The Yale Law Journal


The Separation of National Security Powers: Lessons from the Second Congress

15 Feb 2020

abstract. Whether in the context of President Trump’s controversial border wall or other modern assertions of crisis authority by the executive branch, contemporary discussions of the separation of powers tend to pay far too little attention to the Second Congress’s nuanced, iterative, inter-branch approach to national security emergencies—in particular, the debates leading up to, and enactment of, the Calling Forth Act of 1792. As this Essay explains, that statute, which implemented Congress’s power to provide for the calling forth of the militia to respond to domestic emergencies, provides not only powerful evidence of the role that the Constitution’s contemporaries believed each branch could play in those moments, but also a potential roadmap for today’s Congress to reclaim a meaningful role in overseeing the broad powers it has increasingly delegated to the executive branch over time.


President Trump’s declaration of a “national emergency” for the sole purpose of using otherwise-appropriated military construction funds to build a wall along the U.S.-Mexico border1 provoked an understandable storm of controversy and criticism.2 It also belatedly focused meaningful public attention on a problem that national security scholars have been warning against for years: Congress’s systematic over-delegation of authority to the President to respond to a surprisingly broad array of real or invented (or, at least, overblown) crises.3 From the National Emergencies Act (NEA)4 to the Insurrection Act;5 from the Trade Expansion Act of 19626 to the International Emergency Economic Powers Act;7 and from the 2001 Authorization for Use of Military Force (AUMF)8 to the 2002 AUMF in Iraq,9 federal law today delegates a staggering amount of power to the President in these conditions—including, as in the NEA, the power to decide whether these conditions are even present.10

Thanks to the Supreme Court’s 1983 decision in INS v. Chadha,11 it has become practically impossible for Congress to claw back these authorities once they have been delegated. Although the NEA, like the War Powers Resolution, included a legislative veto to allow Congress to unilaterally terminate presidential invocations of the delegated authorities,12 Chadha held that such legislative vetoes are unconstitutional because they fail to comply with Article I, Section 7’s bicameralism and presentment requirements.13 As Congress memorialized in a 1985 amendment to the NEA, in practice that means that the legislature can only override a President’s actions under those statutes by passing a new statute14—which inevitably requires veto-proof supermajorities in both chambers.15 Given how closely divided Congress has been in recent years, a veto override today effectively requires significant support from the President’s own party. But with partisan allegiances in Washington increasingly crowding out institutional concerns—and with the separation of powers increasingly giving way to the separation of parties16Chadha effectively forecloses congressional overrides absent a President who is deeply unpopular even within his own party.17

Instead, the principal checks on abuses of these emergency authorities over time have been political, as successive Presidents have been reluctant to stretch these statutes to their limits lest they suffer adverse political consequences. These political checks, however, have proven ineffective on the Trump Administration, which has been willing to push those authorizing statutes to, if not beyond, their limits.18 All of this should provoke—and, at least in some circles, has provoked—a discussion about what a new generation of reforms to statutes delegating national security powers should look like.19

As this Essay explains, that conversation would do well to learn from the Second Congress, which met from 1791 to 1793.20 Not only did the Second Congress carefully grapple with some of the same questions about delegating national security powers to the President (at a time when Congress was out of session for much of the year), but some of the checks and balances it wrote into federal law could provide a model for potential reforms today. To that end, Part I introduces the Second Congress’s approach to delegated national security powers, with a special focus on the Calling Forth Act of 1792, and Congress’s changes to that statute after the Whiskey Rebellion. Part II identifies three lessons that current debates should take away from this important, early experience with the separation of emergency powers. First, this history is a powerful reminder of the alarming breadth of one of the most quietly important contemporary delegations of emergency power—the Insurrection Act, which authorizes the President to call out the federal military to respond to a broad (and poorly defined) array of domestic crises. Second, the 1792 statute included both ex ante and ex post checks on the President that could provide equally effective constraints on the use of broad emergency powers by modern presidents. Third, the debates over those checks provide strong evidence that these checks were not viewed at the Founding as raising serious constitutional concerns. If anything, they were seen as necessary to avoid an even graver concern—that Presidents would abuse national security powers for purely pretextual reasons, with obvious implications for the liberty of the people. Those lessons should resonate increasingly loudly to contemporary ears.

