The Supreme Court now has a job to do: Enforce the Constitution’s bar on congressional delegation, whether to agencies or the president. The Court will have plenty of opportunities to do this. For one, on April 4, 2025, two days after Trump announced his sweeping “reciprocal” tariffs, a small Florida stationary importer called Emily Ley Paper sued him over tariffs he imposed in February and March. The firm’s complaint, filed by lawyers at the New Civil Liberties Alliance (full disclosure: I am a member of its board of advisers), charges that his action—ostensibly allowed under the International Emergency Economic Powers Act—unconstitutionally infringed on Congress’s power to tax. Many other suits against his tariffs and other exercises of delegated power have followed.

In this article, I will show how the Court can do this job and thereby force Congress to do its job in making the laws.

Why the Constitution Bars Congress from Delegating Its Power to Make Law

As Alexander Hamilton wrote in Federalist No. 75, “The essence of the legislative authority is to enact laws, or, in other words, to prescribe rules for the regulation of the society.” So, the Constitution gives Congress the power to impose tariffs, other taxes, regulations, and other rules of private conduct. For simplicity, this article calls these “laws.”

Congress must enact laws through the Constitution’s legislative process. For the Constitution to allow Congress to delegate the power to impose tariffs to other branches of the government, for example, would violate the patriots’ vow of “no taxation without representation.” To see why, consider a statute that empowers the president to impose tariffs to protect domestic producers from unfair competition from abroad. The legislators could take credit for providing protection but shift blame to the president for the increased cost of imports and any unfair competition not halted. As David Mayhew in his 1974 book Congress and other political scientists have shown, delegation lets legislators claim credit for popular promises and shift responsibility for unpopular results.

Such responsibility is necessary to anchor taxation, including tariffs, on representation of the people. Under the original Constitution, the members of the House of Representatives were the only officials elected by the people themselves. To keep these representatives closely accountable to the people, the Constitution subjected them to judgment by voters at the polls every two years. In contrast, the Constitution originally let states’ legislatures choose how to elect their US senators and members of the Electoral College, who select the president. But now Trump—who, being twice elected, cannot run for reelection—has unilaterally imposed tariffs comprising one of the largest tax increases in US history.

Delegation of the power to make rules governing private conduct would conflict not only with the “consent of the governed,” as the Declaration of Independence promised, but also with the promotion of “domestic Tranquility,” as the Constitution sought. When legislators bear responsibility for a law, they must debate its results with each other and discuss them with constituents. If it is a tariff rule, the legislators would necessarily focus on its effects on prices and domestic industry. The legislative process thus helps the nation look before it leaps. In contrast, President Trump’s tariffs were based upon assumptions about their economic effects that were soon proven wrong and thereby forced him to change course several times in just weeks. That is why we have been on a tariff roller coaster.

The debate in Congress also would help voters understand the stakes and the feelings of the people on the other side. Moreover, it is far easier to clearly state a compromise on a tariff rate than on how to balance such abstract goals as protecting domestic industries from unfair competition and protecting consumers from startling price increases. As Brook Manville and Josiah Manville find in their 2023 book The Civic Bargain, history shows that in a democracy, domestic tranquility requires the ability to compromise.

While requiring Congress itself to make the law—that is, the rules regulating society—the Constitution does empower the executive to enforce the law and the courts to interpret it and apply it in cases. The Constitution also gives the executive discretionary powers in matters other than lawmaking, such as relations with other nations and the management of the armed forces and other parts of the executive branch.

Drifting into Delegation

During the Constitution’s first century, Congress aimed to enact all the laws, but then gradually began to feel free to delegate. One justification for shifting to delegation was the hope that the agencies or the president would do a better job for the people because they have staffs of experts. Yet, experts are available to Congress, too. Another justification was that bureaucrats supposedly were removed from political pressure, but it turns out that agencies and the president are not so immune. Beginning more than a half century ago, Ralph Nader and his “Raiders” showed that delegation to agencies allowed elected officials to privately pressure the agencies to go easy on campaign contributors but blame the agency for the resulting harm to the public. Now, despite the experts on his staff, President Trump flubbed in predicting the consequences of the tariffs he announced on April 2.

