The Republican justices are fighting over who should really run the government

The most important thing to understand about President Donald Trump’s tariffs, at least as a political matter, is that tariffs are an issue that largely unites Democrats and that divides Republicans.

Shortly after the Supreme Court struck down many of Trump’s tariffs, former Senate Republican Leader Mitch McConnell put out a statement praising the decision. And, of course, the Court’s Republican majority split 3-3 on whether to allow the tariffs, while the three Democratic justices joined together to oppose them.

Indeed, the Court’s decision in Learning Resources v. Trump (2026), the tariffs case, revealed philosophical divides among the Republican justices that go far beyond their stances on Trump’s trade policy.

While the three dissenters — Justices Clarence Thomas, Samuel Alito, and Brett Kavanaugh — appear to believe that the question of how much power should be wielded by the president depends on which party controls the White House, Justice Neil Gorsuch took a more principled (if also quite extreme) stance against allowing any president to wield too much authority over federal policy. Justice Amy Coney Barrett, meanwhile, laid out a more modest framework than Gorsuch’s, which potentially gives Congress a greater role in shaping how the government should function.

In fairness, it’s far from clear just how much the theoretical disagreement between Gorsuch and Barrett actually matters, as the two justices vote together nearly all the time on cases involving questions of executive power. Both of the two Republicans joined the Court’s 2024 decision holding that Trump may use the powers of the presidency to commit crimes, although Barrett did disagree with some minor sections of the majority opinion. And both supported Biden-era decisions limiting a Democratic president’s power to govern.

Still, the two justices’ competing opinions in Learning Resources do reveal fundamentally different ideas about the role of the judiciary. Gorsuch is among the Court’s most outspoken judicial supremacists, and his opinions suggest that his Court should invalidate many federal policies even when those policies are authorized by an act of Congress. Barrett, by contrast, suggests that her Court should take a more humble approach when the two elected branches do not share the justices’ preferences — even as she also concludes that Trump’ s tariffs went too far.

Though this disagreement did not shape the outcome of Learning Resources, it could matter a great deal in a future Democratic presidency, if Congress authorizes that president to act in a way that Republicans find objectionable. Gorsuch’s opinion indicates that he will likely strike down such an act of Congress — or, at least, to read it narrowly enough to prevent the president from doing as he or she pleases. Barrett’s opinion, meanwhile, suggests that she might side with Democrats in such a case, at least if Congress leaves her with no other reasonable option.

The Constitution gives Congress, and not the president, the power to “lay and collect taxes, duties, imposts, and excises.” The question in Learning Resources was whether Congress, in a 1977 law giving the president some authority to “regulate…importation or exportation,” delegated to the president its power to decide how much imported goods should be taxed.

Six justices, including Gorsuch and Barrett, joined a majority opinion arguing that this 1977 law does not permit Trump to levy tariffs because the word “regulate” does not include the power to tax (although, oddly enough, Gorsuch’s concurring opinion includes a paragraph suggesting that the power to “regulate” includes the power to levy tariffs, so it’s not entirely clear that he agrees with the majority opinion that he joined).

The primary disagreement between Gorsuch and Barrett hinges on what should have happened if Congress had written a law that actually did give Trump the power to levy tariffs. Suppose that the law had used the verb “tax” as well as “regulate.” Gorsuch believes that the Constitution limits Congress’s power to delegate authority to the president, and thus, such a law would likely be unconstitutional. Barrett, on the other hand, would likely permit Congress to delegate this power to the president if it did so using sufficiently explicit language.

Gorsuch, for his part, seeks to relitigate a fight that his side lost nearly a century ago. The idea that the Constitution limits Congress’s power to delegate policymaking authority to the executive is hardly a new idea. It even has a name: the “nondelegation doctrine.”

The nondelegation doctrine had a brief moment in the sun in 1935, when the Court struck down two federal laws that gave President Franklin D. Roosevelt an absolutely extraordinary amount of power. One statute, for example, gave Roosevelt nearly limitless authority to “impose such conditions” on private businesses “as the President in his discretion deems necessary to effectuate the policy” declared in the law.

But, outside of these two decisions in 1935, the Supreme Court has never invoked the nondelegation doctrine to strike down any law. Modern decisions permit Congress to delegate authority to the executive so long as it “lays down by legislative act an intelligible principle” that executive officials may use to determine how they are supposed to exercise their authority. And even Gorsuch concedes in his Learning Resources opinion that the nondelegation doctrine was abandoned a long time ago. Though, he also claims that, in doing so, “the Court pushed aside its long-held skepticism of claims to extraordinary delegated powers and began affirmatively encouraging them.”

So, the nondelegation doctrine is out of step with the Court’s current precedents. And the constitutional argument against those precedents is extraordinarily weak. In a 2020 paper criticizing the doctrine, law professors Julian Davis Mortenson and Nicholas Bagley point out that the First Congress enacted — and President George Washington signed — numerous statutes that made sweeping delegations of power to the executive branch, including a law delegating the entirety of Congress’s power over patents to members of Washington’s cabinet.

Nevertheless, Gorsuch is probably the Court’s most vocal proponent of reviving nondelegation, and, for a brief moment in 2019, it appeared that he had five votes to do so. Ultimately, however, the Court’s Republican majority embraced a different framework, known as the “major questions doctrine,” which allows the Court to veto executive branch actions, but that does not claim that Congress may never permit the president to set ambitious policies.