I. the second congress and the militia

The First Congress tends to receive the lion’s share of academic attention with respect to how the Constitution was originally implemented,21 but it was the Second Congress that broadly confronted the question of presidential power in the national security sphere. Although earlier statutes had provided specific authorization,22 at the tail end of its first session in May 1792, the Second Congress grappled with how to more generally implement Congress’s powers under the Militia Clauses: “[t]o provide for calling forth the Militia to execute the Laws of the Union, suppress Insurrections and repel Invasions,”23 and “[t]o provide for organizing, arming, and disciplining, the Militia, and for governing such Part of them as may be employed in the Service of the United States.”24

As to the militia, on May 2, 1792, Congress passed what has become known as the Calling Forth Act. Section 1 of the Act focused on two of the three circumstances in which the militia were to be “called forth”: to suppress insurrections and repel invasions. As it provided,

whenever the United States shall be invaded, or be in imminent danger of invasion from any foreign nation or Indian tribe, it shall be lawful for the President of the United States, to call forth such number of the militia of the state or states most convenient to the place of danger or scene of action, as he may judge necessary to repel such invasion, and to issue his orders for that purpose, to such officer or officers of the militia as he shall think proper; and in case of an insurrection in any state, against the government thereof, it shall be lawful for the President of the United States, on application of the legislature of such state, or of the executive (when the legislature cannot be convened) to call forth such number of the militia of any other state or states, as may be applied for, or as he may judge sufficient to suppress such insurrection.25

Section 2 dealt with the more controversial category of cases in which the militia was to be called forth: “to execute the laws of the union.” Indeed, the bulk of the debate over the bill in the House of Representatives focused on section 2, and the widespread concern that, unlike with respect to repelling invasions or suppressing insurrections, the President could abuse the authority to call out the militia to “execute the laws of the Union” in circumstances in which it was not strictly necessary.26 There was widespread agreement that there should be some cases in which the President had such power, but a lack of consensus as to exactly which cases. The result was language that did not provide substantive constraints on the types of laws the militia could be called forth to execute, but rather the inclusion of two significant procedural constraints on the President’s authority to use military force at home:

[W]henever the laws of the United States shall be opposed, or the execution thereof obstructed, in any state, by combinations too powerful to be suppressed by the ordinary course of judicial proceedings, or by the powers vested in the marshals by this act, the same being notified to the President of the United States, by an associate justice or the district judge, it shall be lawful for the President of the United States to call forth the militia of such state to suppress such combinations, and to cause the laws to be duly executed. And if the militia of a state, where such combinations may happen, shall refuse, or be insufficient to suppress the same, it shall be lawful for the President, if the legislature of the United States be not in session, to call forth and employ such numbers of the militia of any other state or states most convenient thereto, as may be necessary, and the use of militia, so to be called forth, may be continued, if necessary, until the expiration of thirty days after the commencement of the ensuing session.27

In other words, Congress’s delegation of power to the President did not turn on a unilateral determination that the circumstances justifying a resort to extraordinary authorities were present, but rather one made by the President in concert with an Associate Justice of the Supreme Court or a U.S. district judge (there were no freestanding circuit judges at the time). And if circumstances necessitated involving the militia from a different state, the President would have to obtain Congress’s approval. If Congress was out of session (as it often was before the Civil War), the statute allowed the President to act unilaterally, but it also imposed a sunset on this unilateral authority—so that it expired thirty days after Congress was next in session.28

These two constraints—ex antejudicial approval and an ex post sunset—both came into play two years later when President Washington utilized the 1792 statute to put down the Whiskey Rebellion. As Matthew Waxman explains, the episode began as a protest of an excise tax on liquor distilled in the United States, which “was strongly opposed by communities in western Pennsylvania and elsewhere along the frontier—communities that felt distant from the national government and whose economy depended heavily on liquor production.”29 When hundreds of protesters attacked a federal tax inspector in the summer of 1794, President Washington “refused to tolerate this defiance of federal authority.”30 Deliberately but decisively, he obtained the requisite certification from Justice James Wilson,31 issued the necessary proclamation to trigger the 1792 statute, and called forth nearly 13,000 militiamen to (successfully) quell the uprising.32 And when the Third Congress (which had been in recess since June 3) reconvened on November 3, its very first legislative action was to authorize President Washington’s continued use of militias from outside of Pennsylvania for an additional three months—overriding the thirty-day sunset in the 1792 Act.33 There is no record in the House that the extension provoked significant debate (the Senate’s deliberations at this time were secret).

If Congress had stopped there, the precedents set by the 1792 Act might be more familiar to us. But, perhaps taking the wrong lessons away from the Whiskey Rebellion, and motivated by President Washington’s report on the crisis,34 Congress in 1795 dramatically weakened the checks that had been essential to the passage of the 1792 statute. First, it removed the requirement of an antecedent court order—“leaving the President as the sole arbiter of when circumstances necessitated calling forth the militia.”35 Second, it eliminated the ban on calling forth militia when Congress was in session from states other than those in which the crisis was unfolding.36 And even the more modest procedural requirement in the 1792 statute—that a proclamation to insurgents to disperse be issued before calling forth the militia—was watered down. Under the 1795 statute, the proclamation could be contemporaneous37—that is, rather than a formal prerequisite to calling out the militia, now, it was little more than “notice to the rebels that the troops were on their way.”38 Thus, the meaningful separation-of-powers precedent Congress set in the 1792 statute quickly faded from view—to be obliterated entirely when Congress, in 1807 (and without any apparent debate),39 authorized the use of federal regulars—the standing army and navy—in most of the circumstances in which the militia could be called forth under the 1795 law.40