The Courts and Delegation

The week before Trump announced his “reciprocal tariffs,” the Supreme Court heard arguments in Federal Communications Commission v. Consumers’ Research. The case concerns the FCC’s authority to collect a tax on telecommunications service and use it to subsidize access for certain customers across the country. Consumers’ Research argues that Congress unconstitutionally delegated its legislative power to create the tax to the FCC (and the FCC then delegated authority to set the tax to a private entity, the Universal Service Administrative Company).

One might think that the Court would easily find the statute to be an unconstitutional delegation of legislative power. In recent years, five current justices signaled their desire to limit delegation of the power to make law. Justice Neil Gorsuch did so in his dissenting opinion in Gundy v. United States (2019), in which Chief Justice John Roberts and Justice Clarence Thomas joined. A fourth justice, Samuel Alito, stated in a concurring opinion that he would join them if a majority of the justices agreed. Shortly thereafter, Brett Kavanaugh became a justice and signaled his agreement, thus making a majority.

Yet, the Court did not return to the delegation issue until FCC v. Consumers’ Research. Even more strangely, Consumers’ Research argued for a weak test of unconstitutional delegation: that the delegation is unconstitutional only if the statute fails to guide the agency in making laws. The Supreme Court adopted such a test after President Franklin Roosevelt blasted the justices for striking down New Deal programs, including in two delegation cases, Schechter Poultry v. US (1935) and Panama Refining Co. v. Ryan (1935), in which the Court ruled by margins of 9–0 and 8–1 respectively. Roosevelt infamously responded with his court-packing threat and, though Congress rejected the idea, the Court—fearful of losing its independence, as the story goes—denied subsequent delegation challenges provided that the statute “intelligibly” guided officials by stating goals they must pursue in making laws. The goals, however, often provided only fuzzy, feel-good guidance. This, the Court misleadingly stated, is the “intelligible principle” test.

The Unintelligible Intelligible Principle Test

Antonin Scalia, writing in Regulation in 1980 when he was a University of Chicago law professor and not yet on the bench, explained that the intelligible principle test is inherently toothless:

The relevant factors are simply too multifarious: How significant is the power in question…? How technical are the judgments left for executive determination…? What degree of social consensus exists with respect to those nontechnical judgments committed to the executive…? And most imponderable of all—how great is the need for immediate action…?

The article did not suggest a better test, but Scalia nonetheless urged the Court to limit delegation of legislative power. He argued:

The unconstitutional delegation doctrine, far from permitting an increase in judicial power, actually reduces it. For now that judicial review of agency action is virtually routine, it is the courts, rather than the agencies, that can ultimately determine the content of standardless legislation. In other words, to a large extent judicial invocation of the unconstitutional delegation doctrine is a self-denying ordinance—forbidding the transfer of legislative power not to the agencies, but to the courts themselves.

Scalia’s argument makes sense as far as it goes, but it stops short of dealing with a concern that could give justices heartburn. Striking down delegations without a clear standard of how little guidance to the agency is too little could well spark new attacks on the Court for making policy choices rather than enforcing a constitutional rule. Indeed, a few years after becoming a justice, Scalia wrote in Mistretta v. United States (1989):

I fully agree with the Court’s rejection of petitioner’s contention that the doctrine of unconstitutional delegation of legislative authority has been violated because of the lack of intelligible, congressionally prescribed standards to guide the Commission…. [T]he scope of delegation is largely uncontrollable by the courts.

The difficulty of explaining how little guidance is enough to justify striking down a delegation was on display in FCC v. Consumers’ Research. The parties attacking the delegation tried to skirt the unintelligibility of the intelligible principle test by using another label, “clearly delineate,” to describe the required guidance. Yet, the change in label does nothing to deal with the underlying problem: There is no ready way to state how clear is “clearly” enough, just as there is no ready way to state how intelligible is “intelligible” enough. During oral arguments, the justices made mincemeat of the Consumers’ Research argument that the statute failed to provide sufficient guidance because it stated no upper limit on the size of the tax that could be imposed by delegated authority. For example, Justice Amy Barrett asked whether a sky-high upper limit of “3 trillion [dollars], 3 billion, whatever, … would solve the problem.” The response she got was “Absolutely.” If so, that would produce a win for the opponents of the delegation in the case, but it would erect a limit on delegation that Congress could easily evade by stating ridiculously loose limits on the agency’s power. That would be silly.

