Reassessing the Legislative Veto: The Statutory President, Foreign Affairs, and Congressional Workarounds

Abstract

A chief reason that the President is insufficiently constrained when exercising statutorily-delegated power, it is claimed, is the Supreme Court’s disallowance of legislative vetoes in its decision in INS v. Chadha, a claim that intensified during the Trump administration. This article challenges this account, arguing that the availability of the legislative veto was less important before Chadha to congressional-executive relations than legal scholars commonly assume, and that, to the extent that the legislative veto was (or would have become) important for checking some exercises of statutorily-delegated authority, Congress has developed a host of effective workarounds in the years since Chadha.

1. INTRODUCTION

There have long been complaints about the growth of presidential power, a phenomenon memorably referred to by Arthur Schlesinger (1973) as the rise of the “imperial presidency.” These complaints sometimes suggest that the problem is principally one of executive aggrandizement: presidents act unilaterally and Congress often fails to respond, and over time there is an accretion of power flowing to the White House. But this account describes, at most, only part of the story. Many of the broad and controversial actions that presidents take are based on statutory delegations. These actions therefore fall into the first category of Justice Jackson’s canonical framework from Youngstown Sheet & Tube Co. v. Sawyer, where the President’s power is “at its maximum, for it includes all that he possesses in his own right plus all that Congress can delegate” (343 U.S. 579, 635 (Jackson, J., concurring)).¹ During the Trump administration, for example, the “travel ban,” the re-imposition of sanctions against Iran, the shifting of funds to construct the southern border wall, and many other controversial actions, were all based, at least principally, on statutory delegations rather than on claims of independent presidential authority.² Many of President Biden’s initial actions in office, including his efforts to undo some of his predecessor’s actions, have similarly been based on statutes rather than the Constitution. The modern President is, in large part, a “statutory President” (Stack 2005; Stein 2018; Kovacs 2020).

Of course, the connection between the growth of presidential power and statutory delegations has not gone completely unnoticed. Legal scholars who have focused on this connection have emphasized how the Supreme Court, in its 1983 decision in INS v. Chadha, eliminated a key congressional check on presidential uses of delegated power—legislative vetoes. Legislative vetoes allow one or both houses of Congress, or congressional committees, to override particular executive branch actions carried out pursuant to statutory delegations. Without the option of a legislative veto, it is argued, Congress is faced with the choice of either delegating less discretion to the executive or having to rely on new statutes to overturn exercises of the discretion, potentially by having to override a presidential veto. As a practical matter, it is said, Congress cannot choose the former. The result, therefore, is that Congress has lost much of its ability to control executive action.

The loss of the legislative veto is thought to be particularly significant in the area of foreign affairs. The growth of presidential power has been especially robust in this area, in part because the courts tend to play less of a role there than they do in domestic affairs. Recognizing the functional advantages that the executive branch has over the other branches in managing foreign affairs and the relative lack of judicial expertise in this area, the courts often either decline to hear challenges to presidential foreign affairs actions or give substantial deference to the executive branch when they do hear them.³ Without significant judicial checks, it is thought that much of the constraint on executive power relating to foreign affairs must come from Congress. Yet the concern is that, if Congress is required to enact a statute in order to limit executive action, which potentially requires obtaining a two-thirds vote in both the House and the Senate to override a presidential veto, it will have only a minimal ability to provide a check on presidential foreign affairs actions.

Developments during the Trump administration appeared to confirm this critique of Chadha. In rare displays of bipartisanship, congressional majorities expressed opposition to a number of the Trump administration’s foreign affairs actions, including its support for the Saudi-led war in Yemen, its declaration of a national emergency to justify building a southern border wall, and its targeted killing of an Iranian general. But each time Congress lacked the votes to overcome a presidential veto. Chadha was identified by many as a chief culprit in the failure of Congress to provide a check on these and other exercises of presidential power. For example, commenting on Trump’s declaration of an emergency to justify building a southern border wall, Richard Pildes (2019) observed that Chadha “decimated the policy scheme Congress had created for overseeing the President’s declaration of emergency powers—and nearly 200 other federal statutes in which Congress since the 1930s had created a similar scheme.”

This article challenges this critique of Chadha. Drawing in part on studies of the legislative veto by political scientists, this article argues that the availability of the veto was less important before Chadha to congressional-executive relations than is commonly assumed. It also argues that, to the extent that the legislative veto was important (or would have become important) for checking some exercises of statutorily-delegated authority, Congress has developed a host of formal and informal workarounds in the years since Chadha, and there are numerous additional steps that Congress could take to constrain the statutory President other than using legislative vetoes. Critics of Chadha have been unduly focused on statutory overrides of presidential action, neglecting the many other ways in which Congress retains control over statutory delegations. In light of this descriptive account, it is far from clear, this article contends, that Congress has lost much leverage vis-à-vis the President as a result of Chadha. The article also explains why the functional case for allowing legislative vetoes is more debatable than many critics of Chadha have acknowledged.

This is an especially good time for a reappraisal of Chadha. With a new presidential administration in place, and unified party control of Congress and the White House, there are significant calls for reforms aimed at cabining presidential authority and, relatedly, enhancing legislative authority over the executive branch.⁴ In some ways, this period may end up resembling the period in the 1970s following the Vietnam War and the Watergate scandal in which Congress was especially active in attempting to cabin executive authority, including in foreign affairs.⁵ Not coincidentally, the 1970s was also a time in which Congress substantially increased its inclusion of legislative veto provisions in statutes, including in the foreign affairs area. In considering separation-of-powers reforms today, however, it is vital to have a clear sense of what, precisely, the current problems are with respect to broad presidential authority, and what aspects of congressional-executive relations have or have not been responsible for these problems. This article aims to contribute to this effort by clearing away some common misconceptions about the legislative veto. The recent change in presidential administrations is also important, this article contends, for an additional reason: it makes it easier to consider the potential effects of separation-of-powers reforms on the operations of government over time as political alignments change, rather than just focusing on a particular administration, a problem that has plagued some of the recent criticisms of Chadha.

Section 2 of this article describes the historical use of the legislative veto mechanism, Congress’s disapproval of this mechanism in Chadha, and the claims by legal scholars about its negative impact on the separation of powers. Section 3 critiques the claim that Chadha significantly undermined Congress’s ability to check its statutory delegations, a claim that intensified during the Trump administration. As that section explains, legislative veto provisions did not play a major role in congressional-executive relations in most areas before Chadha, especially in foreign affairs, and, to the extent that they were important in select areas, Congress has developed a number of ways of working around the decision. Section 4 presents three case studies that illustrate the descriptive critique, relating to war powers, arms sales, and emergency declarations—all areas where critics of Chadha have argued that the decision has had a particularly significant impact. Section 5 explains that, although Chadha is often thought to represent a triumph of formalism over functionalism, the functional case for allowing legislative vetoes is debatable. Section 6 concludes.

2. THE RISE AND FALL OF THE LEGISLATIVE VETO

This Section briefly describes Congress’s enactment of legislative provisions from 1932 until the Chadha decision in 1983. It also describes the Supreme Court’s reasoning in Chadha and explains why the decision is generally thought to completely disallow legislative veto provisions. Finally, it describes the claim, which is frequently made by legal scholars and which intensified during the Trump administration, that Chadha substantially undermined the U.S. system of checks and balances.

2.1 The Rise of the Legislative Veto

Depending on how they are framed, legislative veto mechanisms allow either a committee of Congress, one of the two houses of Congress, or both houses acting jointly to override executive action that is taken pursuant to a statutory delegation of authority (Breyer 1984). Even when they involve both houses of Congress, legislative veto provisions do not take the form of legislation. To enact a statute, Congress must present a bill to the President for signature, and, if the President vetoes it, the bill cannot become law unless Congress overrides the veto with two-thirds votes in both houses.⁶ Two-house legislative vetoes, by contrast, take the form of “concurrent resolutions” that take effect after majority approval by the House and Senate, without being presented to the President. These concurrent resolutions are distinct from “joint resolutions,” which are statutes that require presidential approval or an override of a presidential veto.

Most commentators trace the legislative veto to 1932, during the Hoover administration, when Congress delegated to the President the authority to reorganize the executive branch, subject to disapproval within sixty days by either house of Congress (Act of June 30, 1932, ch. 314, § 407, 47 Stat. 382, 414 (1932)).⁷ Hoover had been seeking to persuade Congress to delegate reorganization authority to him, and at one point proposed the legislative veto as a compromise, although his Attorney General later concluded that it was unconstitutional (Fisher 2014, p. 159). Congress subsequently used the veto provision to disapprove of Hoover’s reorganization proposals, which he made after having lost to Franklin Roosevelt in the presidential election (76 Cong. Rec. 2125–26 (1933)). Congress then renewed the reorganization authority shortly before Roosevelt took office and eliminated the legislative veto provision. But when Congress renewed the authority again in the late 1930s, it added in a two-house veto provision (Millett & Rogers 1941). In 1949, Congress changed the veto provision back to a one-house veto, and it continued to add the provision in its renewals of reorganization authority during subsequent presidential administrations (Fisher 2014, p. 159).

During this period, Congress began including veto provisions in other types of legislation, including in legislation relating to deportation, as discussed below. In 1941, Congress included a two-house veto provision in the Lend-Lease Act, which allowed President Roosevelt to provide defense materials to Great Britain and other nations involved in the war in Europe. Under the Act, the authority delegated to the President would cease if Congress declared by concurrent resolution that it was no longer necessary (Pub. L. No. 77-11, § 3(c), 55 Stat. 31, 32 (1941)). Roosevelt signed the Act into law even though he privately contended that the veto provision was unconstitutional.⁸

Not coincidentally, this was also a period in which the power of the presidency, and the executive branch more generally, was on the rise, in both domestic and foreign affairs. The difficulties of regulating a modern economy and addressing the challenges posed by the Great Depression gave rise to a vast administrative state housed within the executive branch. In 1946, Congress enacted the Administrative Procedure Act (APA) to regulate this new element of American governance (Pub. L. No. 79-404, 60 Stat. 237 (1946)), but the Act has been construed not to apply to actions taken directly by the President (see Franklin v. Massachusetts, 505 U.S. 788, 800-01 (1992)), and it also has a variety of limitations and exceptions, including an exemption from the rulemaking requirements for military and foreign affairs functions (5 U.S.C. § 553(a)(1)). Instead of including a legislative veto provision in the APA, Congress included what would today be called a “report-and-wait” requirement, providing that most administrative regulations had to be published at least thirty days before their effective date, which in theory gave Congress the opportunity to block them by statute (5 U.S.C. § 553(d)).

