroad breed global blowback, which come back to haunt the lawbreaker somewhere else. So perceived lawlessness by the second Trump administration will dilute American Smart Power, and an “America First” strategy will devolve into “America Alone.”
Paradoxically, as Rebecca Ingber incisively notes, “[t]he broader distaste for international institutions [exemplified by Trump and others] is in part a fear that Americans will lose control over the rules that govern us. [But the] irony is that each of these concerns becomes more realistic the less the United States engages with the process of shaping international law and institutions.” (emphasis added) So radical disengagement will foster further frictions. Even Trump loyalists, like Marco Rubio and Elise Stefanik—the likely nominees for Secretary of State and U.S. Permanent Representative to the U.N.—will want their diplomatic counterparts to listen to America. In time, those appointees may start to push against global isolation and for engagement—as Rex Tillerson and Nikki Haley sometimes did in the first term—if only to enhance their own personal standing and influence.
Over the medium term, the main strategy for avoiding major democratic backsliding will become resilience. Litigants challenging unlawful policies are likely to win at least some early victories from federal judges. Trump’s undisciplined impulses will make it hard for his administration to maintain a durable coherent approach on any complex issue. Over time, sustained opposition to deep structural change will likely shift the focus of Trump’s short attention span to easier symbolic targets.
II. Rebuilding the Dam: Longer-Term Reform
Even while rule of law advocates are plugging the dam, they must help rebuild it. My book’s penultimate chapters addressed a range of reform proposals to restrain unilateral warmaking, and clarify international lawmaking and unmaking, intelligence oversight, information control, and protection of the democratic electoral process. This Symposium has offered constructive “friendly amendments” to a number of these proposals, which I generally accept. Let me briefly comment on three: restraining secret law (here); maintaining America’s relationship with international law (here) by clarifying constitutional restraints on unilateral unmaking of treaties and international agreements (here and here); and promoting better readings of presidential power (here) by avoiding “Youngstown category manipulation.”
Unveiling Secret Law. Even a new Congress that backs the president should know the sources of the president’s legal advice. Dakota Rudesill incisively highlights the urgency for Congress to impose greater transparency regarding secret law—nonpublic law that exists in all three branches—which can give legal cover for overbroad and unaccountable uses of law.
Trump’s threat to use the Vacancies Act and recess appointments raises red flags that many senior executive lawyers will be put in place without proper vetting or assurances that they will follow Congress’s laws. After 9/11, classified OLC memos, including the notoriously wrong “torture memo,” reinterpreted laws without running through normal internal or inter-agency processes. Trump’s second term may similarly seek to invoke secret law to justify executive orders directing mass deportations, domestic deployment of the military, and dismantling key components of the administrative state. To combat this—and future abuses of power under either party—Congress should, as both Rudesill and I urge, enact the DOJ OLC Transparency Act to reduce the executive’s ability to withhold legal memos used to depart from public understanding of the law. Requiring legal transparency should be a constitutional responsibility, not a partisan issue. Republicans have a strong incentive to hold the next Democratic administration accountable under the same transparency requirements. As Rudesill argues, reporting all OLC secret law to Congress could build momentum to fix other areas of insufficient transparency.
Unmaking Treaties and International Agreements. My initial Symposium post asked whether “Donald Trump [could] by tweet unilaterally withdraw the United States from every treaty, agreement, and international institution to which the United States is a party?” As a matter of constitutional law, the legal question is whether the president has an inherent unilateral power to terminate any and all treaties and executive agreements, or whether the Constitution requires a more fact-sensitive rule requiring more congressional participation to terminate different kinds of agreements, depending on their subject matter? As I have detailed elsewhere, neither the Constitution’s text nor structure answer this question definitively. While recent historical practice has seen unilateral executive withdrawals, very few of those have been contested by Congress. INS v. Chadha (1983) graphically showed that the Court may declare unconstitutional even an institutional practice (e.g., legislative vetoes) in which a coordinate branch has acquiesced for many decades.