The Court did, however, react to Scalia’s concern that uncontrolled delegation gives the courts vast power to make policy by interpreting statutes. In 1984, before Scalia became a justice, the Court held in Chevron v. Natural Resources Defense Council that courts should defer to agencies’ interpretations of statutes. Chevron, however, allowed agencies to broaden their own powers through interpretation. Last year, the Court overruled Chevron in Loper Bright Enterprises v. Raimondo.

Meanwhile, the Court has embarked on another effort to limit agencies’ ability to grow their power, holding that they should not be able to tackle “major questions” without explicit statutory authorization. For example, in West Virginia v. Environmental Protection Agency (2022), the Court ruled the agency could not deploy the Clean Air Act in a novel way to cut greenhouse gas emissions without Congress clearly giving it that power. The major questions doctrine does, however, raise the question of how courts can have any more success in explaining how major a question must be to be deemed a “major question” than courts have had in explaining how intelligible a principle must be to be deemed an “intelligible principle.” In any event, even if the major questions doctrine does require Congress to make clear it wishes to delegate power over a major issue, it does nothing to require Congress to take responsibility for the answer to the question, such as how big a cut in emissions there should be and at what cost to consumers.

A Judicially Manageable Test

There is, however, hiding in plain sight a judicially manageable test. It is the test in the Constitution itself: Congress must enact the rules of private conduct. That is, the lawmakers in Congress must themselves make the law.

This test rides on a difference of kind: whether Congress has stated a rule rather than stated enough to guide an agency in making a rule. In a 1985 article and in my 1993 book Power Without Responsibility, I argued that the original meaning of the Constitution requires Congress to enact the rules, and this provides a judicially manageable test. Justice Gorsuch cited my article and book in his opinion in Gundy.

Under this test, Congress cannot empower the president or an agency to impose tariffs, even if the statute states the goals that the tariffs should pursue. Nor can Congress mimic a rule but fail to state one by, for example, passing a statute stating that importers must pay whatever tariff an agency determines would strike a sensible balance between various goals. In contrast, Congress can enact a law establishing a tariff on imported crude oil set at, say, 10 percent of the price of domestic crude in the open market at the port of entry. Such a law may still require interpretation by agencies and the courts, such as deciding precisely how to determine that price. Nonetheless, Congress would still be accountable for the law’s results, popular and unpopular.

Is This Test Legislatively Workable?

The reason that the parties that opposed delegation in FCC v. Consumers’ Research did not rely upon the Constitution’s original test is, I suspect, fear that the Court would think there are too many federal rules needed in modern times for Congress to make all of them.

I argued in Power Without Responsibility that Congress could have made all the federal rules if it had left more purely local matters to state and local government and, in matters that must be regulated nationally, it had used simpler forms of regulation. An example of this is the cap-and-trade approach to controlling acid rain that Congress enacted in the 1990 Amendments to the Clean Air Act. To impose a declining cap on total power plant emissions of pollutants that make acid rain, the statute used a formula to allocate among the plants the right to emit these pollutants such that the total of the rights would equal the cap and allowed the emitters to buy and sell these rights. This gave all sources a profit-based motive to figure out how to cut such emissions. Cap-and-trade, unlike most of the rest of the Clean Air Act, delivered the pollution control promised at a cost lower than was predicted. Presidential candidate Barack Obama praised cap-and-trade as being better for the country than the ordinary command-and-control method of regulating pollution because it gave people more clean air bang for the buck. Moreover, with cap-and-trade, one rule cut emissions from all power plants, whereas delegating to the agency the job of determining how much each plant could cut emissions could have required hundreds of rules.

