Presidential Unilateralism Is Bad. But Not for War Powers.

In 1990, just before the Cold War’s sudden end, a young Yale Law School professor who had previously served as an attorney-adviser in the Justice Department’s Office of Legal Counsel (OLC), Harold Koh, published “The National Security Constitution: Sharing Power After the Iran-Contra Affair.” (I was a research assistant for that book.) Over the subsequent 34 years, Koh has written a great deal of scholarship; served as assistant secretary of state for democracy, human rights and labor (in the Clinton administration) and as the State Department legal adviser (in the Obama administration); sued the U.S. government concerning various foreign relations matters; and represented the United States and other countries before international tribunals.

In light of these rich and varied experiences, Koh has now updated his original book in “The National Security Constitution in the 21st Century.” Koh says the “intervening time has given me different perspectives on some of the issues addressed in the original book,” but “the core of my thinking remains the same.” This review highlights some of what is different and some of what is the same, especially with regard to Koh’s views on presidential war power.

“The National Security Constitution” 1.0

The 1990 book centered on the then-notorious but now-largely-forgotten “Iran-contra affair.” This was the scandal in President Reagan’s second term that involved illegal arms sales to Iran to secure the release of American hostages, and the use of the proceeds from the arms sale to illegally fund Nicaraguan rebels known as the contras. Koh used this episode to distinguish between “the Curtiss-Wright vision of unrestrained executive discretion” and Justice Robert Jackson’s Youngstown vision of a “policy-making process in which the three branches of government all play integral roles.”

Curtiss-Wright, as most Lawfare readers will know, is the 1936 Supreme Court decision that upheld the constitutionality of a statute that criminalized certain arms sales based on a presidential finding that an arms sale ban would promote peace in a border war between Bolivia and Paraguay. The decision contained influential dicta about the “exclusive power of the President as the sole organ of the federal government in the field of international relations,” which required that he have “a degree of discretion and freedom from statutory restriction.” And Youngstown is the 1952 decision that rejected President Truman’s effort to seize steel mills during the Korean War. Justice Jackson’s influential concurrence established three categories of presidential powers, which fluctuate depending on congressional support, silence, or restriction. Jackson reasoned that Truman’s seizure was illegal because he had acted contrary to statutory limitations and his power was not exclusive.

“The National Security Constitution” 1.0 argued for Jackson’s vision of “balanced institutional participation” by all three branches of government as opposed to the unilateralist vision supposedly contemplated by Curtiss-Wright. The book posed the question “why the president almost always wins in foreign affairs.” Koh’s insightful answer came in three parts: (a) The president has institutional incentives to take the initiative in foreign affairs and the power to construe laws in ways that allow him to do so; (b) Congress has institutional incentives to acquiesce to presidential initiative; and (c) federal courts, mainly by use of justiciability doctrines, have tolerated executive aggrandizement. This arrangement, Koh argued, violated “the Constitution’s mandate of balanced institutional participation in national security decision making.” To rectify this state of affairs, Koh called for “a new national security charter,” a framework statute that would reorient power and incentives among the three branches to restore balanced institutional decision-making.           

“The National Security Constitution” 2.0

The 2024 version of “The National Security Constitution” maintains the same basic framework: an orienting distinction between the “unilateralist vision” of Curtiss-Wright and “Youngstown’s vision of shared powers,” and the same basic three-part explanation for why the president almost always wins in foreign affairs. The Iran-contra affair recedes in importance, and Koh adds a long and opinionated history of foreign relations law since 1990. The thrust of his story is that Republican presidents—Reagan, the second President Bush, and President Trump—embraced the Curtiss-Wright vision of “unrestrained executive power,” while Democratic Presidents Clinton, Obama, and Biden “have swung the pendulum partially and sporadically back toward shared power, but new external threats and the legislative weaknesses of all three leaders forced them to employ ad hoc unilateralist tactics as well.” Koh now has a much longer list of needed reforms.

In the 2024 version of the book, the distinction between Curtiss-Wright unilateralism and Youngstown’s balanced institutional participation is in some contexts analytically unhelpful. Koh sometimes associates the Curtiss-Wright vision with exclusive presidential power that denies the need for “congressional concurrence” for most foreign relations decisions. But other times he associates it with something very different: a “diluted role” for Congress or a lack of “meaningful participation” by Congress (emphasis added). This equivocation elides a central challenge of U.S. foreign relations law: Congress has authorized by statute a great deal of presidential action, often in broad terms, that presidents often rely on—sometimes faithfully, sometimes opportunistically—when exercising power. An analogous issue arises under the Youngstown vision. Jackson blessed presidential acts pursuant to “an express or implied authorization of Congress” (emphasis added), but the game ever since—in the executive branch, and in courts—has come in figuring out whether Congress has in fact authorized the presidential action in question. It all turns on the breadth and clarity of the congressional authorization and thus one’s method of statutory interpretation in particular cases.