The only Supreme Court case on point, Goldwater v Carter (1979), found nonjusticiable President Jimmy Carter’s termination of a particular bilateral mutual defense treaty, in accordance with its terms. But tellingly, none of the nine justices in Goldwater endorsed a unilateral one-size-fits-all transsubstantive power of presidential treaty termination. To the contrary, the proposition on the merits that garnered four votes was “that different termination procedures may be appropriate for different treaties” (William Rehnquist J. concurrence at 1003) (emphasis added). A fifth justice, Justice William Brennan, also voted for the president’s power to terminate the treaty at issue, not because the president had a general inherent power of unilateral treaty termination, but because in that particular context, the subject matter at issue (recognition) empowered him to do so. For that reason, my book argued for a more fact-sensitive, context-specific constitutional “mirror principle,” whereby a comparable degree of congressional input should be required for agreement termination as for entry (311-315).
The Symposium pieces on this topic—by Professors Sean Murphy and Ed Swaine and Catherine Amirfar and Ashika Singh—both agree that the United States needs a more nuanced constitutional rule to govern withdrawal from international agreements.
Amirfar and Singh further agree that because presidential powers exist on a continuum, the constitutional rule in this area must consider the degree of congressional approval and the constitutional allocation of institutional authority over the subject matter. They generally endorse the “mirror principle” with complementary “friendly modifications” that they refer to as a “tailored out” approach, that would build in greater flexibility to address situations where the mirror principle might not be possible to apply. Their “tailored out” approach would require nuanced assessment of the text and subject matter of particular agreements—for example, international trade agreements—in light of the tripartite framework of congressional approval suggested in Justice Robert Jackson’s landmark concurrence in Youngstown Sheet & Tube Co. v. Sawyer.
In particular, Amirfar and Singh argue that unilateral executive withdrawal by Trump from the Paris Climate agreement—which was initiated by the executive with congressional awareness and approval and broadly implicates Congress’s commerce powers—would infringe upon congressional powers. They further suggest that Congress incorporate requirements for withdrawal into congressional authorizations for an agreement, and enact legislation to set out a withdrawal process from certain agreements, for example, Section 1250A of the 2023 National Defense Authorization Act (NDAA), which expressly prohibits the president from withdrawing from NATO or using any appropriated funds for that purpose without congressional permission and ensures that withdrawal be scrutinized with caution by a reviewing court.
Murphy and Swaine are sympathetic to a mirror principle “as a better aspiration for interbranch cooperation,” agreeing that we should “avoid[] a rigid rule whereby the president can always terminate agreements unilaterally.” Just because the president can unilaterally terminate a sole executive agreement made within the scope of his plenary executive power does not mean that he can terminate unilaterally the new NAFTA (USMCA) or another congressional-executive agreement that was entered only after an extensive congressional debate and vote in an area of core congressional subject-matter authority. While we disagree on precisely what the law currently requires, they recognize that Congress or the Senate might limit presidential termination when authorizing an agreement or through a later statute, and that if Trump sought to disregard such limits, his power would fall into the Youngstown III category, where his authority is at its “lowest ebb.”
But the difficult question is what to do if Congress, or the Senate, approves an agreement but is silent about any future termination procedure. Especially if a congressional-executive agreement is self-executing, and functions like a statute, it is hard to see where the president acquires the unilateral constitutional power to repeal that congressional-executive agreement.
Murphy and Swaine agree that it is “somewhat anomalous to allow the president to repeal the ‘law of the land’ when it is in the form of a treaty but not in the form of a statute.” They further acknowledge that presidential authority over appointments—where the president needs Senate confirmation to make the appointment, but may fire that official unilaterally—“is not precisely comparable to the issue of terminating international agreements.” As I have argued elsewhere, recent Supreme Court doctrine in Zivotofsky v. Clinton has reduced barriers to justiciability by narrowing the political question doctrine. So an agreement termination properly contested by Congress would likely be justiciable. During the second Trump administration, litigation should clarify whether, absent a congressional condition on exit, the constitutional default rule is that the president possesses a one-size-fits-all unilateral termination power, or whether that termination power varies, depending on how much congressional input went into entering the agreement.