It’s not at all clear, however, where this major questions doctrine comes from, and the Court has never issued a majority opinion explaining which statute or constitutional provision authorizes the justices to block a federal policy because they deem it to be too economically or politically significant. In Learning Resources and in a handful of previous cases, however, Gorsuch and Barrett laid out competing theories that, they claim, justify the Court’s decision to create this novel doctrine.

Gorsuch’s view, which he first articulated in his concurring opinion in NFIB v. OSHA (2022), is that the major questions doctrine can be found within penumbras and emanations emerging out of the nondelegations doctrine. Gorsuch even claimed, falsely, that “for decades courts have cited the nondelegation doctrine as a reason to apply the major questions doctrine.” Similarly, in Learning Resources, Gorsuch describes the major questions doctrine as a constitutional limit on Congress’s power to give authority to the president. “The Constitution,” he claimed, “lodges the Nation’s lawmaking powers in Congress alone, and the major questions doctrine safeguards that assignment against executive encroachment.”

Barrett’s approach, meanwhile, is far less legalistic. Concurring in Biden v. Nebraska (2023), Barrett argued that the doctrine is implicit in a parable involving a babysitter.

Barrett imagined a parent who hires a babysitter to take care of her children over a weekend and who “hands the babysitter her credit card and says: ‘Make sure the kids have fun.’” She then imagines that this babysitter uses the card to pay for a trip to “an amusement park, where they spend two days on rollercoasters and one night in a hotel.” According to Barrett, while the babysitter’s decision to take the kids to Disney World may have complied with the parent’s instructions in the “literal” sense, they were not a “reasonable” understanding of what the parent wanted.

There are many problems with this analogy. Among other things, when an employee follows an employer’s instructions, but does so in a foolish or wasteful way, the employer’s recourse is typically to fire that employee, not to sue them. Some states even have laws which outright forbid employers from suing their workers under circumstances like the one described in Barrett’s Parable of the Babysitter.

Similarly, if Congress passes a law giving power to a president or cabinet secretary, and that official uses that power irresponsibly, the recourse should be for Congress to repeal the law or to remove the official via impeachment, not to file a federal lawsuit seeking to nullify the official’s action.

Either way, Barrett’s version of the major questions doctrine is more modest than Gorsuch’s. Barrett does not claim that a parent could never give a babysitter explicit instructions to take their child to Disney World. And she similarly does not argue that Congress could not write a law which very clearly gives the president the power to levy tariffs.

Barrett does appear to believe that, if a federal statute can be read two different ways, then the Court should prefer the interpretation that would prevent a presidential administration from doing anything too ambitious. But she also appears to concede that presidents may make very significant policy decisions when Congress clearly gives them the power to do so. As Barrett writes in her Learning Resources opinion, “if the Constitution permits Congress to give the Executive a particular power, who are we to get in the way?”

Ultimately, in other words, the disagreement between Gorsuch and Barrett comes down to a dispute over which branch has the ultimate authority to determine what the president can and cannot do. Gorsuch’s position is that the Supreme Court is the ultimate authority, and that his Court should simply strike down acts of Congress that give the president more power than he wants the president to have. Barrett, on the other hand, appears to believe that Congress may write laws giving the president powers that she is personally uncomfortable with, even though Congress must do so fairly explicitly.

If you are wondering how this disagreement could matter, consider this hypothetical: In West Virginia v. EPA (2022), the Court’s Republican majority struck down EPA regulations seeking to reduce greenhouse emissions by power plants, claiming that those regulations violated the major questions doctrine. Suppose that, in 2029, a Democratic Congress enacts a new law which provides that “The Environmental Protection Agency shall promulgate regulations governing power plant emissions that ensure that each plant uses the best system of emission reduction, which may include the regulations that the Supreme Court rejected in West Virginia v. EPA.”

If a Democratic administration then relied on this statute to reinstate the exact same regulations that the Court repudiated in West Virginia, Gorsuch likely would strike them down again as unconstitutional. But what about Barrett? The approach Barrett laid out in Nebraska and Learning Resources suggests that she might defer to Congress if it wrote a statute that so explicitly authorized the same regulations repudiated by a previous Supreme Court decision.

That said, Barrett’s past behavior calls into question whether she would actually permit Congress to delegate such authority to a Democratic president, as the federal policy she voted to strike down in Nebraska was very clearly authorized by Congress.

That case involved the Biden administration’s attempt to cancel many billions of dollars in student loans. To justify this policy, Biden’s Education Department relied on a broadly worded statute permitting the Secretary of Education to “waive or modify any statutory or regulatory provision applicable to the student financial assistance programs...as the Secretary deems necessary in connection with a war or other military operation or national emergency” such as the Covid-19 pandemic.

This statute, moreover, was full of language signaling that Congress intended to delegate to the secretary complete discretion to alter borrowers’ student loan obligations. Among other things, it contained a provision stating that the secretary’s power remains intact “notwithstanding any other provision of law, unless enacted with specific reference to” the law permitting the secretary to waive or modify student loan obligations.

And yet, despite this very explicit language, all six of the Court’s Republicans voted to block Biden’s student loan forgiveness program, including Barrett.

So, while some of the Republican justices have, at times, endorsed a relatively modest version of the major questions doctrine, none of them have displayed such modesty in an actual case. The practical holding of Nebraska is that the Supreme Court may veto any policy promulgated by an executive branch official, even when that policy is clearly authorized by federal law.