Also during this period, the United States was addressing the challenges posed by a World War, and then of operating as a global superpower during the subsequent Cold War. Almost inevitably, the presidency became a more powerful institution. Congress itself recognized the need for broad presidential authority to handle the complexities of domestic and foreign affairs, and it delegated substantial discretionary authority to the executive branch in numerous statutes. After a brief insistence by the Supreme Court on limiting the breadth of such delegations during the New Deal,⁹ the Court retreated, and it had in any event already shown more acceptance of broad statutory delegations in the area of foreign affairs.¹⁰

The legislative veto emerged as one of the ways in which Congress sought to check this vast authority it was delegating to the executive branch. Although commentators vary in their estimates of the precise number, from 1932 until the time of the Chadha decision, Congress included hundreds of legislative veto provisions in a wide variety of statutes.¹¹ Many of these provisions were enacted in the 1970s, in the wake of the Vietnam War and the Watergate scandal, and many of those 1970s provisions concerned foreign affairs.¹² Notable examples include:

• The 1973 War Powers Resolution, which Congress enacted over President Nixon’s veto, provided that “at any time that United States Armed Forces are engaged in hostilities outside the territory of the United States, its possessions and territories without a declaration of war or specific statutory authorization, such forces shall be removed by the President if the Congress so directs by concurrent resolution” (Pub. L. No. 93-148, § 5(c), 87 Stat. 555, 556–57 (1973) (codified at 50 U.S.C. § 1544(c))). This provision still exists in the Resolution, despite Chadha, but it is generally thought to be unconstitutional.

• The 1974 Nelson–Bingham Amendment to the Foreign Military Sales Act required the President to provide Congress with detailed information concerning upcoming government-sponsored arms sales in excess of $25 million, and it gave Congress twenty calendar days to veto the sale by concurrent resolution (Pub. L. No. 93-559, § 45, 88 Stat. 1795, 1814 (1974)). Two years later, in the Arms Export Control Act, Congress broadened the reporting obligation and gave itself thirty days to veto proposed arms sales (Pub. L. No. 94-329, sec. 211, § 36, 90 Stat. 729, 743 (1976)).¹³

• The 1976 National Emergencies Act, which was designed to centralize congressional control over presidentially declared emergencies under numerous other statutes, provided that any national emergency declared by the President would cease if Congress voted to terminate it by concurrent resolution (Pub. L. No. 94-412, § 202(a)(1), 90 Stat. 1255 (1976)).¹⁴ Subsequent emergency power statutes, including most notably the 1977 International Emergency Economic Powers Act, were made subject to the provisions of the National Emergencies Act.

• The 1978 Nuclear Non-Proliferation Act, which authorized the President to approve the export of nuclear materials, equipment, and technology to non-nuclear weapons countries under various circumstances, allowed Congress to block such export through concurrent resolution (Pub. L. No. 95-242, §§ 303, 304, 306, 307, 401, 92 Stat. 120, 130, 134, 137–39, 144–45 (1978)).¹⁵ The Act also authorized the President to negotiate nuclear cooperation agreements with other nations, but it stated that such agreements had to be approved by a concurrent resolution of Congress (§ 104, 92 Stat. at 122–23).

Also during the 1970s, Congress deviated from the APA’s usual report-and-wait approach and “extended the legislative veto to a series of agency programs involving rulemaking” (Bruff & Gellhorn 1977, p. 1371). These included agency programs relating to education, health, energy, and elections (see id.). Meanwhile, there was intense controversy over the constitutionality of the legislative veto mechanism. Critics argued, in particular, that it allowed Congress or its sub-units to take legally binding actions without following the procedures required by Article I, Section 7 of the Constitution for enacting legislation. Ultimately, this controversy came to a head in the Chadha litigation.

2.2 INS v. Chadha

Chadha concerned Congress’s use of a one-house veto to override a decision by the Attorney General to suspend statutorily-mandated deportations. Historically, Congress had used private bills to grant individuals relief from deportation (Mansfield 1941, p. 285).¹⁶ In 1940, however, it granted the Attorney General the authority to suspend deportations based on grounds of extreme hardship, subject to a congressional ability to override suspensions pursuant to a concurrent resolution (54 Stat. 670, 672, § 20(c) (1940)). This provision was amended in 1952 to allow for congressional override through a one-house resolution (66 Stat. 163, 216, § 244(b) (1952)).

The Attorney General delegated the suspension authority to the Immigration and Naturalization Service (INS) within the Justice Department, which had immigration judges review and make decisions about requests for suspension of deportations. In 1974, after a hearing, an immigration judge suspended the deportation of Jagdish Chadha, a citizen of Kenya, who had overstayed his student visa. In late 1975, Representative Joshua Eilberg, who chaired the House Judiciary Subcommittee on Immigration, Citizenship, and International Law, introduced a resolution in the House opposing the suspension of the deportation of Chadha and five other individuals (H. Res. 926, 94th Cong., 1st Sess., 121 Cong Rec. 40247 (1975)). The House passed the resolution without debate or recorded vote, and Eilberg’s resolution was not made available to members of the House before it was voted on (Craig 1988, p. 22). Eilberg did state on the House floor, however, that “[i]t was the feeling of the committee, after reviewing 340 cases, that the aliens contained in the resolution [Chadha and five others] did not meet these statutory requirements, particularly as it relates to hardship; and it is the opinion of the committee that their deportation should not be suspended” (121 Cong. Rec. 40800 (Dec. 16, 1975)).¹⁷

After the government subsequently ordered Chadha to be deported, he appealed to the U.S. Court of Appeals for the Ninth Circuit, arguing that the legislative veto was unconstitutional. The INS agreed with Chadha’s argument, so the Ninth Circuit allowed the House and Senate to file amicus briefs in support of the veto provision. The Ninth Circuit held that the veto provision was unconstitutional (Chadha v. Immigration and Naturalization Service, 634 F.2d 408 (9th Cir. 1980)), and the Supreme Court affirmed.¹⁸ Meanwhile, the D.C. Circuit had held legislative veto provisions unconstitutional in two decisions relating to agency rulemaking.¹⁹

In invalidating the legislative veto provision in Chadha, the Court emphasized that Article I, Section 7 of the Constitution requires that legislation be approved by both houses of Congress and presented to the President for approval, something that does not occur with legislative vetoes (462 U.S. at 946–51). The Court then reasoned that the exercise of the veto in this case was “essentially legislative” because it had “the purpose and effect of altering the legal rights, duties, and relations of persons” (id. at 952). The Court further explained: “The one-House veto operated in these cases to overrule the Attorney General and mandate Chadha’s deportation; absent the House action, Chadha would remain in the United States. Congress has acted and its action has altered Chadha’s status” (id.).

The Court in Chadha also emphasized that “the fact that a given law or procedure is efficient, convenient, and useful in facilitating functions of government, standing alone, will not save it if it is contrary to the Constitution” (id. at 944). Relatedly, it declined to defer to the political branches’ decision to include legislative veto provisions in numerous statutes, noting that “our inquiry is sharpened, rather than blunted, by the fact that congressional veto provisions are appearing with increasing frequency in statutes which delegate authority to executive and independent agencies” (id.).²⁰

The Court’s reasoning appeared to invalidate all legislative veto provisions, a point emphasized by Justice Powell in his concurrence and Justice White in his dissent (id. at 959-60 (Powell, J., concurring); id. at 967 (White, J., dissenting)). Shortly after issuing its decision in Chadha, the Court summarily affirmed the D.C. Circuit decisions that had invalidated legislative veto provisions relating to agency rulemaking (see Process Gas Consumers Grp. v. Consumer Energy Council, 463 U.S. 1216 (1983)).

Chadha has been read as disallowing not only congressional override provisions like the ones at issue in these cases, but also what could be called “legislative permission” (or “positive veto”) provisions whereby the executive is granted the authority to do something only if it obtains the approval of a committee, a congressional house, or both houses.²¹ As noted above, there was precedent for such provisions even before 1932. The basic rationale of Chadha is that Congress can only create new legal effects by acting through the full legislative process, and that rationale appears to apply to legislative permission provisions as well as to classic legislative veto provisions.²² As will be explained in Section 3, this point is important when considering some of the ways that Congress has responded to Chadha.

2.3 Claims about Chadha’s Significance

The conventional wisdom, at least among many legal scholars, is that Chadha significantly undermined the constitutional system of checks and balances, especially in the area of foreign affairs. Under the Constitution, legislative, executive, and judicial powers are assigned to different branches of the federal government. By parceling out authority in this manner, the constitutional structure is thought to provide a check on efforts by any one branch to expand its authority. As James Madison argued in The Federalist Papers, the three branches would, “by their mutual relations, be the means of keeping each other in their proper places” (Federalist No. 51, p. 320).²³

Chadha, it is said, undermined this constraint. As a practical matter, the argument goes, Congress must delegate significant discretionary authority to the President. The effect of this arrangement, in practical terms, is that the President ends up with significant authority to make law, despite the Constitution’s assignment of that authority to Congress. Before Chadha, it was much easier for Congress to overturn exercises of this lawmaking authority because Congress did not need to enact a new statute, which would potentially require overcoming a presidential veto. The legislative veto mechanism, according to this claim, helped preserve our system of checks and balances in a world of broad statutory delegations.

This is the gist of the argument that Justice White made in his celebrated dissent in Chadha, in which he contended that:

Without the legislative veto, Congress is faced with a Hobson’s choice: either to refrain from delegating the necessary authority, leaving itself with a hopeless task of writing laws with the requisite specificity to cover endless special circumstances across the entire policy landscape, or, in the alternative, to abdicate its lawmaking function to the Executive Branch and independent agencies.²⁴

Many legal scholars have made similar claims, especially with respect to presidential power over foreign affairs. Harold Koh (1990, pp. 129, 142) has described legislative vetoes as “linchpins of the post-Vietnam era framework legislation” and has argued that, when combined with the Supreme Court’s decision in Dames & Moore v. Regan (which suggested a generous approach to finding statutory authority for presidential actions relating to foreign affairs), Chadha “radically undercuts the Youngstown vision of balanced institutional participation in the national security process.” Jules Lobel has contended that Chadha “struck down as unconstitutional a form of the legislative veto which provided a critical congressional check in the War Powers Resolution, [the National Emergencies Act], and [the International Emergency Economic Powers Act]” (Lobel 1989, p. 1416). Oona Hathaway (2009, p. 197), focusing in particular on the effect of Chadha on presidential power over the making of international agreements, has argued that “[t]he Court’s decision in Chadha upended foreign relations law in the United States to an extent rarely appreciated.”²⁵

These claims intensified during the Trump administration, especially in response to some of Trump’s invocations of statutorily-delegated emergency authority. Richard Pildes (2019) argued, for example, that Chadha might be the “most consequential” Supreme Court separation-of-powers decision “in terms of the real-world balance of power between Congress and the president.”²⁶  Aziz Huq (2019) argued that Chadha “addressed a nonexistent problem of congressional overreach even as it created a new peril of executive abuse,” especially with respect to declarations of national emergencies. Tess Bridgeman and Stephen Pomper (2019) described Chadha as dealing “a heavy blow” to Congress’s ability to check presidential uses of force. Geoffrey Manne and Seth Weinberger (2019), commenting on presidential invocations of emergencies, suggested that “[p]erhaps it’s time for the Supreme Court to reconsider the legislative veto.”²⁷

A related concern raised by legal scholars is that many of Congress’s broad delegations of authority before Chadha were premised on the assumption that Congress would have the legislative veto as a check. Because the veto provisions were generally viewed as severable, however,²⁸ Congress was left simply with the broad delegations, and amending those delegations is difficult because that, too, potentially requires overcoming a presidential veto. As Kristen Eichensehr (2021, p. 1275) has recently argued, Chadha, when combined with the presumption in favor of severability, may have “locked in” broad delegations to the executive that would not have occurred otherwise. Stephen Vladeck (2020) similarly has argued that “[t]hanks to the Supreme Court’s 1983 decision in INS v. Chadha, it has become practically impossible for Congress to claw back these authorities once they have been delegated.” To the same effect, Hathaway (2009, p. 254) has observed that, “[w]ith its decision in INS v. Chadha, the Supreme Court pulled away this last strand of congressional power over ex ante congressional-executive agreements, leaving behind only the now-unsupervised broad delegations of power from Congress to the President.”

