INS v. Chadha Killed the Legislative Veto and Left Congress Fewer Tools

Imagine a legislature that wants to delegate power to the executive and keep a remote control for taking it back. Pass a broad law, hand the agency the keys, and reserve the right to snatch them away with a vote of one chamber, with no second chamber and no signature. It is a clever arrangement. It is also, as the Supreme Court told Congress in 1983, a way of making law without the Constitution's permission.

The case that said so began with one man's deportation, and it deserves a careful reading, because it is among the most consequential separation-of-powers decisions of the modern era and one of the most misunderstood. Let me give you the facts, the holding, the dissent, and what Congress did next.

The man at the center of the case

Jagdish Rai Chadha was born in Kenya and came to the United States on a student visa. His visa expired. Under the Immigration and Nationality Act, the Attorney General had the power to suspend the deportation of an alien who had been physically present in the country for seven years, was of good moral character, and whose deportation would cause extreme hardship. That authority had been delegated to the Immigration and Naturalization Service. In 1974 an immigration judge held that Chadha met the standard and suspended his deportation. The ruling was reported to Congress, as the statute required.

Here is where the story turns. Section 244(c)(2) of the Act gave either house of Congress the power to veto the suspension, by resolution, within a set period after the case was reported. No second chamber. No presentment to the President. In December 1975, with the clock about to run out, a House subcommittee chairman moved to veto the suspensions in a group of six cases, Chadha's among them, and the House passed the resolution without recorded debate. Chadha's deportation was back on.

He challenged the veto. By the time the case reached the Supreme Court, the executive branch had switched sides and no longer defended the law, so the Senate and the House authorized their own intervention to defend it. That fact alone tells you how rare the clash was: the House and Senate, as institutions, defending their power against the President and a private litigant at the same time.

The constitutional text

Chadha is a case about words, so let us read them. Article I, Section 1: "All legislative Powers herein granted shall be vested in a Congress of the United States, which shall consist of a Senate and House of Representatives." Two houses, one Congress.

Then the text of Article I, Section 7, Clause 2, the Presentment Clause: "Every Bill which shall have passed the House of Representatives and the Senate, shall, before it become a Law, be presented to the President of the United States." And Clause 3 closes the door on the obvious dodge: "Every Order, Resolution, or Vote to which the Concurrence of the Senate and House of Representatives may be necessary (except on a question of Adjournment) shall be presented to the President of the United States."

Read that last clause once more. The Framers anticipated that Congress might try to legislate under other labels, an "order" or a "resolution" or a "vote." They said that anything that needs the concurrence of both houses, and that has the force of law, goes to the President. The legislative veto was just such a vote, except that it took one house rather than two. The Framers' answer to "can one house do it alone?" is in Article I, Section 1. No.

What the Court held

The Supreme Court decided INS v. Chadha on June 23, 1983, by a vote of 7 to 2. Chief Justice Burger wrote the opinion of the Court. His reasoning had a simple logic, and I would invite you to follow it step by step, as he did.

1. Was the House's action legislative in character? Yes. Burger asked whether it altered "the legal rights, duties, and relations of persons" outside the legislative branch. Chadha's right to remain in the country, determined by the Attorney General's delegated authority, was changed by the House's vote. A vote that changes the law is legislative, whatever its label.
2. Then it must satisfy Article I. Legislative acts require passage by both houses and presentment to the President. The House's one-house veto was neither bicameral nor presented.
3. The Constitution lists the exceptions. Burger pointed out that the Framers did specify when one house may act alone: the House's power to initiate impeachments, the Senate's power to conduct impeachment trials, to approve treaties and presidential appointments. Where the Framers wanted a single chamber to act, they said so. The absence of any such grant here counted against Congress.
4. Efficiency is not the test. The Court acknowledged that the legislative veto was a convenient device and that it might be defended as an administrative necessity in a complex age. It said that convenience and efficiency are not the primary objectives of democratic government. The Framers' choices for lawmaking were deliberately cumbersome, in order to preserve liberty. The passage still repays reading: bicameralism and presentment were designed as friction, and the friction is the feature.