Congress has not enacted cap-and-trade to deal with other pollutants because that would expose its members to blame for the pollution control costs. In a 2009 Wall Street Journal op-ed, former chair of the Environmental Defense Fund Richard Stewart and I argued that the 2009 Waxman–Markey climate change bill that President Obama supported masqueraded as a cap-and-trade bill, but it was really a “bait and switch” that shifted the hard choices on emissions control to unelected officials. (The bill passed the House but failed in the Senate.) Also, in a 2013 article, former Environmental Protection Agency deputy general counsel Bill Pedersen and I argued that Congress’s failure to enact a cap-and-trade program to cut fine particulates and instead delegating the tough decisions to the EPA has shortened the life expectancy of the average American by months.

Not long after Power Without Responsibility was published, some members of Congress asked me to design a bill to stop Congress from passing the buck on regulation. I based my suggestion on an idea that James Landis, the New Deal’s guru of regulation, floated in his 1938 book The Administrative Process: Congress should vote on agency actions. I also borrowed from Justice Stephen Breyer’s 1984 Georgetown Law Journal article showing how Congress could craft a statute that would force both houses of Congress to vote on agency actions by a deadline and thereby circumvent filibusters.

The result of these efforts was a bipartisan bill dubbed the Congressional Responsibility Act of 1995. It gained traction at first, but it ultimately foundered on two obstacles: First, many legislators did not want to shoulder the responsibility of voting on these laws. Second, there were simply too many regulations for Congress to vote on them all. My argument in Power Without Responsibility that there would not be too many rules if Congress had never begun to delegate overlooked the reality that it had delegated for generations. That has made American society, especially its economy, rely upon a regulatory system based upon delegation of lawmaking power that calls for a huge number of new regulatory rules every year. Requiring legislators to suddenly take responsibility for all of them would likely cause widespread disruption. We are path dependent.

Instead, Congress enacted the Congressional Review Act of 1996. It dealt with these two objections by giving legislators the option of voting on recently promulgated laws. Congress rarely does so because, as noted, its members do not want the responsibility and, in any event, presidents would likely veto rejections of laws promulgated during their administrations. The statute thus gave the legislators who voted for it credit for wanting to take responsibility without them actually taking any. President Clinton signed it. In contrast, the Congressional Responsibility Act would force legislators to vote and presidents would generally not veto rules designed by their appointees.

In response to the limitations of the Congressional Review Act, some legislators have, since 2009, introduced a bill, dubbed the Regulations from the Executive in Need of Scrutiny (REINS) Act, that would greatly expand Congress’s power and responsibility in overseeing delegations. The legislation’s most important provision would require Congress to approve “major rules” drafted by the agencies before the rules can take effect, with an exception for presidentially declared emergencies or other urgent need, in which case the rule would still require congressional approval within 90 days of implementation.

REINS is dressed up to look anti-regulatory, given the anti-regulatory rhetoric of its sponsors and the bill’s title that identifies the executive branch—not the legislative—as being in need of scrutiny. The problem is, a bill dressed up to look anti-regulatory is likely to put off almost all Democrats and some Republicans from swing states, and so will never get past a Senate filibuster.

I was asked to testify in hearings on REINS. I said the bill was designed to let its sponsors take a pro-responsibility stance without actually having to take responsibility. I have not been asked to testify since.

How Congress Could Shoulder Responsibility

A new version of the Congressional Responsibility Act bill that exempts less important rules from the requirement that Congress approve them would provide a workable way for Congress to begin to take more responsibility. Unlike the Congressional Review Act, it would force Congress to vote on rules. Indeed, it would use Justice Breyer’s design to prevent legislators from using the filibuster to avoid taking responsibility. Like REINS, it would exempt less important rules. Unlike REINS, it would be stripped of anti-regulatory rhetoric (and thus off-putting costuming). Also unlike REINS, the definition of rules upon which Congress must vote would include not only those that would be very costly to the economy but also those that would be very protective of the public. This is what I recommended in my 2017 book D.C. Confidential. I suggest that Congress should, at least for starters, focus on new rules that would have a major impact.

To do so, Congress could enact a statute that

• defines a major rule,
• requires agencies to propose new major rules to Congress rather than promulgate them,
• establishes a process that would force the House and Senate to vote up or down on each such rule by a short deadline, and
• then, if the rule passes both houses, presents it to the president for signature or veto.