Many of the complaints about Chadha, as the above examples indicate, have been focused on issues relating to foreign affairs. There appears to be greater acceptance of the demise of the legislative veto in the domestic policy sphere. This is in part because it is generally thought that the accretion of presidential power has been especially extensive in foreign affairs.²⁹ In addition, domestic delegations are more likely to involve administrative agency implementation as opposed to direct presidential action, and the agencies are subject to various checks on their authority, including judicial review. Relatedly, the judicial checks on the President tend to be lower in the foreign affairs context,³⁰ potentially increasing the need for legislative checks.

One of the many observers who has emphasized this distinction between domestic and foreign affairs is Joseph Biden, in a law review article that he wrote as a senator, shortly after Chadha was decided (see  Biden 1984). Biden thought that the government “can and will function, for the most part, more effectively in the absence of the legislative veto,” and that “Congress still has ample means for controlling the agencies” (id. at 685, 688 (emphasis added)). Indeed, he suggested that the Court in Chadha “has done Congress a service” (id. at 693). He explained:

By doing away with the legislative veto, the Court may have helped clear the lobbyists off Capitol Hill, and has stemmed the movement toward full-blown congressional review of agency regulations. Congress can now concentrate on its most important role—shaping national policy on issues that affect the public as a whole. It is almost possible to say that the Supreme Court has saved Congress from itself. (Id.)

But Biden suggested that the veto was still needed for foreign policy. Discussing in particular the checks and balances with respect to foreign arms sales and war powers, he argued that “Chadha strikes at the foundations of carefully structured systems for sharing power between the executive and legislative branches” (id. at 685). Whether Biden still holds this view today now that he is president is unclear.

In short, many commentators view Chadha as having had a consequential and negative impact on the separation of powers between the legislative and executive branches, especially in the context of foreign affairs. For expositional purposes, I will refer to this view as the conventional wisdom, although, as will become evident, it is not the conventional wisdom among political scientists. This view is admittedly intuitive: by eliminating one of Congress’s tools for overseeing the executive branch, it seems reasonable to assume that the Court in Chadha undermined Congress’s ability to control the exercise of its statutory delegations. This article aims to unsettle that intuition.

3. A CRITIQUE OF THE CONVENTIONAL WISDOM

Drawing in part on the political science literature on the subject, this Section critiques the claim that Chadha’s disallowance of legislative vetoes significantly undermined Congress’s ability to control exercises of statutorily-delegated authority. It begins by describing the relatively limited effect of the legislative veto on congressional-executive relations prior to Chadha, and it then describes the various ways that Congress has responded to Chadha, in some instances by limiting executive authority and in other instances by fashioning workarounds that have allowed it to retain influence.

3.1 Legislative Vetoes Before Chadha

In thinking about the effect of Chadha, it is useful to begin by considering the record of Congress’s use of the legislative veto mechanism prior to the decision. During the fifty-year period from the first adoption of a legislative veto provision in 1932 until the Supreme Court’s decision in Chadha, Congress enacted hundreds of veto provisions, but it used them infrequently—an average of less than one use per statutory provision during this long period (Gilmour & Craig 1984, p. 374). More importantly, the use of the legislative veto was concentrated in just a few areas. Other than for overturning deportation suspensions such as in Chadha, the veto was primarily used in three contexts: executive reorganization, spending deferrals, and agency regulations (Gibson 1992, p. 30; Gilmour & Craig 1984).³¹ Given that much of the focus of the critique of Chadha has centered on foreign affairs, it is especially noteworthy that, before Chadha, there was not a single instance in which Congress used a legislative veto to overturn a presidential foreign affairs action, whether concerning the exercise of emergency powers or any other issue (Gilmour & Craig 1984, p. 374; Pomerance 1985, pp. 260–61). This fact is consistent with Congress’s well-documented passivity in the foreign affairs area, due in part to the relatively low interest that its constituents tend to have in this area.³²

Of course, the threat of the use of the veto, whether express or implicit, might have had an important constraining effect on presidential action even when the veto was not used. Such an effect would not necessarily be evident merely by observing the instances in which the veto was invoked.³³ A number of political scientists have studied this issue, however, including with respect to foreign affairs, and they have found much less effect than has been assumed by legal scholars.

In their study of the effect the legislative veto mechanism prior to Chadha, Robert Gilmour and Barbara Craig found that, for some foreign affairs actions, such as arms sales, the threat of the veto may have had some effect. They pointed to a number of instances in the 1970s and early 1980s in which there were legislative veto threats (sometimes prompted by lobbying from the pro-Israel group AIPAC), and presidents modified an arms sales proposal in response (although they suggest that this might have happened even without the veto threat) (Gilmour & Craig 1984, pp. 375–76).³⁴ In other foreign affairs areas, however, Gilmour and Craig found little effect from the existence of the veto provisions: “When Congress has not demonstrated a strong intention to use its veto power, as in most foreign trade and war powers situations, policies have not been affected” (id., p. 382).³⁵ Moreover, other studies indicated that, even in the arms sales area, the threat value of the veto was declining in the years leading up to Chadha.³⁶

In a different study, Joel Aberbach (1989) analyzed fourteen congressional oversight tools, including the legislative veto. Based on both statistics of use and surveys of staffers from the 95th Congress (in the late 1970s), he found that the legislative veto was one of the least effective of all of the tools (pp. 134–36).³⁷ Much more effective, he found, were tools such as informal staff communications and oversight hearings. In light of this finding, Aberbach noted that Chadha “does not seem quite so significant for oversight as one might have guessed” (id., p. 134).

Jessica Korn reached a similar conclusion in her 1996 book on the subject. Based on a number of case studies, involving both domestic and foreign policy, she found that “members [of Congress] never found the legislative veto shortcut useful for achieving their desired policy outcomes” (Korn 1996, p. 117). As she explained:

When executive branch actions clearly violated bipartisan congressional expectations, members responded swiftly and emphatically by enacting, or making credible threats to enact, new law. And in the more common situation in which members either were not certain as to the best way of implementing a particular policy, or had specific opinions that would not easily be supported by a majority of their colleagues, they preferred to participate informally in policy implementation through report-and-wait provisions rather than the legislative veto. (Id., pp. 117–18.)

In other words, members of Congress typically found that they either did not need the shortcut of the veto or did not want to be pressured into using it. Korn concluded that, “[s]ince the legislative veto never served as a significant oversight tool, its invalidation did not weaken congressional power to exercise control over policy outcomes” (id. at 79).

By contrast, Martha Gibson argued in a 1992 book that Chadha was important, but not in the way that many legal scholars have assumed. According to her account, for some foreign policy issues, where Congress tends to be assertive (for example, arms sales to countries in the Middle East), the legislative veto had helped to reduce the intensity of inter-branch conflict by encouraging behind-the-scenes consultation and compromise. And she argues that after Chadha, Congress became more assertive with respect to arms sales, not less, by making its disagreements with the President more public and by enacting statutes to block presidential action. The “ultimate effect of the Chadha decision,” she concluded, “has been to increase congressional obstruction of the president in foreign policy rather than decrease it” (Gibson 1992, p. 26). The key point here is that Chadha did not, even under this account, result in a reduction of congressional control of executive action pursuant to statutory delegations, even in foreign affairs.³⁸

One can of course quibble with the methodologies used in the above studies, and even the studies at face value do not show that the legislative veto mechanism was entirely inconsequential prior to Chadha. But this literature does suggest a more cautionary take on that question than is found in much of the legal scholarship that discusses Chadha.³⁹

In considering the threat effect of the legislative veto before Chadha, it is also worth keeping in mind that the constitutional validity of legislative vetoes was not settled prior to the decision. Starting with the Hoover administration, numerous presidents and their legal advisers had argued that legislative vetoes were unconstitutional, and presidents had often issued signing statements or other objections to that effect (Bruff 2006, p. 233; Wheeler 2008, p. 100).⁴⁰ Moreover, as discussed above, many of the legislative veto provisions in existence at the time of Chadha, especially ones relating to foreign affairs, had been enacted during the 1970s, at the height of the constitutional controversy (Brough 2020, pp. 456–59; Cooper & Hurley 1983, pp. 3–4). Such constitutional uncertainty likely reduced the threat value of the veto provisions in the pre-Chadha period.⁴¹

* * *

The above discussion merely suggests that the legislative veto may not have been a significant source of congressional leverage in many areas prior to Chadha. It does not necessarily show that Chadha’s elimination of the option of a legislative veto is unimportant today. It might be that there is more need today for this sort of check on the executive—for example, because other checks (such as informal norms) have diminished as a result of heightened partisanship, or because executive power has grown since the decision. Of course, a greater need would not necessarily translate into a change in congressional behavior, but perhaps Congress’s incentives or ability to overcome collective action limitations have somehow changed since Chadha, such that it would be more likely than in the past to use this tool of oversight. Or, on the other hand, it might be that heightened partisanship would have led to greater use of the legislative veto as a means of obstructing the operations of government.

For a variety of reasons, it is difficult to make confident predictions about what would have happened. Among other things, imagining a world in which the Court decided Chadha differently presents a host of counterfactual questions—about what the Court would have said,⁴² about how Congress would have responded to the anti-Chadha decision, and how the executive would have in turn responded to Congress. Nevertheless, there is one thing that we can observe, which is how Congress responded to Chadha itself.

Critics of Chadha often emphasize how, in a number of instances, Congress responded to the decision by stripping out the legislative veto provisions and replacing them with provisions requiring a full statute to override executive action. They also suggest (somewhat inconsistently) that Congress found itself stuck after Chadha with broad delegations that it would not have made in the absence of an enforceable legislative veto. A closer examination reveals that the picture is more complicated: In some areas, Congress has reduced executive authority altogether. In others, Congress has replaced veto provisions with report-and-wait provisions that, as a practical matter, appear to yield a lot of the same results as vetoes. And in still other areas, Congress has continued to enforce its veto ability through informal means.

3.2 Post-Chadha Reductions in Executive Authority

It is true, as commentators have noted, that in a number of statutes, Congress responded to Chadha by replacing one- or two-house veto provisions with a requirement of a joint resolution, which is a statute that requires presidential signature or an override of a presidential veto. This is what happened, for example, with legislation concerning export controls (see Pub. L. No. 99-64, § 301(b), 99 Stat. 120, 160 (1985) (codified at 42 U.S.C. § 2153(d))), national emergencies (see Pub. L. No. 99-93, § 801, 99 Stat. 448 (1985) (codified at 50 U.S.C. § 1622(a)(1))), and, importantly, arms sales (see Pub. L. No. 99-247, 100 Stat. 9 (1986) (codified at 22 U.S.C. § 2776)). But this tells only a small part of the story.