This is not just a narrow constitutional question, but a policy challenge critical to the future of global governance and America’s relationship to international law and institutions in a post-Trump world. As Ingber observes, “if a reckless president withdraws the United States from critical institutions or treaties, they may face few political consequences for doing so, and the United States may never be able to rejoin.” More than 40 years ago, during the Goldwater litigation, D.C. Circuit Judge George MacKinnon foresaw the danger of “an ambitious or unreasoned President disengaging the United States from crucial bilateral and multilateral treaties with the stroke of a pen.” As Ingber notes, even if Trump cannot singlehandedly destroy international law and institutions, the United States may well emerge from his presidency lacking a coherent “strategic vision of how to engage international law and institutions going forward.” America’s post-World War II global leadership position could erode, diminishing U.S. influence and capacity to negotiate agreements, promote accountability, and engage in international dispute-resolution for generations to come.
Reducing Youngstown Category Manipulation. Deeks and Eichensehr usefully highlight what they call “Youngstown Category 1 Complacency”: an interpretive move whereby courts too easily conclude that spotty historical precedent, ambiguous congressional silence, and perceived “congressional shoves” push unilateral presidential action up into Youngstown Category 1, where the president’s power is deemed to be at its constitutional maximum. As my book chronicles (255-56), the risk that the courts will expand “Youngstown category manipulation” is one of the greatest dangers of the Trump 2.0 era.
Beyond the pages of this Symposium, Professor Jack Goldsmith chides me for
treat[ing] presidential actions pursuant to broad but clear congressional authorizations as examples of presidential unilateralism that defy the Youngstown vision. … Koh decries presidential actions pursuant to this broad but clear delegation [in 8 USC § 1182(f), as construed in Trump v. Hawaii] as a “statutory Curtiss Wright” approach to immigration—a strange formulation, since Congress clearly authorized broad presidential discretion.
But Goldsmith misunderstands the point: the very question is whether the congressional delegation to the president is in fact “broad but clear.” My new book argues that the courts have increasingly applied Justice George Sutherland’s infamous opinion in U.S. v. Curtiss-Wright Export Corporation to distort Jackson’s Youngstown analysis at both a constitutional and a statutory level.
First, Curtiss-Wright’s dicta positing that the president may exercise inherent constitutional authority as “sole organ” of the nation in foreign affairs has been read to empower the courts and executive branch lawyers to apply a “Youngstown Category Three” theory (256). Under that theory, a court may invalidate even a legislative enactment in direct opposition to presidential will—although in Category III the president’s power should be at its “lowest ebb”—claiming that the statute unconstitutionally invades the president’s exclusive constitutional “sole organ” power. As I point out with respect to the Haitian refugee case (in which I represented the respondents), the first major Supreme Court test of the statute Goldsmith cites, 8 USC § 1182(f),
the president’s direct return of the refugees should have fallen into the third category of Jackson’s Youngstown concurrence, where the executive’s ‘power is at its lowest ebb’ because the president is acting in a manner ‘incompatible with the express or implied will of Congress’ that [as Justice Blackmun put it in his dissent, 509 U.S. at 190] ‘[v]ulnerable refugees shall not be returned.’ But the majority simply rejected this Youngstown claim, predictably citing in dictum Curtiss-Wright. (141)
But a second problem with Sutherland’s opinion is its “statutory Curtiss-Wright” theory of delegation (256), which emphasizes “the unwisdom of requiring Congress in this field of governmental power to lay down narrowly definite standards by which the President is to be governed.” (299 U.S. at 321-22). If applied as an overly deferential canon of statutory construction, this language would enable courts—notwithstanding individual rights, legislative intent, or legislative history to the contrary—to misread statutory authorizations to place presidential actions into Youngstown Category I that properly belong in Categories II or III. Citing this Curtiss-Wright language, a court may conclude that Congress has conferred a greater degree of discretion on the president through a foreign affairs-related statute, because the delegated statutory authority overlaps with or complements the president’s own constitutional foreign affairs powers. But as noted in my book, this judicial tendency
encourage[s] the president to act first then search for preexisting congressional blank checks, rather than seek specific prior or immediately subsequent legislative approval of controversial decisions. … [This] has encouraged executive-branch lawyers to play ‘find the statute,’ that is, to search the U.S. Code for preexisting statutes that they can claim already directly or implicitly authorized the challenged action…allow[ing] the president to avoid asking the legislature for new legislative authority, which might be withheld. (115)
To prevent Youngstown Category manipulation in the Trump 2.0 era, vigilant courts should follow Justice William Douglas’s directive in Kent v. Dulles (1958), declaring that judges must find a clear statutory statement that Congress has authorized the executive act in question before deciding whether to condone executive actions in foreign affairs that infringe upon individual constitutional rights. As Douglas put it in Ex Parte Endo,(1944), applying a “clear statement principle,” rather than a “statutory Curtiss-Wright principle” of statutory construction, requires that judges “must assume, when asked to find implied powers in a grant of legislative or executive authority, that the law makers intended to place no greater restraint on the citizen than was clearly and unmistakably indicated by the language used.”