II. three lessons from the second congress

This history bears on current debates in at least three respects. First, and most directly, the less constraining 1795 Act, as amended in 1807, is the core of what is known today as the Insurrection Act. The Insurrection Act created the President’s remarkably broad authority under current law to use the federal military (and not just the constitutional “militia”)41 to respond to a wide array of domestic emergencies.42 The Insurrection Act has not been used since President George H.W. Bush sent 2500 Army soldiers and 1500 Marines to Los Angeles to help put down the Rodney King riots in 1992.43 But multiple media reports have suggested that the Trump Administration has been contemplating invocation of the Act to enlist the military in immigration enforcement,44 a move that would be a radical and aggressive, but arguably lawful, application of the Act.45

Second, as a policy matter, there is much to commend in the ex ante and ex post checks built into the 1792 statute—judicial certification and a sunset. Taking sunsets first, especially if legislative vetoes are unconstitutional, sunsetting delegated authority within a fixed (and short) temporal window may be the most realistic mechanism for Congress to supervise how these delegated authorities are exercised. That way, “the default is for the President’s authority to lapse in the absence of congressional action, rather than to continue in perpetuity.”46 And although critics of sunsets, especially in the context of national security policies, argue that it is burdensome and unrealistic to expect Congress to repeatedly reauthorize the actions it supports,47 that is exactly what has happened with some of our most important foreign intelligence surveillance authorities, which have periodic sunsets built into them.48

To be sure, the surveillance context suggests that a sunset is no guarantee that, as the date approaches, Congress will vigorously debate whether the relevant authorities should be reauthorized (to say nothing of why). But “they at least put the onus on Congress to endorse the president’s resort to such measures, rather than reject them.”49 If nothing else, such a move should at least enhance democratic responsibility—and, as such, accountability—for these powers.

The same can be said for more express involvement by federal judges—whether in a pure ex ante context (as under the 1792 Act), or at least in relatively short order after the authority is exercised—to ensure its fidelity to the relevant statutory and constitutional mandates.50 Here again, the foreign intelligence surveillance context provides a useful template for modern national security policy—where Congress, as part of the Foreign Intelligence Surveillance Act of 1978, created the Foreign Intelligence Surveillance Court and Court of Review to provide meaningful (if largely secret) accountability for the government’s exercise of its authorities under the Act. In that context, judicial review not only helps to reduce the likelihood that the government is abusing these unique powers in a manner that is inconsistent with either the statutory delegations or the Constitution, but it also provides a stronger foundation of legitimacy for national security policies that might be more controversial if they were never subjected to meaningful judicial oversight.51

Third, and perhaps most importantly, there was no suggestion during the House debates leading up to the 1792 Act, or at any point thereafter, that the checks Congress wrote into the act raised constitutional concerns. Sunsetting the President’s authority to use militia from other states was seen as necessary, not constitutionally problematic. And involving Article III judges, whether in approving emergency declarations ex ante or reviewing them after the fact, was designed to help ensure that Presidents could not act arbitrarily. Congress was not seen as investing Article III judges with an unconstitutional, extrajudicial function, or as inappropriately involving them in national security matters. The history surveyed above therefore strongly suggests that constitutional arguments against such a judicial role are not based on the original understanding of the separation of powers. And it is hard to imagine the argument that the courts’ powers have declined since the Founding. If anything, as Justice Jackson observed in Youngstown, the Constitution’s focus on using the militia for domestic emergencies, and on having Congress define the relevant criteria for doing so, sent a powerful message about the limits of executive power, not judicial power:

Such a limitation on the command power, written at a time when the militia rather than a standing army was contemplated as the military weapon of the Republic, underscores the Constitution’s policy that Congress, not the Executive, should control utilization of the war power as an instrument of domestic policy.52


Congress should “control utilization of the war power as an instrument of domestic policy.”53 And, if the lessons from the Second Congress are any indication, Congress can do so. It is not enough, though, to fight the “last war” by trying to respond to the immediate provocation for the reform (for example, by doing nothing more than amending the NEA to include a sunset).54 As noted above, we have seen similarly broad assertions of executive power under a number of other statutes delegating national security authorities to the President—from the 2001 AUMF to the Trade Expansion Act of 1962. In each of these contexts, it is worth asking the same question: should the President’s determination that “national security” or some other emergency warrants resort to these powers be effectively unreviewable and subject only to override by a veto-proof supermajority of both chambers? If the answer is “no,” then Congress should revisit many (if not most) of its delegations in the national security space, impose sunsets across the board, and, where appropriate, provide more expressly for judicial review of the President’s determination that the circumstances necessary to trigger the relevant authorities are in fact present. And even if such a non-substantive, structural reform were to prove impossible without veto-proof supermajorities of both chambers, current and future federal representatives should still keep the Second Congress in mind before delegating any additional national security authority to the President—be it this President or his successors.

Stephen I. Vladeck is the A. Dalton Cross Professor in Law, University of Texas School of Law.