In some areas, Congress responded to Chadha by requiring the President to obtain a new statute before being able to act at all, often with expedited procedures that limit debate and that ensure that the issue gets to the floor of Congress. This is what Stephen Breyer (while a judge on the First Circuit) had suggested in response to Chadha, calling it a “confirmatory law” approach (Breyer 1984, p. 793). Congress took this approach, for example, for executive reorganization, which was the issue that was the genesis of the modern legislative veto (see Pub. L. No. 98-614, 98 Stat. 3192 (1984)). The requirement of a joint resolution of approval means that the executive merely can propose something and have it subjected to expedited statutory consideration, which in effect gives each house of Congress a veto (Kaiser 1984, p. 251).⁴³ As Louis Fisher noted in a report for the Congressional Research Service in 2005, on the issue of executive organization “the position of the President was weakened. The President now had to obtain the approval of both houses within a specified number of days in order to reorganize executive agencies” (Fisher 2005, p. 3).

This is also ultimately what happened with respect to presidential impoundment of funds, which was one of the few other areas in which the legislative veto had been used frequently before Chadha. Prior to the decision, the President was given the power to defer the spending of appropriated funds, subject to the possibility of a one-house veto (see Pub. L. No. 93-344, § 1013(b), 88 Stat. 297, 335 (1974)). After Chadha, however, the legislative veto provision, which was now invalid, was found by the courts to be inseverable from the rest of the statute, thereby invalidating as well the delegation of deferral authority to the President. Congress then changed the law to require that the President obtain statutory approval to defer spending (see Pub. L. No. 100-119, § 206, 101 Stat. 754, 785 (1987) (codified at 2 U.S.C. § 684)). Under the amended Impoundment Control Act, if presidents wish to defer spending, they must obtain congressional approval of a rescission bill within forty five days after impounding the money. Otherwise, the amount in question must be made available for expenditure. The effect of this change, as Fisher notes, “was to limit the President to routine, managerial deferrals and prohibit the use of deferral authority to delay the spending of funds simply because the President disagreed with the budgetary priorities enacted into law” (Fisher 2005, p. 3).

In short, in two of the areas in which the legislative veto had been used most frequently, reorganization and impoundment, Congress responded to Chadha by requiring that the President obtain statutory approval. In other words, it did the opposite of the conventional story and actually reduced presidential power. Moreover, these examples contradict the claim made by some legal scholars that, after Chadha, Congress was stuck with broad delegations that it could not as a practical matter revise.

3.3 Post-Chadha Workarounds

Even when Congress has not imposed statutory approval requirements, it has found other ways to retain influence over the exercise of its statutory delegations. This section highlights two of these mechanisms: report-and-wait provisions, and informal vetoes. It also briefly describes an array of additional tools that Congress can and sometimes does employ. Critics of Chadha have tended to focus too exclusively on statutory overrides of executive action, neglecting the many other ways that Congress can have influence.

3.3.1 Increased Use of Report-and-Wait Provisions

One way in which Congress responded to Chadha was by increasing its use of “report-and-wait” provisions (Berry 2016, pp. 164, 276). Pursuant to these provisions, certain executive actions cannot take effect for a specified period of time, giving Congress an opportunity to consider whether to enact legislation disallowing the action, pursuant to streamlined legislative procedures. As noted by Korn and others, members of Congress may actually prefer these provisions to legislative vetoes because the provisions place less pressure on Congress to directly oppose proposed executive actions (because concerns do not need to be expressed in terms of a threatened veto). Importantly, this less confrontational posture does not necessarily reduce Congress’s leverage. As Gilmour and Craig (1984) note, report-and-wait provisions “can serve much the same function as a veto with regard to oversight of agency rulemaking or planning” (p. 389). And the Court in Chadha made clear that these provisions are constitutional (462 U.S. at 935 & n. 9).

Report-and-wait is the general approach that Congress has used for administrative regulations. As noted above, this approach predates Chadha. But after Chadha, Congress supplemented it with the Congressional Review Act of 1996. Under this Act, major federal regulations must be reported to Congress and cannot take effect for sixty days, giving Congress an opportunity to override them, using expedited procedures (including the disallowance of filibusters in the Senate) (5 U.S.C. §§ 801–808).⁴⁴ Prior to the Trump administration, Congress had only rarely overturned a regulation through this process, but it did so sixteen times in the 115th Congress (2017–18).⁴⁵ Congressional Democrats also invoked the Act after Biden took office in order to block several administrative regulations proposed near the end of the Trump administration.⁴⁶

In foreign affairs, Congress famously used a report-and-wait approach in response to President Obama’s proposed nuclear deal with Iran, pursuant to which Iran agreed to freeze its nuclear weapons program in return for a relaxation of sanctions. Under the Iran Nuclear Agreement Review Act of 2015 (INARA), Congress mandated that the deal would not take effect for sixty days while Congress had an opportunity to review and potentially disapprove it by statute (Pub. L. No. 114–17, 129 Stat. 201 (codified at 42 U.S.C. § 2160e)).⁴⁷ INARA was novel in applying the report-and-wait approach to a purportedly non-binding agreement, but the approach itself was not novel.

Congress has also used report-and-wait and other monitoring provisions to regulate some types of executive agreements—that is, binding international agreements that are made by the executive branch without submitting them to the Senate for its advice and consent. Consider, for example, nuclear cooperation agreements.⁴⁸ Since the 1950s, Congress has imposed a variety of substantive requirements for such agreements, relating to limitations on the use of nuclear material (see Pub. L. No. 85-479, § 3, 72 Stat. 276, 277 (1958)). Prior to Chadha, it had also provided that such agreements would not take effect if, during a sixty-day review period, Congress passed a concurrent resolution of disapproval (Pub. L. No. 95-242, sec. 401, § 123(d), 92 Stat. 120, 144–45 (1978)). Congress replaced the concurrent resolution provision after Chadha with a joint resolution provision (Pub. L. No. 99-64, § 301(b)(1), 99 Stat. 120, 160 (1985) (codified at 42 U.S.C. § 2153(d))), but it also did a lot more. It required the President to consult with the Senate Foreign Relations Committee and House Foreign Affairs Committee for at least thirty days of continuous session to ensure that the proposed agreement is consistent with the statutory requirements (id., § 301(a)(2), 99 Stat. at 159 (codified at 42 U.S.C. § 2153(b))). It also subsequently required the President to keep those committees “fully and currently informed” of initiatives and negotiations related to new and amended cooperation agreements (Pub. L. No. 110-369, sec. 202, § 123(e), 122 Stat 4028, 4033 (2008) (codified at 42 U.S.C. § 2153(e))). In addition, Congress required that, during a subsequent sixty-day review period, the committees must hold hearings on proposed agreements and submit a report to their respective houses recommending whether the agreements should be approved or disapproved (Pub. L. No. 99-64, § 301(a)(3), 99 Stat. 120, 160 (1985) (codified at 42 U.S.C. § 2153(d))). Finally, it provided that, if a President seeks to exempt a cooperation agreement from the statutory requirements, the agreement will become effective only if Congress enacts a joint resolution approving the agreement (id., § 301(b)(2)).

3.3.2 Continued Inclusion (and Informal Enforcement) of Veto Provisions

It also turns out that, despite their presumed unconstitutionality, Chadha did not eliminate the inclusion of legislative veto provisions in statutes, a point that Louis Fisher flagged many years ago (Fisher 1993). In fact, Congress has included many more legislative veto provisions in statutes enacted after Chadha than in statutes enacted before the decision. This practice could be seen as a form of congressional “self-help” of the sort described by David Pozen (2014) in his account of informal enforcement of the separation of powers.

In a 2016 book, Michael Berry reported that “over 80 percent of all veto statutes enacted in U.S. history were passed after” Chadha (p. 275).⁴⁹ A large majority of these statutes have given the congressional appropriations committees veto authority over particular expenditures (id., p. 276). Although often framed in terms of committee permissions (e.g., for the reprogramming of funds or to exceed informally-agreed-upon spending caps) rather than vetoes, it is generally assumed that they fall within Chadha’s scope.

Importantly, even though these provisions are thought to be unenforceable in court, the executive branch has not ignored them. Despite frequently protesting the provisions in presidential signing statements (Garvey 2010, p. 21),⁵⁰ the executive has since Chadha often chosen to work informally with the relevant congressional committees to address their concerns as expressed in vetoes or threatened vetoes.⁵¹  Chadha presumably does not preclude voluntary compliance with veto provisions. To take one prominent example, in the late 1980s, President George H.W. Bush made an agreement with congressional leaders pursuant to which his administration could provide humanitarian aid to the Nicaraguan contras but would terminate the aid on a specified date unless four congressional committees approved its continuation. Because the agreement was not treated as legally binding and thus did not affect any legal rights or duties, it did not appear to run afoul of Chadha (Glennon 1989).⁵² The key point here is that even when veto provisions seem to matter in congressional-executive relations, it is not because of their judicial enforceability, suggesting that the Supreme Court had less of an impact here than is commonly claimed.

Executive agencies have a variety of out-of-court incentives to work with Congress pursuant to these provisions, including concerns about cutbacks in funding if they fail to do so.⁵³ Congress’s control over appropriations gives it substantial leverage over agencies in these matters; indeed, regardless of Chadha there is already in effect a “one-house veto” over new appropriations (which the executive branch is dependent upon), given that either house can block the renewal of appropriations unless it gets what it wants (Lawrence 2021, p. 1073).⁵⁴ Moreover, Congress can include riders in appropriations bills to retaliate against agencies that decline to cooperate, giving Congress in that setting something close to a two-house veto, since presidents are generally reluctant to block appropriations bills (Fisher 2014, p. 170; MacDonald 2010). This enforcement regime that has flourished after Chadha illustrates how judicial interventions in the separation of powers may have limited effect, especially when those interventions conflict with bipartisan political branch preferences.⁵⁵

3.3.3 Other Mechanisms of Constraint

The above discussion highlights only two of the workarounds that Congress can employ, and often has employed, in order to retain some control over statutory delegations. Congress has a wide variety of other tools at its disposal. They include delegating more narrowly, which, despite Justice White’s “Hobson’s Choice” claim, is at least sometimes feasible. This is what it ultimately did for the deportation suspension authority at issue in Chadha: in the 1990s, Congress substantially restricted the circumstances under which such relief could be granted.⁵⁶ For another example, Congress has used its appropriations power to maintain strong control over how U.S. development aid is parceled out to other countries, and it managed to rebuff Trump administration efforts to cut such aid (Moore 2021, p. 1171). More generally, it is difficult to think of examples after Chadha in which there was majority support in both houses of Congress for reducing the scope of a statutory delegation and Congress was blocked by a presidential veto from doing so. The claim that Congress delegated broad authority to the executive before Chadha as part of a bargain that involved having a legislative veto and was subsequently unable to pull back its part of the bargain after Chadha simply does not hold up.