This concern about Trump 2.0 claiming “broad but clear delegations” from obsolete statutes extends to warmaking as well. As I asked in my initial Symposium post: “Would current law allow [the president] unilaterally to back into a wider war in the Middle East, out of a desire to help Israel fight Hamas, Hezbollah, Iran, and Iran-backed militias in the Red Sea?” If Trump seeks to enter a wider war in the Middle East, he will surely claim delegated power to do so, citing Authorizations for the Use of Military Force that are more than a quarter-century old. But as I have repeatedly argued in urging steps to end Forever War (e.g., here and here), Congress should require Trump to seek new legislative authorization for any new U.S. military actions, rather than claiming that those actions are authorized by prior statutes that were plainly meant for earlier conflicts. Goldsmith claims that I have changed my views of war powers since 1990, now favoring “a vision of unilateral presidential war powers that defies the balanced institutional participation that his book otherwise exalts,” because I question the efficacy of the War Powers Resolution. But as one of my research assistants for the original version of The National Security Constitution, Goldsmith may recall that I expressed similar hesitations about the War Powers Resolution back then (x, 38-40, 174, 189-93). And as I note at the conclusion of my new book, “during the four decades since I first entered the government in the 1980s, my legal views and guiding principles have remained largely unchanged. … I have offered detailed legal defenses and have read no persuasive rebuttals of my written opinions as a government lawyer, which I continue to believe are correct” (338-9).
To be sure, recent Congresses have not pushed back against the Office of Legal Counsel’s view “that Article II authorizes the president to use force without congressional authorization for “limited interventions of constrained nature, scope, and duration,” so long as they do not rise to the level of “war” that Congress must constitutionally declare. But once again, Congress is certainly free to do so if president Trump abuses that authority, as the executive branch did in the 1980s in Chadha, when it successfully challenged the constitutionality of the legislative veto after acquiescing in that device for many decades.
A Problem Bigger than Trump. As my book chronicles, the Founders sought above all to avoid installing a new American king (16-27). If Trump’s re-election has any silver lining, it may be as a wake-up call saying “America, we have a problem”: the threat that autocracy will return because of overconcentration of government power in the executive. I am grateful to those who have commented on my book for grappling so thoughtfully with that issue. If the problem is structural, the answer cannot simply be surviving Trump 2.0, however stressful that may be.
Globally, the problem of autocracy is on the rise. We see autocrats ruling in Belarus, China, Russia, Hungary, Iran, Myanmar, Nicaragua, South Korea, Sudan, Venezuela, and rising even in such traditional democratic strongholds as France, Italy, and Germany. One of many reasons to look to foreign constitutional experience (291-95) is to see how other embattled democracies have responded when their autocratic leaders try to steal their democracy.
Once again, the watchwords of rule of law advocates should be vigilance and rapid response. As a Korean-American, I was horrified when a Korean president with rapidly declining approval ratings invoked martial law to try to steal democracy. The Korean people heroically rose up, announcing by action in just a few hours, that “this is our democracy. It is not yours to steal.” But even in my relief, I wondered, “If something like that should happen here, would Americans respond so ferociously and so fast?”
I harbor no illusions that the reforms proposed by this Symposium will come quickly, but our alternatives, quite simply, are acceptance, apathy, despair, or reform: now or later. Beginning a decades-long national security reform process is not just necessary, but urgently demanded to break inertia and launch a new dynamic of reform. Yes, we live in a time of intense polarization. But as the 21st century unfolds, I refuse to believe that there will not come a time when we can return to a shared national commitment: not to “America First,” but rather, to being “Americans First.” (339)
IMAGE Tetra Images via Getty Images.
The post National Security Resilience and Reform: Trump 2.0 and Beyond appeared first on Just Security.