Burger also drew on the Framers' concern with each of the two safeguards. Presentment gave the President a check against Congress's overreach and encouraged him to guard his own office. Bicameralism protected against the faction that could sweep one chamber, with the second chamber elected differently and on a different schedule. That is the logic of the Federalist: the Founders wanted laws to be hard to pass, so that the passage of one would signify something.

Justice Powell's narrower concurrence

Justice Powell concurred in the judgment, but he wrote separately, and his reason is worth knowing. He thought the majority's sweeping rule was unnecessary. In his view the House had adjudicated a particular person's rights, deciding that Chadha, by name, did not deserve to stay. That, Powell argued, is a judicial function. The legislature is exercising a power belonging to the courts, in a manner that goes beyond lawmaking, and it did so without the procedures of a trial. He would have decided the case narrowly on that ground and left the broader question for another day.

The distinction matters. Powell's approach would have left the broader question of other legislative vetoes open. The majority did not take that route, and its rule is wider. You may think Powell was wiser or you may think the Chief Justice was braver. Either way, you now know the law has two supports, not one.

Justice White's dissent and the number that stunned Congress

Justice White dissented, and Justice Rehnquist wrote a separate dissent that White joined. White's opinion is the best statement of the pragmatic case for the veto. Congress, he said, had come to rely on the device as a way of reconciling its need to delegate broad authority to the executive with its desire to keep some control. In a modern government, where Congress cannot possibly write every detail into a statute, the legislative veto lets it say yes to delegation with a safety catch. Without it, he warned, Congress would either have to write more specific statutes or delegate less carefully.

White counted what was at stake. By his tally, the decision called into question a great many statutory provisions: nearly two hundred statutory provisions with a legislative veto, by the figures he cited. He observed that the decision struck down, in one fell swoop, provisions in more laws enacted by Congress than the Court had cumulatively invalidated in its entire history. Whatever the exact count, the point is that Congress had spent decades building a system on a device that the Court had now removed.

His defense, in the end, rested on the notion that the legislative veto kept ultimate control in the hands of the elected branch, that it was a check on the executive and not an expansion of Congress's power. That is not a silly argument. A conservative should take it seriously, because it is the argument that sounds most like a defense of the people's representatives against the bureaucracy. Why shouldn't Congress hold a string on the agencies it creates?

The steelman, and the answer

Let me state White's case at its strongest, then answer it.

The strongest version: Congress created the agencies, and Congress should be able to correct them. Agencies exercise lawmaking power through rules. If Congress cannot readily reverse a rule it dislikes, the executive's power grows. The veto was a cheap, fast, accountable way to restore balance, and its abolition drove Congress toward ever more detailed statutes, or ever more vague ones.

Here is the answer, in order.

1. The Constitution already provides the correction. Congress can pass a law, over a presidential veto if it has the votes, that repeals or changes a rule. What the veto avoided was not an unconstitutional obstacle but the presidential signature, the very check the Framers put on lawmaking.
2. A one-house veto is lawmaking by the cheapest majority. One chamber, perhaps one committee's motion in a half-empty room, could change the legal rights of someone outside Congress. The Framers' fear was precisely that a legislature would concentrate power in its own hands. Federalist 73 and Federalist 51 describe the President's veto as a protection for the executive department and the public.
3. The device encouraged bad delegation. If Congress can always reach back and veto, it has less reason to write clear statutes in the first place. It may delegate broadly, hold the string, and then claim credit for the agency's good decisions and distance for the bad ones. A constitutional order that forces Congress to decide first and delegate narrowly is a better one.
4. The remedy for too much delegation is less delegation. That argument, which is also the argument behind the nondelegation doctrine, points to the real cure: Congress writes the rules itself, or sets standards in the statute that the agency must follow.

I would add something a talk-radio host might say and a constitutionalist would endorse: if Congress wants to veto something, let it do what everyone else in the country has to do when they want to change the law. Persuade two chambers and a President, or find two-thirds of each. If that is hard, good. That is how you know you are changing the law rather than the mood.