Congress can also include sunset provisions in its delegations so that the executive is dependent on statutory renewal, which effectively gives each house of Congress a veto.⁵⁷ An example of a statute with such a provision is the Defense Production Act, which was enacted in 1950, after the start of the Korean War (Pub. L. No. 81-774, 64 Stat. 798 (1950) (codified at 50 U.S.C. §§4501 et seq.)).⁵⁸ The Act allows the President to, among other things, contract with and incentivize U.S. manufacturers to expand and redirect their production to support national defense and responses to national emergencies, and both President Trump and President Biden invoked it in response to the COVID-19 pandemic. Because of sunset provisions, the authority has to be periodically renewed by Congress. Many commentators have suggested that similar sunset provisions should be added in the war powers and emergency contexts.⁵⁹

Relatedly, Congress sometimes imposes time limits on the exercise of delegated authority. For example, Section 232 of the Trade Expansion Act of 1962 allows the President to impose restrictions on imports when the Commerce Department finds that products are “being imported into the United States in such quantities or under such circumstances as to threaten to impair the national security” (19 U.S.C. § 1862(c)(1)(A)).⁶⁰ But Congress has specified that, upon receiving the Commerce Department’s report, the President has ninety days to decide whether to take action, and thereafter has fifteen days to implement the action and thirty days to submit a written statement to Congress explaining the action or inaction (19 U.S.C. § 1862(c)). The Court of International Trade has treated these procedural limitations as judicially enforceable (Transpacific Steel LLC v. United States, 466 F. Supp. 3d 1246 (C.I.T. 2020)).⁶¹ Similar timing rules apply to trade sanctions imposed by presidents under Section 301 of the Trade Act of 1974 (which has frequently been used in recent years against China).⁶²

Congress can also enhance its ability to monitor the exercise of statutorily-delegated authority by imposing consultation, reporting, and publication obligations on the executive branch. This has been Congress’s approach, for example, to addressing the phenomenon of executive agreements, most of which are concluded based on statutory authority. Instead of trying to limit presidential authority to conclude such agreements, Congress has insisted in the Case-Zablocki Act on reporting and publication (see 1 U.S.C. § 112 b).⁶³ A more elaborate example of congressional involvement in the exercise of delegated authority is the Global Magnitsky Human Rights Accountability Act, which authorizes the President to impose economic sanctions on foreign persons who are identified as engaging in human rights abuses or corruption. The Act requires the President, in deciding whether to impose sanctions, to consider information provided jointly by the chairperson and ranking member of certain congressional committees as well as credible information obtained by foreign countries and nongovernmental human rights organizations, and it also requires the President to respond within 120 days to requests from congressional committee leadership to determine whether a foreign person has engaged in sanctionable activity under the law and whether the President intends to impose sanctions (see Pub. L. No. 114-328, div. A, title XII, subtitle F, Dec. 23, 2016, 130 Stat. 2533, at 22 U.S.C.§ 2656 note).

Another option for Congress is to structure a delegation so that it requires the involvement of particular executive branch actors, or the use of particular executive branch processes of decisionmaking. That is part of the approach, as noted above, in Section 232 of the Trade Expansion Act. Another example is Congress’s regulation of the President’s delegated authority to limit foreign investment in the United States. Congress has given the Committee on Foreign Investment in the United States (CFIUS), an inter-agency committee, a substantial role in decisions in this area. Originally established by executive order in 1975, the CFIUS review process has subsequently been heavily regulated by Congress (see  Jackson 2020). Measures like these can promote greater “internal separation of powers” in lieu of, or in addition to, inter-branch checks and balances.⁶⁴ And there are even more dramatic steps that Congress could consider, such as amending the Administrative Procedure Act so that it would apply to at least some presidential actions involving statutorily-delegated authority.⁶⁵

There are also a number of less formal steps that Congress can take that do not require legislation. It can make its views known through a variety of “soft law” mechanisms, such as statements in oversight hearings and non-binding resolutions, which may influence executive action.⁶⁶ It can also seek informal commitments from the executive branch when it enacts a delegating statute about how the authority will be used—embodied, for example, in Statements of Administrative Action.⁶⁷

 * * *

In light of these and other checks on presidential exercises of statutorily-delegated authority, as well as the uncertainty about the significance of the legislative veto mechanism in the first place, it is difficult to state with confidence that Congress has lost significant leverage vis-à-vis the executive as a result of Chadha. The following case studies further illustrate this point.

4. CASE STUDIES

This Section presents three case studies to demonstrate that Chadha has had less of an impact on congressional-executive relations than many legal scholars have claimed. These case studies have been chosen because they are the most prominent examples of situations in which legal scholars have argued that Chadha especially undermined congressional control over presidential action. As will be seen, the Trump administration complicates the picture a bit, although this article argues against generalizing from that administration.

4.1 War Powers

Many scholars contend that the Constitution was designed to require that presidents obtain congressional authorization before using military force, except when engaging in certain acts of self-defense.⁶⁸ In practice, however, presidents often use military force in non-defensive situations without obtaining congressional authorization. They do so either by claiming constitutional authority as Commander in Chief or by claiming that their uses of force fall within the scope of prior statutes delegating use-of-force authority. Some commentators contend that Chadha is partially responsible for this state of affairs. Because of Chadha, it is argued, the only way for Congress to overturn presidential uses of force is through the enactment of new statutes, which very likely requires overcoming a presidential veto.⁶⁹

There are a number of reasons to question this claim. Consider the use of the legislative veto prior to Chadha. Congress first included a legislative veto provision in war powers-related legislation in 1957, in a joint resolution that was ostensibly designed “[t]o promote peace and stability in the Middle East” (Pub. L. No. 85-7, 71 Stat. 5 (1957)).⁷⁰ This resolution authorizes the President to undertake military assistance programs in the general area of the Middle East and states further that “if the President determines the necessity thereof, the United States is prepared to use armed forces to assist” any Middle Eastern nation or group of nations “against armed aggression from any country controlled by international communism” (id., § 2). Congress specified that the resolution would expire “when the President shall determine that the peace and security of the nations in the general area of the Middle East are reasonably assured by international conditions created by action of the United Nations or otherwise except that it may be terminated earlier by a concurrent resolution of the two Houses of Congress” (id., § 6 (emphasis added)). The resolution is still in effect today, more than sixty years later (see  Waxman 2019a), and Congress has never threatened to use the veto provision, let alone actually used it.

Congress again included a legislative veto provision relating to the use of force in 1964, in the Gulf of Tonkin Resolution. The Resolution, enacted in response to alleged attacks by the North Vietnamese against U.S. navy vessels, provided an arguable statutory foundation for the Vietnam War by stating that “Congress approves and supports the determination of the President, as Commander in Chief, to take all necessary measures to repel any armed attack against the forces of the United States and to prevent further aggression” (Pub. L. No. 88-408, 78 Stat. 304 (1964) (emphasis added)).⁷¹ The Resolution had an expiration provision similar to the one in the 1957 Middle East resolution: by its terms, the Gulf of Tonkin Resolution would expire “when the President shall determine that the peace and security of the area is reasonably assured by international conditions created by action of the United Nations or otherwise, except that it may be terminated earlier by concurrent resolution of the Congress” (id., § 3 (emphasis added)). The Vietnam War escalated and dragged on for many years, and it became highly unpopular, and yet Congress never invoked the concurrent resolution veto provision, even though this was long before Chadha. Instead, Congress finally repealed the Resolution in 1970—through a regular statute, which President Nixon did not veto (Pub. L. No. 91-72, § 12, 84 Stat. 2053, 2055 (1971)).⁷² Congress also began taking additional steps to end the war by, for example, cutting funding—again by ordinary legislation.⁷³

The experience of the Vietnam War prompted the enactment of another legislative veto provision—this time in the 1973 War Powers Resolution (WPR), which Congress enacted over President Nixon’s veto (Pub. L. No. 93-148, 87 Stat. 555 (1973) (codified at 50 U.S.C. §§ 1541–1548)). The WPR, which is still in effect, imposes consultation and reporting obligations on the President and mandates that uses of force must cease after sixty days if they have not received congressional approval. Section 5(c) of this Resolution also provides that when U.S. forces are engaged in hostilities without congressional authorization, “such forces shall be removed by the President if the Congress so directs by concurrent resolution” (50 U.S.C. § 1544(c)). In other words, it contains a two-house legislative veto provision.⁷⁴ After Chadha, Congress left this provision in place, while also providing separately for expedited consideration of joint resolutions of termination (50 U.S.C. § 1546a).

There have been numerous proposed resolutions under Section 5(c), but not one has obtained majority support from both houses of Congress, either in the ten years before Chadha or in the nearly forty years since Chadha.⁷⁵ Nor have scholars found evidence that the existence of the veto mechanism prior to Chadha influenced presidential decisionmaking with respect to the use of force.⁷⁶

One might argue, of course, that, but for Chadha, the veto provision in the WPR would have become important. While it is difficult to disprove such a counterfactual, there are a number of reasons to doubt it. As an initial matter, it seems unlikely that Chadha is the only reason that the full Congress has not invoked Section 5(c).⁷⁷ In fact, some commentators contend that Section 5(c) does not even violate Chadha. Their argument is that the WPR does not delegate authority to the President, and hence that Section 5(c), unlike other legislative veto provisions, is not attempting to overturn exercises of delegated authority.⁷⁸ In any event, even if it were clear that courts would find Section 5(c) to be unconstitutional, there might be good reasons for Congress to nevertheless invoke it—and, indeed, members of Congress often have proposed resolutions under this section, and a few of these resolutions have received majority support from the House. A two-house resolution disapproving a presidential use of force would provide a strong political challenge that might influence the President’s actions regardless of whether it were judicially enforceable.⁷⁹ It might also increase Congress’s ability to have standing to raise constitutional challenges to a president’s use of force.⁸⁰ And it could serve as a “soft law” marker of congressional non-acquiescence in presidential claims of war powers authority.⁸¹

Despite the potential incentives to invoke Section 5(c), the full Congress, as noted above, has never done so. Part of the reason is that many presidential uses of force are not controversial in Congress. Even when they are, Congress is typically reluctant to directly confront the President on the issue. Instead, it is often content merely to wait and see how the operation unfolds before taking a position. Even when some members of Congress want to confront the President, it is difficult for them to muster the majorities in both houses of Congress that would be required to invoke Section 5(c). For example, there were efforts in Congress in 1999 to force an end to President Clinton’s use of force in Kosovo, but these efforts failed to obtain majority support in either the House or the Senate.⁸² The same thing happened in 2011, in connection with President Obama’s use of force in Libya.⁸³ An enforceable legislative veto would have made no difference. Moreover, in the rare instances in which congressional majorities are willing to oppose presidential actions relating to war, they probably do not need the legislative veto in order to do so. For example, when U.S. troops suffered casualties in Lebanon in 1982, Congress pressured President Reagan into agreeing by statute to a 180-day deadline on the military mission (S.J. Res. 159, 97 Stat. 805 (1983)). Similarly, after U.S. troops suffered casualties in Somalia in 1993, Congress enacted a funding cutoff for operations that would take effect after a specified date—again, by means of a regular statute (Pub. L. No. 103-139, § 8151, 107 Stat. 1418, 1475–77).