The statute could also allow for a law to go into immediate effect if the president declares an emergency but cease to have effect if the House or Senate reject it or both houses of Congress have not approved it by a statutory deadline.

This process would make use of agencies’ expertise but require members of Congress and the president to be accountable for the new major laws. Meanwhile, it would be in the agencies’ interest to write such law in ways that could get enacted by these elected officials.

How the Court Can Do Its Job—and Make Congress Do Its

The Supreme Court cannot order Congress to pass such a statute. The justices cannot tell legislators how to do their job (no matter how badly Congress needs to hear that). Nonetheless, the Court must do its job, which includes deciding that Congress has violated the Constitution when a case presents that issue.

In such a case, the justices’ job requires them to overrule the intelligible principle test if the Court’s three requirements for overruling precedent can be met. One requirement is that the precedent be mistaken, and that is clearly the case with the “intelligible principle” test. The second is that the mistake does harm. That is also the case for the intelligible principle test: Congress has used it to take credit for popular promises and shift blame for unpopular results. The result is that legislators work to make themselves look good to voters rather than do good for voters. That is neither “government by the people” nor “for the people.”

The final requirement is that, in overruling a prior case, the Court not unduly upset “reliance interests.” In reliance on the intelligible principle test, Congress has structured most of our statutes dealing with regulation and taxation. The Court could accommodate this reliance by stating that, in other cases, it will apply the correct test only to laws promulgated after a certain date. The Court should set that date to give Congress time to prepare to do its lawmaking job. In other words, the new test would have delayed prospective effect.

The prospective effect of the new test would mean courts would treat rules promulgated before the designated date one way and those promulgated afterward another way. The reason is that the Court, in an exercise of equitable discretion, would seek to avoid undue hardship on the public from the delay needed for Congress to adapt to complying with the Constitution rather than that the Constitution would have different meanings before and after the designated date.

Given also that reliance on delegation has resulted in a legal system with a very large numbers of rules, the Court should accommodate this reliance by stating it will not apply the new test to less important laws if Congress decides to vote only on the more important ones. To take advantage of this leeway, however, Congress would need to adopt a method of identifying the more important rules based upon their effects on society rather than shielding elected officials from hard choices.

Some legislators would want to define these more important rules and structure agency rulemaking in ways that would shield themselves from responsibility. That would not be easy. Existing statutes and agency practices often dictate the scope of new rules so that changes that made previously major laws into many minor ones would be evident. Meanwhile, polling shows that voters believe by overwhelming margins that legislators should enact the regulatory rules. Legislators defining major actions in a way that would let them duck responsibility for important issues would be highly visible, especially after the Court weighed in. All this would give lawmakers responsibility for ducking responsibility.

Another exception could allow new laws to have brief temporary effect in an emergency. Longstanding practices of equitable remedies suggest the Court would tolerate the emergency exception.

In sum, both Congress and the Court need to do their jobs. That would ensure we have a government based upon the consent of the governed.

Readings

• Breyer, Stephen, 1984, “The Legislative Veto after Chadha,” Georgetown Law Journal 72: 785.
• Landis, James M., 1938, The Administrative Process, Yale University Press.
• Manville, Brook, and Josiah Manville, 2023, The Civic Bargain: How Democracy Survives, Princeton University Press.
• Mayhew, David, 1974, Congress: The Electoral Connection, Yale University Press.
• Scalia, Antonin, 1980, “A Note on the Benzene Case,” Regulation 4(4): 25–28.
• Schoenbrod, David, 1985, “The Delegation Doctrine: Could the Court Give It Substance?” Michigan Law Review 83(5): 1223–1290.
• Schoenbrod, David, 1993, Power Without Responsibility: How Congress Abuses the People through Delegation, Yale University Press.
• Schoenbrod, David, 2017, DC Confidential: Inside the Five Tricks of Washington, Encounter Books.
• Schoenbrod, David, and William F. Pedersen, 2013, “The Overwhelming Case for Clean Air Act Reform,” Environmental Law Reporter 43(10971).
• Schoenbrod, David, and Richard B. Stewart, 2009, “The Cap-and-Trade Bait and Switch,” Wall Street Journal, August 24.