Concept
Applying the Doctrine
Nondelegation doctrine
Applying the Doctrine
A statute that hands power to an agency can go wrong in two independent ways, and the first task in any fact pattern is to see which one is in play. The nondelegation doctrine bars Congress from transferring its legislative power to another branch.1 INS v. Chadha addresses the opposite risk: that Congress, having delegated, will keep a way of controlling the power that skips Article I. Any exercise of legislative power with the purpose and effect of altering the rights, duties, or legal relations of persons outside the legislative branch must conform to bicameralism and presentment, so a one-house legislative veto of an executive decision is unconstitutional.2 Both questions belong to the law of separation of powers, where many disputes pit Congress against the executive.3
The two failures run in opposite directions, so repairing one does nothing for the other. When Congress delegates too much, policy choices pass to a body that never had to win the approval of two Houses and the President, and the cure lies in the statute: a clearer standard, added procedures, or a different recipient. When Congress keeps control through a device such as a one-house veto, Congress itself skips those requirements, and the cure is to act as Article I prescribes. A statute can therefore pass the first inquiry and fail the second, or the reverse.
Interpretation comes first because the statute's meaning fixes what the constitutional test is applied to. A court cannot say whether a statute's instructions guide the delegate until it knows what the statute tells the delegate to do. Gundy v. United States illustrates the point. The Court read 34 U.S.C. § 20913(d), part of the Sex Offender Registration and Notification Act, to require the Attorney General to apply the Act's registration requirements as soon as feasible to offenders convicted before its enactment, and it held that this delegation “easily passes constitutional muster.”4 An answer that recites the intelligible principle test without first construing the statute skips the step that may decide the case.
Step 2 asks the central question, and steps 3 through 5 adjust how much a court will demand in answer. The constitutionality of a delegation of rulemaking or other discretionary authority may turn on its subject matter, the identity of its recipient, and the presence or absence of accountability mechanisms such as mandatory procedures or judicial review.5 Private recipients are suspect. They may be self-interested, and they are not subject to the presidential oversight that constrains agencies, which is why the Fifth Circuit has said the “private nondelegation doctrine” retains continuing force. Delegations to the President, the states, or tribal governments, who hold independent constitutional or sovereign authority, are less likely to be problematic.6 Subject matter can also favor the delegation: the President's powers in military matters and foreign affairs justified delegating the power to define court-martial offenses and to activate a statutory arms embargo.7
Step 6 is a question of statutory meaning, and it can be raised alongside step 1. It comes late on the list because steps 2 through 5 reveal how large the claimed power is. It also asks something different from the intelligible principle test. That test asks how much guidance Congress supplied. The major questions doctrine, discussed earlier in this chapter, asks how clearly Congress spoke about a power of vast economic and political significance. A statute can therefore satisfy the first inquiry and still fail to authorize the agency's action, so a complete answer runs both.
The two inquiries, and the independence of the delegation steps from step 7, show up when a single problem contains two statutes.
The first statute passes every delegation step and fails at step 7. It states a policy with boundaries, names a public official as recipient, and attaches notice, a hearing, and judicial review. The cancellation clause is the problem. A resolution of one House that changes the legal position of employers and workers outside Congress is an exercise of legislative power that bypasses the other House and the President, so it fails under Chadha however sound the delegation beside it.8 Whether the invalid clause drags down the rest of the statute is a severability question that this checklist does not answer.
The second statute fails for the opposite reason. Congress kept no control, so step 7 raises nothing, but the delegation lacks every feature that steps 2 through 5 look for. It gives the pen to an interested private group, and delegations to private entities are suspect because they may be self-interested and escape presidential oversight, as the cases the Fifth Circuit relies on, Carter Coal and Schechter Poultry, reflect.9 Its only standard is the Secretary's own belief that the rules will help the economy, which gives a court little to measure the Secretary's choices against, and it dispenses with the mandatory procedures and judicial review that can make a thin standard workable.10 Removing the veto from the first statute leaves the soundness of its delegation unchanged, and adding a veto to the second would only add a second defect to the first.
Step 7 reaches any device by which Congress or its agents change legal rights outside the process Article I prescribes. The legislative veto and the line-item veto are both invalid exercises of legislative authority that fail bicameralism or presentment.11 In Clinton v. City of New York, the Court declared the line-item veto unconstitutional because “[i]n both legal and practical effect, the President has amended two Acts of Congress by repealing a portion of each,” and repeal, like enactment, must comply with Article I procedures.12 A measure that clears both Houses and receives the President's signature meets the requirement Chadha announces, so a joint resolution that nullifies an agency rule on those terms is not a legislative veto in that sense.13 The devices discussed earlier under congressional agents as executors belong in this step as well. The question to put to each device is who must agree before legal rights change.
Calibrate the answer to how courts behave. The intelligible principle test is lenient, as the earlier discussion of modern leniency shows, and Gundy held a delegation to pass it easily.14 An answer that predicts a pure nondelegation holding against a modern statute with a stated purpose is therefore betting against the odds. Arguments with better prospects work through interpretation, because courts invoke the nondelegation doctrine to support narrow constructions of the scope of delegated statutory discretion.15 The major questions argument is the other strong option, since it demands clear authorization without invalidating anything. Pure nondelegation arguments are still made. In Jarkesy, the Fifth Circuit held that a delegation to the SEC of authority to choose between administrative and judicial enforcement of alleged securities fraud violated the doctrine because the statute lacked an intelligible principle to guide the choice, although the Supreme Court affirmed on other grounds.16 A stricter test to replace the intelligible principle standard is best presented as an argument about where the law may go, not as the rule a court applies today.
This checklist covers one slice of separation of powers. The principle does not merely govern relations between the two political branches, and the Court has relied on it to prevent congressional and executive encroachment on the judicial branch, so a statute that passes every step here may still raise a separate question about the courts.17