The best evidence that an enforceable veto provision might have made a difference emerges from the Trump administration. Two instances in particular stand out. First, in 2019, Congress passed a joint resolution directing the President to end U.S. involvement in the Saudi-led war in Yemen, but President Trump vetoed the resolution, and Congress lacked the votes to override the veto.⁸⁴ Importantly, however, an enforceable legislative veto provision in the WPR likely would have made no difference in that situation. President Trump maintained that U.S. assistance to the Saudis in Yemen did not constitute an involvement of U.S. forces in hostilities under the WPR;⁸⁵ if not, Section 5(c) was inapplicable. The failure of the WPR to define what constitutes an involvement in hostilities is one of several weak points in the statute. Trump also suggested that a legislative effort to restrict his actions in this situation would interfere with his Article II authority as Commander in Chief,⁸⁶ a claim that he could also have made against a legislative veto provision. This illustrates a broader point: legislative vetoes only work for presidential actions based on statutes, but many presidential war powers actions are ostensibly based, at least in part, on the President’s constitutional authority. In any event, shortly after Biden took office, he ended U.S. support for the Saudi-led war in Yemen (Hubbard & Almosawa 2021).

The second instance was when the Trump administration used a drone strike in 2020 to kill an Iranian general, Qassem Soleimani, while he was in Iraq. In response, Congress voted to approve a joint resolution, which “direct[ed] the President to terminate the use of United States Armed Forces for hostilities against the Islamic Republic of Iran or any part of its government or military, unless explicitly authorized by a declaration of war or specific authorization for use of military force against Iran” (S.J. Res. 68, 116th Cong. (2020)). President Trump vetoed the resolution, however, and it never took effect. It might seem that the lack of an enforceable legislative veto was important at least in this situation. Congress’s action there, however, was mainly just political theater: by the time Congress passed the proposed statute—months after the killing of Soleimani—there were no ongoing hostilities between the United States and Iran. As Trump noted in his veto message, “the United States is not engaged in the use of force against Iran” (Message to the Senate 2020). As a result, there was no use of force to terminate. If, as it seems, the goal of the resolution was to send a message of concern to the President, Congress accomplished that goal. In addition, in carrying out the strike on Soleimani, the Trump administration had relied in part on post-WPR statutory authorization (the 2002 AUMF), something that likely could not be overridden by the use of Section 5(c) even if that section were otherwise valid (Anderson & Taylor 2020).

This last point invites the question whether the 2001 and 2002 AUMFs would have been better statutes if they had contained enforceable legislative veto provisions, in the sense of allowing a stronger congressional check on presidential uses of force under them. Other than the Soleimani strike, which was already completed before Congress could have invoked such a provision, it is difficult to think of presidential actions under those statutes that would have been overridden by majorities in both houses of Congress. For example, Presidents Obama and Trump used the 2001 AUMF to wage a substantial conflict against the Islamic State, but there is no indication that congressional majorities wished to stop them.⁸⁷ In fact, Obama went so far as to invite Congress to pass a law specifically authorizing the conflict with the Islamic State and restricting his AUMF authority, but it took no action (Peters 2015; Bradley & Goldsmith 2016, p. 637). The bigger issue with those AUMFs is that they seem to go on forever, even as circumstances change, and Congress is never forced to decide what it wants to authorize. This is why some commentators have called for the inclusion of sunset clauses in such statutes.⁸⁸ The virtue of sunsets, as opposed to a legislative veto mechanism, is that they would force Congress to take action to reapprove the use of force, rather than engaging in its usual shirking in this area. Moreover, it is possible that including legislative veto provisions in AUMFs would just encourage Congress to draft them even more broadly than it has already done.

In sum, even during the Trump administration, it seems unlikely that an enforceable legislative veto provision would have made much of a difference in the war powers area. In any event, as emphasized throughout this article, it is hazardous to generalize about separation-of-powers reforms based primarily on one administration.

4.2 Arms Sales

Under the Foreign Assistance Act of 1961, Congress delegated to the executive branch broad authority to sell arms to foreign nations. In 1974, Congress added a two-house legislative veto provision to this legislation, in the Nelson–Bingham Amendment (Pub. L. No. 93-559, § 45, 88 Stat. 1795, 1814 (1974)). Under this provision, the President could not proceed with a sale if Congress objected within twenty days by concurrent resolution, a period that was later changed in the Arms Export Control Act to thirty days. That concurrent resolution provision was never used before Chadha, but there is some evidence that its existence might have had an effect on some arms sales decisions by the executive, especially with respect to sales in the Middle East.⁸⁹ It is difficult to know with certainty because it is very possible that presidents would have altered the sales for political reasons once they were confronted with congressional opposition, regardless of whether Congress had the votes to pass a veto.⁹⁰ Nor is it clear that any such influence was positive; most of the examples suggesting that the veto provision had an influence involve lobbying by special interest groups.

In any event, it is far from clear that Congress has lost much leverage in this area as a result of Chadha. Indeed, as discussed earlier, there is evidence indicating that Congress may have become more assertive in the arms sales area after Chadha. As with many examples, the picture changed somewhat during the Trump administration.

Congress has used two mechanisms after Chadha to retain an influence on arms sales: report-and-wait, and informal holds. Under the Arms Export Control Act as amended after Chadha, the executive branch must formally notify Congress at least thirty calendar days before concluding a major arms sale to another country (which is triggered by various threshold amounts) (22 U.S.C. § 2776(b)).⁹¹ To block a sale under this process, Congress must pass a joint resolution of disapproval, which requires the enactment of a statute, potentially over the President’s veto. Congress has never successfully blocked a sale through this process, although, as discussed earlier, it also never successfully used a legislative veto mechanism to block an arms sale prior to Chadha. The key advantage of report-and-wait is that it allows Congress time to express concerns about potential arms sales before they take effect.

This formal review process for arms sales has been supplemented since even before Chadha with an informal hold process (Kerr 2021, p. 1). Under that process, the State Department submits a preliminary notification of a prospective major arms sale to the Senate and House foreign affairs committees before the executive branch takes further formal action.⁹² As a matter of practice, the executive branch will not typically initiate the formal review process for arms sales if a committee member has placed a hold during the informal notification stage.

Using both formal and informal processes, Congress has a long history of involving itself in arms sales, especially in the Middle East, and it is difficult to see any significant reduction in such involvement after Chadha. As a Congressional Research Service report published in 2020 notes, “members’ questions over specific arms sales to states in the region have helped frame the terms of the arms sales debate, and have shaped the broader, sometimes contentious relationship between the executive and legislative branches over U.S. foreign policy.”⁹³ As discussed earlier, some studies indicate that congressional assertiveness in this area actually increased after Chadha.⁹⁴ And Congress appears to have genuine influence. The Congressional Research Service explains that “Congress has—by expressing strong opposition to prospective arms sales, during consultations with the executive branch—affected the timing and the composition of some arms sales, and may have dissuaded the President from formally proposing certain arms sales” (Kerr 2021, p. 4). To take one recent example, during the Trump administration, several members of Congress used the informal review process to hold up arms sales to Turkey (see  Browne & Hansler 2020).

That said, some of the traditional processes of congressional involvement appear to have eroded during the Trump administration. The report-and-wait process is subject to an exception, which predates Chadha, for situations in which the President certifies “that ‘an emergency exists’ which requires the sale … to be made immediately ‘in the national security interests of the United States.’” (Kerr 2021, p. 6.) When invoking this exception, the President must provide Congress with a “detailed justification for his determination, including a description of the emergency circumstances” that necessitated the action and a “discussion of the national security interests involved” (id.). The Trump administration controversially invoked that exception in 2019 when concluding $8.1 billion in arms sales to Saudi Arabia and other Gulf states, claiming the expedited sales were needed to deter Iranian aggression in the region. Congress passed several bills in an effort to block the sales but was unable to overcome presidential vetoes (see  Demirjian & Itkowitz 2019). (Presidential declarations of emergencies are discussed more generally below in Section 4.3.)

By invoking the statutory emergency exception, the Trump administration also bypassed informal holds that had been placed on the sales in question (which had been placed because of concerns about civilian casualties resulting from the military operations of the Saudi-led coalition in Yemen and because of Saudi Arabia’s murder of the journalist Jamal Kashoggi). The Trump administration apparently also contemplated abandoning the informal review process altogether after members of Congress placed a hold on certain additional arms sales to Saudi Arabia (LaForgia et al. 2020; Stark 2020). In November and December 2020, the Trump administration gave formal notice of an intent to conclude a major arms sales agreement with the United Arab Emirates and Saudi Arabia, without fully complying with the informal review process (Thomas et al. 2020, p. 14; Ryan & DeYoung 2020).

It is difficult to know whether the actions of the Trump administration represent a trend. After taking office, President Biden promptly paused the Trump arms sales, but then decided after a review to proceed with some of them (Crowley 2021; Crowley & Wong 2021). The key point here is that, at least prior to the Trump administration, Congress appears to have retained significant leverage over arms sales, notwithstanding Chadha. The Trump administration’s questionable use of the emergency exception and its threat to abandon the informal review process appear to have reduced Congress’s leverage, making the absence of a legislative veto mechanism more significant. But there is no indication so far that the Biden administration will follow the Trump administration’s hardball approach in this area, and congressional leaders have already been using the informal hold process to exert influence on the Biden administration’s arms sales decisions (see  Gramer 2021).

None of this is to suggest that the arms sales process is currently ideal. It may be, for example, that Congress should amend the relevant law to limit the emergency exception, or to require affirmative congressional approval for certain sales, as some commentators have suggested.⁹⁵ Such changes would be perfectly consistent with Chadha. If anything, the availability of an enforceable legislative veto provision—which was never used to block an arms sales agreement before Chadha—might reduce Congress’s incentive to enact these other statutory limitations by providing a false sense of control over the statutory delegation.

4.3 Declarations of Emergencies

The issue of emergency power, more than any other, lies at the heart of recent claims about the significance of Chadha. Unlike many other countries’ constitutions, the U.S. Constitution does not specifically provide for the exercise of emergency powers (Goitein 2020). It does of course expressly or implicitly give the President and Congress powers that may be needed in times of emergency. It also allows for the suspension of the writ of habeas corpus during time of invasion or rebellion (U.S. Const. art. I, § 9). But otherwise it is silent about powers specific to emergency conditions.

There are over a hundred statutory provisions, however, that delegate authority to the President in times of emergency, and they allow the President to determine whether a relevant emergency exists (Elsea et al. 2020; Brennan Center 2020). In the 1976 National Emergencies Act (NEA), Congress sought to centralize its regulation of presidentially-declared emergencies. The NEA stated that existing emergencies based on statutory authority would terminate in two years and that any newly-declared emergencies would expire in a year, unless terminated earlier by presidential proclamation, although it allowed presidents to renew them each year without limitation. For these newly-declared emergencies, the NEA requires the President to issue a proclamation in the Federal Register and transmit it to Congress, and to identify the statutory authority being invoked. The NEA also originally contained detailed provisions giving Congress the ability to terminate presidentially-declared emergencies through a two-house legislative veto, using expedited procedures (Pub. L. 94-412, § 202(a)(1), 90 Stat. 1255 (1976)). After Chadha, this provision was removed and replaced with a provision allowing Congress to terminate emergencies through a joint resolution—that is, through a statute (Foreign Relations Authorization Act, Fiscal Years 1986 and 1987, Pub. L. No. 99-93, 99 Stat. 405, 448 (1985)).