What Congress did in response

The legislative veto did not vanish overnight. Congress kept writing such provisions, and agencies continued to honor many of them informally, treating them as "committee approval" arrangements. But the principle was now clear, and the next generation of reformers looked for a constitutional substitute.

They found one in the Congressional Review Act of 1996. Signed by President Clinton in March 1996 as part of a small-business regulatory law, the Act requires every agency to submit new rules to Congress before they take effect. Congress may then pass a joint resolution of disapproval. The resolution must pass both houses and be presented to the President, exactly as Chadha demands. Its special feature is expedited procedure in the Senate: debate is limited and a simple majority suffices, so a minority cannot stall it with a filibuster. If it becomes law, the rule has no effect, and the agency is barred from issuing a "substantially the same" rule without new statutory authority. You can read the details in The Congressional Review Act Lets Congress Veto a Rule, if It Moves in Time.

The Act is Chadha-compliant, and it is also weak in practice. A President who favored the rule will veto the resolution. Until 2017 the Act had produced exactly one successful disapproval, in 2001. The lesson is the one the Chief Justice drew: you may not shortcut the Constitution, and the proper route is slow. If the proper route is too slow for Congress's purposes, that is a political problem to be solved by winning elections, not by redesigning the Constitution.

Chadha also pushed Congress toward other levers it did not lose: the power of the purse, which includes attaching conditions to appropriations; oversight hearings; confirmation of officers; and statutes that require an agency to report before acting. See Make Congress Vote on Every Major Rule, and Watch Accountability Return for the reform that follows from Chadha's logic: if lawmaking must go through both houses, then make Congress own the biggest rules.

Why this is a conservative case

Some conservatives look at Chadha and see a loss. Congress, the more representative branch, had a tool against the unelected agencies, and the Court took it away. I understand the feeling. But I think the opinion belongs in the conservative canon, for three reasons.

First, it takes the text seriously. Article I, Section 7 means what it says, and no convenient practice, however long it has lasted, overrides it. Second, it protects the most important structural idea in the Founding: that lawmaking should be hard, so that liberty is easy. The same principle that stops a one-house veto stops one-house regulation (see The Nondelegation Doctrine Says Congress Cannot Hand Its Job to Someone Else), one-agency lawmaking, and a President who wants to legislate by phone and pen. A conservative cannot consistently invoke bicameralism against executive overreach and then wave it away when Congress wants a shortcut.

Third, Chadha puts the burden where it belongs: on Congress. If the legislature wants to control the administrative state, it must legislate. That means long, hard, public votes. It is the opposite of the culture of delegation that has built the agencies in the first place. The Court did not weaken the legislature by closing the shortcut. It reminded the legislature what its job is.

A last word about the man himself. Chadha was not a villain or a hero, just a student whose paperwork collided with the Constitution. The decision gave him the result the statute had initially provided, a suspension of deportation approved by the executive that no single chamber could revoke. That is what the rule of law looks like at the scale of one household: procedures that apply to everyone, including the powerful.

The case in brief

1. The Constitution requires that anything with the force of law pass both houses and be presented to the President (Article I, Sections 1 and 7), and the Framers made that process cumbersome on purpose.
2. In INS v. Chadha (1983) the Court held 7 to 2 that a one-house legislative veto was legislative action and therefore unconstitutional without bicameralism and presentment.
3. Justice Powell would have decided on the narrower ground that the House adjudicated one person's rights, and Justice White's dissent warned that scores of statutes were affected.
4. The best defense of the veto, that Congress needs a string on its delegations, is answered by the fact that Congress already has constitutional levers and that the real cure for unchecked agencies is to delegate less.
5. The Congressional Review Act of 1996 is the Chadha-compliant substitute, and it works only when Congress and the President agree.

What you can do: Read Article I, Section 7 at the Constitution page, then look up how your Representative voted on the last major bill that delegated rulemaking power to an agency, using the scorecard.