The difficulty becomes apparent when Koh treats presidential actions pursuant to broad but clear congressional authorizations as examples of presidential unilateralism that defy the Youngstown vision. For example, Koh notes that Congress in the immigration context—most notably, 8 U.S.C. § 1182(f), the main statute at issue in Trump v. Hawaii, the travel ban case—has “de facto delegated enormous discretion to the executive branch through the ‘one size fits all’ penalty of removal combined with a broad delegation of authority to the president to determine how the laws should be ‘faithfully executed.’” Koh decries presidential actions pursuant to this broad but clear delegation as a “statutory Curtiss Wright” approach to immigration—a strange formulation, since Congress clearly authorized broad presidential discretion. Yet absent some robust constitutional nondelegation constraint, which Koh does not advance (he does decry a foreign affairs exception to the nondelegation doctrine, and judicial deference doctrines tailor-made to foreign affairs), one wonders why this isn’t a clear example of the Jackson Youngstown vision since Congress so clearly delegated authority over the issue in broad terms to the president.

The puzzle grows when Koh moves to trade policy. He complains about President Trump’s wielding “broad delegated statutory authority to engage in unilateralist acts” that Koh thinks made for bad trade policy. Why were those acts unilateralist if they rested on clearly delegated authority? The “unilateralist” criticism of presidential action loses its meaning if it includes not just presidential action based on Article II independent or exclusive power (presidential action without statutory authorization) but also presidential action pursuant to broad statutory authorization.

To be sure, it is often the case—though not in the immigration and trade examples Koh mentions—that presidents opportunistically interpret statutes to “permit executive initiatives extending far beyond the intended scope of those statutes,” as Koh puts it. This is a major weakness in the Jackson Youngstown framework, especially in those contexts when presidents act in ways not subject to judicial review. A prominent example is President Obama’s 2014 interpretation of the 2001 Authorization for Use of Military Force (2001 AUMF) to apply to the Islamic State. This interpretation was at best a very big stretch that significantly broadened (in terms of the enemy, geographic scope, and time) the scale of the war that began on 9/11. But even though Koh decries the “forever war” and says it should be ended, he mentions this vital episode in a single passing sentence, without criticism, in a paragraph about how Obama embraced a “narrowed legal theory” of the scope of the post 9-11 conflict compared to the Bush era. And in a footnote he cites a blog post he wrote (after returning to Yale Law School from the Obama administration) in support of the legality of the extension of the 2001 AUMF to the Islamic State—an event of much greater defiance of the Jackson Youngstown vision than presidential acts pursuant to clearly delegated congressional authority in the immigration and trade contexts.

Koh on War, 1.0 v. 2.0

The 1990 book was published before the Cold War ended and before the first U.S. invasion of Iraq. The Vietnam War is mentioned prominently, but Koh’s main point related to presidential war powers in that conflict is that President Lyndon Johnson construed the 1964 Gulf of Tonkin Resolution “as broad congressional authorization to escalate” the war. Koh devoted his war powers analysis mainly to small and short affairs: President Ford’s Mayaguez rescue; Reagan’s invasion of Grenada; strikes against Libya and Iranian oil platforms; the dispatch of troops to Lebanon (which was not short); and President George H.W. Bush’s sending of troops to El Salvador, the Philippines, and Panama.

Koh said little in 1990 about the president’s Article II power to engage in these actions without congressional authorization beyond noting that the War Powers Resolution (WPR) “derogate[s] from Congress’s war-making powers by allowing the president too much freedom to make covert and short-term war and to commit military forces overseas without a clear purpose.” He focused instead on how presidents had circumvented the WPR—by using the 60-day clock as permission for short uses of force; by relying more extensively on covert operations, which the WPR does not regulate; and by finding loopholes to send U.S. troops into harm’s way beyond the 60-day limit (as in Lebanon). Koh proposed amending the WPR to expressly authorize the president to engage in some short-term military actions (such as rescuing American citizens); to require more robust presidential consultation with Congress; to enhance sanctions and penalties for noncompliance with the WPR; and to authorize members of Congress to sue the executive branch for noncompliance with the WPR.

War has played a significantly larger role in U.S. foreign policy in the intervening 34 years: two large-scale military actions against Iraq (both authorized by Congress); the now-23-year-old sprawling multicountry and ever-morphing “forever war” that began on 9/11, that has cost more than $8 trillion, and that is ostensibly authorized mainly by the 2001 AUMF; numerous uses of force without congressional authorization, including notable humanitarian interventions such as Kosovo in 1999 and Libya in 2011 most prominently and also Bosnia (1995), Haiti (1994), and Somalia (1992); and one-off strikes ostensibly in “self-defense,” including Clinton’s 1998 cruise missile strikes against terrorist targets in Sudan and Afghanistan in response to the Africa embassy bombings and Trump’s targeted killing of Iranian Gen. Qassem Soleimani in January 2020 (which the administration also justified under the 2002 Iraq AUMF). “The National Security Constitution” 2.0 is a shocking reminder about how commonplace and extensive war has become as a tool of U.S. foreign policy since 1990.