Since the enactment of the NEA, presidents have declared 71 emergencies, 40 of which are still in effect (Brennan Center 2021). The vast majority of these emergencies have been declared pursuant to authority granted in the International Emergency Economic Powers Act (IEEPA), which was enacted in 1977 (Harrell 2020). IEEPA was designed to narrow the President’s emergency authority under the Trading with the Enemy Act of 1917 (TWEA), but it nevertheless grants the President broad authority to regulate international economic transactions after declaring a national emergency (Casey et al. 2020). IEEPA is typically the statute that presidents invoke when imposing economic sanctions on other countries and their officials. A number of the emergencies still in effect have been renewed repeatedly, such as the emergency declared by President Carter during the Iranian hostage crisis.⁹⁶

Presidential use of emergency power became especially controversial during the Trump administration, most notably in connection with Trump’s use of an emergency declaration to justify his diversion of funds to build a wall on the border with Mexico, after Congress declined to approve the funding he had requested for the wall (Baker 2019). In doing so, he invoked a statute that authorizes the President, during emergencies that require the use of the armed forces, to initiate military construction projects “not otherwise authorized by law that are necessary to support such use of the armed forces” (10 U.S.C. § 2808). Congress approved a joint resolution that would have terminated this emergency, but Trump vetoed it (Message to the House 2019). But this was not the end of the matter: near the end of Trump’s presidency, Congress managed to overturn Trump’s redirection of funds for the wall, in a statute enacted over Trump’s veto (Edmondson 2021).⁹⁷ Even before that, courts were resisting Trump’s invocation of emergency authority in this context.⁹⁸

As discussed in Section 2.3, a number of commentators have argued that Chadha is largely to blame for the lack of sufficient checks and balances relating to emergency power. However, there are a number of reasons to doubt this claim. First, it seems unlikely that the possibility of using a legislative veto played much of a causal role in Congress’s broad delegations of emergency authority to the President. These delegations are not in the NEA itself; they are in numerous other statutes, including many enacted or amended since Chadha.⁹⁹ Moreover, the NEA adopted a legislative veto mechanism at the height of the constitutional controversy over that mechanism, when Congress could not have been confident about its ultimate fate in the courts. In addition, when Congress amended the NEA after Chadha to replace the concurrent resolution provision with a joint resolution provision, it could have attempted to cabin the President’s emergency powers in light of its loss of the veto, but it did not do so.

It may be, of course, that Congress would have used the two-house veto mechanism since then if Chadha had come out the other way. Politically, this would be most likely to occur during times when both houses of Congress have been controlled by a party other than the President’s party. Even then, however, most of the presidential declarations of emergency since Chadha have not been particularly controversial in Congress. Perhaps for that reason, prior to 2019, during the Trump administration, Congress had never invoked the joint resolution provision that replaced the legislative veto provision, whether successfully or unsuccessfully. In fact, even though the NEA specifies that Congress is supposed to meet every six months to review whether to use the joint resolution mechanism to end an emergency (see Pub. L. No. 94-412, § 202(b), 90 Stat. 1255, 1256), it had not even done that prior to 2019.

As a result, much of the recent claim that Chadha matters in this area comes down to a focus on the Trump administration, and in particular the controversy over the border wall. Even for that example, it is noteworthy that Congress eventually overrode the presidential action, although it required overcoming a presidential veto to do so. President Biden, moreover, formally ended the border emergency after taking office,¹⁰⁰ and he subsequently canceled contracts for border wall construction (Miroff 2021). Biden also ended another Trump-declared emergency under IEEPA that was used as the basis for imposing sanctions against officials of the International Criminal Court,¹⁰¹ and he retracted Trump executive orders that attempted to ban the Chinese-owned TikTok and WeChat software applications (Exec. Order No. 14,034, 86 Fed. Reg. 31,423 (June 9, 2021)).

In sum, even for the issue of emergency power, it is difficult to find a compelling case that Chadha has made much of a difference, especially if one abstracts away from the Trump administration. To be sure, it may be desirable for Congress to trim some of its delegations of emergency authority, as a number of commentators have been suggesting. For example, it might make sense to make the initial duration of presidentially-declared emergencies shorter, or to require congressional approval for their renewal. But, as with the war powers and arms sales examples, the existence of an enforceable legislative veto might actually discourage such reform efforts, because the veto mechanism could give Congress a false sense of having a meaningful check.¹⁰² The key point, once again, is that Chadha is not the problem.

            * * *

It is worth emphasizing that the above case studies were selected because they are purportedly the best examples of where Chadha has undermined checks and balances. Even these examples, however, provide relatively little support for the claim. Indeed, even during the Trump administration, enforceable legislative vetoes likely would have made only a modest difference—not at all on war powers, perhaps some delay in an arms sales package to the Middle East (which Biden ultimately approved in part anyway), and probably earlier resistance to the construction of the southern border wall (which Congress eventually opposed more successfully by overriding a Trump veto). To the extent that checks and balances are not working well in those areas (and this article makes no claim that they are), it appears to be primarily for reasons unrelated to the demise of the legislative veto.

5. FORM AND FUNCTION IN CHADHA

The reasoning of the Court in Chadha is, for the most part, highly formalistic—relying, for example, on notions of what is “essentially legislative” (462 U.S. at 952),¹⁰³ and discounting arguments grounded in efficiency,¹⁰⁴ or in the historical practice of interbranch relations concerning the veto.¹⁰⁵ Commentators have often criticized the decision’s formalism,¹⁰⁶ and have argued that the decision is problematic when considered on functional grounds.¹⁰⁷

This article’s principal response to that argument is descriptive—that Chadha had less of an impact on congressional control over statutory delegations than is commonly assumed. This Section explains why, even if the decision were more consequential, the decision’s functional desirability would still be subject to reasonable debate. Section 1 points out shortcomings in some of the standard functional critiques of Chadha, and Section 2 describes some of the ways in which legislative veto mechanisms can potentially undermine good governance.

To be clear, this Section makes no claim that the costs of having enforceable legislative vetoes would outweigh the benefits. Such an assessment, assuming it were feasible, would almost certainly require additional empirical information. In addition, as indicated earlier, any assessment of the effect of Chadha on our separation-of-powers system would need to consider difficult counterfactual questions about how congressional-executive relations would have evolved if the Supreme Court had come out the other way (or if it had abstained from deciding). Such an assessment would also ideally consider the effect of legislative vetoes on a system-wide basis and not merely based on a select number of instances in which such mechanisms appear to be either useful or harmful. The goal of this Section is much more modest—simply to explain how it might be possible, despite the formal nature of the Court’s reasoning in Chadha, to develop a functional defense of the decision.

5.1 Standard Functional Critiques of Chadha

Critics of Chadha have claimed that the Supreme Court neglected to give sufficient weight to the positive role that legislative veto mechanisms can play in the modern regulatory state. This section highlights weaknesses in some of the standard critiques.

5.1.1 Undermining the “Balance” Between the Branches?

One common critique of Chadha is that it undermined a proper “balance” of authority between the branches.¹⁰⁸ The claim is that the substantial delegations of authority required by the rise of the administrative state shifted a lot of authority to the executive branch, and that legislative vetoes were a necessary adjustment to maintain a balance of authority.

The central problem with this argument is that, as others have pointed out, we have no agreed-upon baseline for assessing the proper balance of authority between the branches.¹⁰⁹ Even if one adopts an originalist methodology (and many critics of Chadha are not in fact originalists), this is unlikely to produce any clear sense of an appropriate balance. Moreover, even if there were an agreed-upon baseline, legislative-executive relations have evolved in numerous conflicting directions since the Founding, and since Chadha, and assessing the overall impact of these many changes on some sort of balance is an impossible task. As Liz Magill has cogently observed, “we do not know what balance means, how to measure it, or how to predict when it might be jeopardized” (Magill 2001, p. 605). Nor is there a coherent theory about which developments that might affect institutional balance merit legal responses, as Eric Posner has pointed out (Posner 2016, pp. 1709–10).

This is not to argue, of course, that our separation of powers system is working well. The Madisonian idea that each branch would have sufficient tools and incentives to resist encroachments on its constitutional authority, to the extent that it ever made sense, has been substantially undermined by the rise of hyper-partisanship.¹¹⁰ The incentives of members of Congress often do not line up with the long-term institutional prerogatives of the legislative branch. To the extent that Congress imposes checks on executive authority, these checks often depend more on partisan disagreement than on institutional loyalty. Moreover, even without regard to partisanship, Congress faces a host of collective action problems that make it difficult for it to oppose unilateral executive actions.¹¹¹ But these problems are all unrelated to Chadha’s disallowance of legislative vetoes. Moreover, while legislative vetoes—especially of the one-house or committee variety—might reduce some of the collective action problems, they might also, as discussed below in Section 2, exacerbate problems of partisan gridlock.

5.1.2 Allowing Drift from the Enacting Congress’s Preferences?

Another functional critique of Chadha, made most prominently by William Eskridge and John Ferejohn (1992), is that the legislative veto mechanism helps ensure that the executive branch does not stray too far away from the median preferences of the Congress that originally delegated the statutory authority. Using spatial models, they describe how agency policies are likely to drift away from median congressional preferences and how legislative veto mechanisms could in some instances limit such drift. As they explain, “agency lawmaking can sometimes result in drastic shifts in policy from one Presidency to the next,” whereas “[t]he introduction of a two-house legislative veto has the effect of moving policy outcomes back toward those that would occur under the original understanding” (pp. 538, 543).

This argument is unpersuasive for several reasons. First, as a number of commentators have noted, it gives insufficient attention to the likelihood of congressional drift. ¹¹² When there are shifts in congressional party alignment and divided government, the legislative veto is itself likely to result in deviations from the enacting Congress’s median preferences. Party control of the houses of Congress shifts more frequently than party control of the White House, and divided government is more common in recent decades than unified government.¹¹³ Moreover, in the regulatory context, there is at least an administrative law check on agency drift from a statutory regime, a check that does not apply to exercises of the legislative veto.

In addition, Eskridge and Ferejohn overstate the ability of the executive branch to operate near the presidential veto override threshold without congressional resistance or correction. In practice, Congress has a number of ways to resist presidential action without triggering a presidential veto. For example, Congress can use omnibus bills and appropriations legislation to package provisions in a way that makes it less likely that they will be vetoed.¹¹⁴ Perhaps in part for that reason, presidential vetoes have been relatively rare during the last several administrations—less than two vetoes per year.¹¹⁵ And sometimes Congress is able to overcome a presidential veto.¹¹⁶

5.1.3 Upsetting an Institutional Bargain?

Another critique of Chadha, made most prominently by Aziz Huq (2014), suggests that it upset an institutional bargain between the legislative and executive branches—one that both institutions found to be useful for modern governance.

Huq contends that bargains between government institutions “are often (if not always) a desirable means of resolving constitutional ambiguities, adapting to changed conditions, and realizing new policy goods” (p. 1686). Under his account, if the legislative and executive branches want to trade away some of their authority, courts should generally allow the trade. Although the President normally has the authority to block legislative changes through the Article I, Section 7 process, Huq suggests that the courts should have allowed the President to give up some of that authority by accepting legislative veto provisions. Huq quotes with approval the Supreme Court’s statement in NLRB v. Noel Canning that judges should “hesitate to upset the compromises and working arrangements that the elected branches of Government themselves have reached” (p. 1602).¹¹⁷

This institutional bargain argument may be persuasive for some issues, but it is less so for legislative vetoes. A close examination of the history of the legislative veto does not reveal any institutional agreement about its desirability, let alone its legality. As noted earlier, presidents and their legal advisers frequently objected to legislative veto provisions from the very beginning of the practice during the Hoover administration, a point that the Court noted in Chadha.¹¹⁸ While presidents have not always been consistent in their objections, it is difficult to see in the historical practice a genuine institutional agreement, especially one that should bind presidential successors. Moreover, by the time of Chadha, the veto landscape had shifted dramatically, as Congress during the 1970s had greatly expanded the number of veto provisions and had inserted them into sensitive areas of foreign affairs. So even if there had been an earlier institutional bargain (for example, for the issue of executive reorganization), it is not clear that it continued to apply in light of the changed circumstances.

It can be argued, of course, that if presidents found the legislative veto provisions objectionable, they should have vetoed the bills containing them. But it is difficult to treat a decision by a president not to do so as evidence of a genuine institutional agreement about the desirability or legality of the provision. The modern practice of constitutional signing statements arose in part so that presidents could signal their doubts about particular provisions without blocking legislation, and presidents routinely have used such statements to object to legislative veto provisions.¹¹⁹

This institutional bargain argument is also in tension with the common claim that, as a practical matter Congress has no choice but to delegate broad discretionary authority to the executive branch. This is the “Hobson’s Choice” argument that Justice White made in his dissent in Chadha. If it has no choice, then there is no reason for the executive branch to compromise and accept a legislative veto mechanism. There is no genuine bargain to be had. Nor, in any event, did Chadha prevent inter-branch bargaining over the control of statutorily-delegated authority. As we have seen, the decision left substantial room for the branches to develop workarounds, which have flourished in the wake of the decision.

Finally, it is worth remembering that not all things that the branches agree to are good for governance. Increasing the veto points in Congress and further empowering committees and individual members of Congress may not be good for governance, especially given how partisan and non-institutional Congress has become. When individuals or private firms are bargaining, we assume that they are aiming to maximize their own welfare, but this assumption is harder to maintain about Congress. Therefore, any analogy to private party bargains (which Huq acknowledges is imperfect (pp. 1603, 1655) has limited value when applied to the separation of powers.

5.1.4 Reduced Transparency?

Yet another functional critique of Chadha is that it reduced the transparency of congressional action. Neal Devins and Louis Fisher (2015) have contended that, “[i]n many cases, the Court’s decision [in Chadha] simply drove underground a set of legislative and committee vetoes that used to operate in plain sight” (p. 127). Others have echoed this claim, usually just citing back to Devins and Fisher.¹²⁰

It is difficult to evaluate this claim because unfortunately neither Devins and Fisher, nor anyone else, has attempted to document it. Without more evidence, skepticism is warranted. As Chadha itself illustrates, the legislative veto mechanism was not a model of transparency. The suspension of Chadha’s deportation occurred without any public discussion or recorded vote. Moreover, the informal veto regime that has flourished after Chadha also existed before Chadha. As discussed above in Section 3.1, the veto was not used frequently in practice, so any influence it had was largely informal. Devices like informal holds on arms sales predated Chadha. And the informal incentives that the executive branch has had since Chadha to accommodate congressional wishes, such as concerns about the renewal of appropriations, certainly existed prior to the decision.¹²¹ In fact, studies of the effect of the legislative veto prior to Chadha suggested that it resulted in significant, non-transparent interactions with administrative agencies.¹²²

In short, whatever “underground” there is in terms of congressional-executive relations does not appear to be the result of Chadha, and more evidence would be needed to establish that Chadha exacerbated the situation. It is also worth keeping in mind that at least some of the workarounds appear to have increased congressional engagement and transparency. Certainly this is true of the shift in some areas to legislative approval requirements, which involve direct voting and accountability, but it also may be true of the report-and-wait systems that have proliferated since Chadha, since they may be more likely to trigger public hearings, public complaints by members of Congress that are designed to put pressure on the executive, and proposals for legislation.¹²³

5.2 Potential Functional Problems with Legislative Vetoes

Legislative vetoes come in a variety of forms. They can involve committee vetoes, one-house vetoes, or two-house vetoes. In Chadha, for example, the Court was confronted with a one-house veto. As scholars documented before Chadha, these various veto mechanisms can create a number of problems for good governance.¹²⁴ In particular, they can increase the opportunities for special interest group influence, increase the likelihood of partisan gridlock, and reduce bureaucratic rationality in the administration of statutes.

First, as Harold Bruff and Ernest Gellhorn (1977) explained before Chadha, legislative veto mechanisms can increase the ability of special interest groups to influence government decisionmaking (pp. 1413, 1417).¹²⁵ Instead of being required to persuade majorities in both legislative chambers, and potentially also the executive branch, these groups may need to persuade only one chamber or one committee—or, as a practical matter, one committee chairperson. Some of that is true, of course, even in the absence of legislative veto mechanisms, but it is even truer with them. As Bruff and Gellhorn explained, “[i]n certain subtle ways, the presence of congressional review [under the veto provisions] allowed the influence of special interest groups in Congress to affect the substance of rules outside the [agency] comment process.”¹²⁶ Indeed, as we have seen, in the one foreign affairs area where the legislative veto seemed to have some effect before Chadha—that is, arms sales—it seemed primarily due to its facilitation of special interest group influence.¹²⁷

Second, these veto mechanisms can increase the likelihood of partisan gridlock. By allowing sub-units of Congress to block or overturn executive action, legislative vetoes allow the party that does not control the White House another means of impeding government action. As David Martin (1982) explained before Chadha, “[s]everal features intrinsic to the legislative veto device, … make impasse a significant risk under any such [legislative veto] scheme” (p. 280). Such gridlock is especially likely during times of divided government. With that in mind, it is useful to remember that, although there is (barely) unified government at the moment, divided government has been the norm in the post-Chadha era: during the thirty eight years between 1983 and 2021, there was divided government in all but ten years.¹²⁸ By most accounts, partisanship has also grown significantly during the forty years since Chadha,¹²⁹ which means that, if anything, an enforceable legislative veto would be more problematic today in terms of its effect on governance. And there is every reason to believe that these political trends will continue for the foreseeable future (Drutman 2021). In short, the political situation since Chadha has made the legislative veto mechanism less rather than more desirable.¹³⁰

Third, and perhaps most significantly, legislative veto mechanisms can increase the arbitrariness of government decisionmaking. As Lisa Bressman (2020) has noted, “[a] legislative veto may be exercised without public hearing, report, or statement of reasons, and may be passed without recorded vote,” and thus it “does not have the qualities of the administrative action it reverses—such as participation, transparency, and rationality” (p. 520). In Chadha itself, an individual’s ability to remain in the country was disallowed by a one-house veto that was issued without debate, explanation, or recorded vote.¹³¹

To put it differently, legislative vetoes effectively convert sub-units of Congress into the functional equivalent of administrative agencies, without the rationalizing forces of administrative law.¹³² Moreover, because the vetoes do not involve the regular legislative process, they bypass a lot of the rationalizing forces associated with Congress’s internal bureaucracy. This bureaucracy, which includes numerous nonpartisan expert bodies within Congress (such as the legislative counsels’ offices, the Congressional Budget Office, and the Government Accountability Office), “furthers Congress’s own internal separation of powers and safeguards the legislative process from executive and interest-group encroachment” (Cross & Gluck 2020, p. 1544). This point about rationalizing forces provides a functional explanation for why the Supreme Court has shown greater willingness to police delegations by Congress to its own agents than to administrative agencies.¹³³

To be sure, direct presidential actions, unlike agency actions, are not subject to the rationalizing forces of the APA, and some of President Trump’s actions, such as the travel ban, were thought by many observers to be arbitrary or poorly conceived. Even for the travel ban, however, there was extensive judicial review, and, in part because of that, the order was revised twice after significant additional vetting within the executive branch (see Trump v. Hawaii, 138 S. Ct. 2392, 2404 (2018)). None of that occurs with the use of the legislative veto.

That said, some of the functional concerns are probably lower for two-house vetoes than for other types of vetoes, especially in particular settings. It is possible, for example, that on balance a two-house veto in the emergency declarations context would be functionally desirable as a safeguard against potentially egregious abuses of authority. It is difficult to think of a constitutional rule, however, that would allow for such vetoes but disallow others that would be functionally more problematic.¹³⁴ If not, any functional assessment of Chadha’s impact needs to consider the veto landscape more broadly. In any event, for reasons discussed earlier, the two-house veto is also the least likely to make a difference in congressional-executive relations.

In light of the potential functional problems, it is perhaps not surprising that functional concerns about legislative vetoes have crossed the ideological spectrum. The vote in Chadha was 7-2, with liberal Justices like Brennan and Marshall, who were not known as formalists, joining with conservative, formalist justices like Burger. Moreover, although Justice Scalia and Professor Laurence Tribe were ideologically far apart, one thing that they agreed on was that legislative vetoes were bad for governance (see supra notes 125, 132). And, as Dennis Hutchinson has noted in his biography of Justice White (1998, p. 445), while White’s dissent was ostensibly functional, it “ignores the unsavory aspects of the legislative veto in action and seems unwilling to weigh competing concerns at comparable levels of care.”¹³⁵ Conversely, even though the reasoning in Chadha is highly formal, the Court seemed aware of, and was potentially influenced by, functional concerns.¹³⁶

To be clear, this article makes no claim that Chadha’s disallowance of legislative vetoes has prevented these political ills. As Section 3 made clear, Congress has developed a host of workarounds since Chadha, and it is quite possible that the workarounds replicate some or all of the functional problems associated with legislative vetoes. The point is simply that, to the extent that having a judicially enforceable legislative veto would matter to some extent, one should not assume that its elimination has been more harmful than beneficial. The dominant narrative in legal commentary has been that the legislative veto was a good innovation for modern governance and that the Court in Chadha allowed formalism to override functionalism, but this narrative is too simplistic.

6. CONCLUSION

It has been nearly forty years since the Supreme Court decided Chadha, and it is time, this article argues, for a reassessment of the decision and its effects. Much of the criticism of Chadha in recent years was prompted by concerns that were particular to the Trump administration, and the shift to a new administration provides a useful opportunity to reflect more broadly about the role of the legislative veto in the separation of powers. The claim here is not that there is an ideal balance of authority between the branches, assuming that one could discern such an ideal. Rather, it is simply that whatever imbalance exists is not especially attributable to Chadha.

A number of broader lessons also emerge from this study. First, to the extent that legal scholarship makes claims about how Congress and the executive branch interact, it is important for it to engage with political science work that has studied these interactions. Second, proposals for separation-of-powers reform should consider how congressional–executive relations have operated over time, including across changes in party control, rather than just at a particular moment in time that may not be representative. Finally, the workarounds and informal enforcement regimes that have flourished since Chadha illustrate the limits of judicial interventions relating to the separation of powers, especially when those interventions conflict with bipartisan political branch preferences. One problem with critiques of Chadha is the tendency of legal scholars to place the Supreme Court at the center of accounts of the separation of powers, when in fact political dynamics often play a predominant role.

REFERENCES