Perhaps because there has been so much more war since 1990, and because Koh was involved in war-related decision-making while serving in the Clinton and Obama administrations, Koh says more in the 2024 book about the president’s unilateral Article II power to use force without congressional authorization. He adopts the Office of Legal Counsel view—which he had embraced in prior scholarship—that Article II authorizes the president to use force without congressional authorization for “limited interventions of constrained nature, scope, and duration.”

This view of the president’s Article II war powers is a significant step beyond where Koh seemed to be in 1990. And it is a consequential counterpoint, in the most consequential of contexts (war), to any theory of balanced institutional participation in foreign affairs. Koh invokes the OLC theory mainly in the context of justifying the domestic legal basis for two “limited” humanitarian interventions that he strongly favors: Clinton’s 1999 use of force without congressional authorization in Kosovo, which involved over 30,000 U.S. sorties during a 78-day air campaign; and Obama’s 2011 use of force without congressional authorization in Libya, which involved thousands of sorties over 227 days. Importantly, the same OLC theory operates in nonhumanitarian intervention contexts and has been especially important in modern “light footprint” warfare involving drone strikes, cyber and missile attacks, and special operations. These unilateral presidential uses of force—which Koh acknowledges are the dominant forms of U.S. modern warfare—will almost always have a “constrained nature, scope, and duration” within the meaning of the OLC opinions.

Koh pushes back on many executive branch legal positions in the book but not on the OLC view of the president’s Article II war powers. His only affirmative argument for the OLC view, as best I can tell, is that “Congress has ... largely acquiesced in [the OLC] interpretation.” Koh’s theory of acquiescence in “The National Security Constitution” 2.0 is hard to pin down. Sometimes he invokes “the high standards for ‘constitutional acquiescence’ set by Justice Frankfurter’s Youngstown opinion,” under which Koh says “[e]ven four decades of consistent executive practice would not rise to Justice Frankfurter’s level of historical ‘gloss’ unless the other affected branch of government affirmatively ‘acquiesced’” (emphasis added). He does not appear to apply this demanding standard, however, in the constitutional war powers context. In a book that emphasizes the “constitutional primacy of balanced institutional participation,” it is surprising that Koh accepts Article II war unilateralism so casually.

The WPR in theory constitutes a statutory constraint on presidential uses of force under Article II. Koh the scholar defends at length his support in government of the Obama administration’s arguments that the 2011 conflict in Libya fell short of “hostilities” and thus did not implicate the termination provisions of Section 5(b) of the WPR. This issue has been thoroughly litigated, and I will not repeat the arguments pro and con. What is significant now is that Koh thinks that the WPR’s current structure is “increasingly obsolete” in an era of light footprint warfare. Yet his proposed WPR reforms would do little to check the broad conception of the president’s unilateral Article II war powers he embraces.

Koh says Congress should establish a “core consultative group” with continuing responsibility for war powers problems that the president must “consult regularly” but without binding consequence. He thinks Congress should “pre-authoriz[e]” certain uses of force such as rescuing American citizens abroad. It should also create “specialized legislative procedures” that would require Congress to vote on a proposed use of force. For example, he says that “Congress should adopt a stricter framework to regulate and oversee the use of force in Iran, Somalia, Syria, and Yemen and against emerging terrorist groups.” But he does not specify that framework beyond describing a process for a fast-track vote that only works if Congress musters the will (which, as Koh explains, it usually lacks) or if the president introduces a bill seeking such approval (which presidents rarely do). Koh says with hope that Congress would “likely” support relaxation of the WPR’s “strict durational limits for the limited purpose of completing” future humanitarian missions (I am skeptical) and adds that Congress’s main concern should be on curbing “open-ended” (an undefined term) military missions. He adds that a reformed WPR should “grant individual members of Congress statutory standing” to seek relief under the WPR in federal court, but this would accomplish little under the current law of justiciability. Koh also thinks that WPR reform should “consider” regulation of “targeted killing conducted via drone, artificial intelligence, or cyberspace,” perhaps (Koh does not take a firm view) with ex ante or ex post judicial review. And he believes that Congress should wind down the forever war.

It is hard to assess these proposals, in part because they are (as Koh acknowledges) politically unrealistic, and in part because they are in crucial places so vague. But as written, they would not meaningfully restrain the dominant forms of presidential warfare, including light footprint warfare, that his conception of Article II war powers permits.

***

It is not surprising that some of Koh’s views have evolved over the course of his career. Given his many experiences in the arena, and massive changes in the world, it would be surprising if they had not evolved. As Koh says, quoting Justice Harry Blackmun, “Life is long, and you learn things.” Nor is it surprising, or troublesome, that Koh sometimes took different positions in government than he took as an academic, or that, as he says, “some presidents based unilateral actions in part on advice that I gave as a government lawyer.” The role of the government lawyer is entirely different from that of the legal academic. As Koh says, quoting former OLC head and former law professor Walter Dellinger: “[T]here are powerful and legitimate institutional reasons why one’s views might properly differ when one sits in a different place.” What is surprising, and to me puzzling, is that in a capstone academic book about the dangers of presidential unilateralism in the conduct of U.S. foreign relations, Koh articulates a vision of unilateral presidential war powers that defies the balanced institutional participation that his book otherwise exalts.