The who’s and what’s of presidential power

Major Questions is a recurring series by Adam White, which analyzes the court’s approach to administrative law, agencies, and the lower courts.

Please note that the views of outside contributors do not reflect the official opinions of SCOTUSblog or its staff.

The 2025-26 Supreme Court term’s first cases on presidential power are shaping up to be a study in contrasts.

Last month, when the court heard oral arguments in Learning Resources v. Trump, President Donald Trump’s sweeping assertion of the executive branch’s tariff powers was met with palpable judicial skepticism, and not just among the progressive justices. Some of the conservative justices, perhaps Neil Gorsuch most prominently among them, voiced serious concerns over the scale of presidential power at stake, and the migration of power from the legislative branch to the executive.

Yet in next week’s oral arguments in Trump v. Slaughter, on Trump’s refusal to heed statutory limits in firing two Democratic leaders of the Federal Trade Commission, the justices might seem much less skeptical of Trump’s assertion of power.

A majority of the court already told us so, more or less, in a May 22 order in a similar case on the firing of members of the National Labor Relations Board and Merit Systems Protection Board, who had been appointed by then-President Joe Biden. Staying the D.C. Circuit’s own order freezing those firings, the Supreme Court’s interim order explained that the government is “likely to show” that the NLRB exercises “considerable executive power,” and thus that the president has inherent constitutional power to remove the agency’s leaders regardless of laws that limit a president’s discretion to fire them at will. And the NLRB statute’s limits on a president firing NLRB members is strikingly similar to the statute that limits presidents in firing FTC commissioners.

In the months since that initial order in the NLRB case (Trump v. Wilcox) the Supreme Court has repeatedly stayed lower-court orders to block Trump’s firing of the heads of several multimember agencies – firings that did not attempt to comply with statutes purporting to limit the president’s removal powers. In July, the court stayed an order from the U.S. Court of Appeals for the 4th Circuit that had blocked Trump’s firing of three of the five members of the Consumer Product Safety Commission, noting that the case was “squarely controlled” by the court’s May order in the NLRB case. And two months later, in September, the justices issued a similar ruling (without opinion) staying the U.S. Court of Appeals for the D.C. Circuit’s order blocking Trump’s firing of the FTC commissioner.

Surveying the court’s recent approach to Trump’s agency firings, Justice Elena Kagan’s dissent blasted “the majority” for “hand[ing] full control of all those agencies to the President. He may now remove—so says the majority, though Congress said differently—any member he wishes, for any reason or no reason at all.”

Even if one disagrees (as I have) with Kagan’s view of the these cases and their relationship to the longstanding Humphrey’s Executor precedent (which I’ll explain below), her criticism raises a broader point that is well worth considering. Why would a justice who is keen to reinforce statutory limits on presidential power in the tariffs case not be equally inclined to reinforce statutory limits on presidential power in this other context?

The answer boils down to two kinds of cases on presidential power. I call them the “who” cases and the “what” cases. The “who” cases involve presidents asserting constitutional power to fire agency heads; the “what” cases involve presidents asserting statutory power to set policy. The Roberts court sees these two questions very differently – and this is not a recent development.

The “who” cases were the subject of my September and October columns, and what I have described above, which focus on debates around Trump’s claims of constitutional power to fire the heads of independent agencies despite their statutory protections. As I noted, the court’s recent interim orders on Trump’s firing of independent agency heads follow a long string of cases in which the Roberts court has broadly defended presidential power to do so. In cases on the Public Company Accounting Oversight Board (Free Enterprise Fund v. PCAOB), the Federal Housing Finance Agency (Collins v. Yellen), and, most notably, the Consumer Financial Protection Bureau (Seila Law v. CFPB), the court has repeatedly struck down statutes creating new forms of agency independence that, according to it, unconstitutionally constrained the president’s removal power for executive officers.

To be sure, the court has not completely renounced all forms of agency independence. In particular, it has stopped well short of overturning Humphrey’s Executor v. United States, the landmark 1935 ruling that Congress can give at least some special protection to agencies whose “duties are neither political nor executive, but predominantly quasi-judicial and quasi-legislative.” Specifically, Humphrey’s Executor affirmed Congress’ power to enact laws granting those officials a fixed term, and prohibiting presidents from firing them without specific grounds – say, “for cause,” as the Federal Reserve law vaguely puts it, or “for inefficiency, neglect of duty, or malfeasance in office,” as the FTC law elaborates.

Still, the court’s default constitutional rule (since 1926’s Myers v. United States) has been to recognize the president’s constitutional power to remove executive agency heads for any reason or no reason at all. And any exceptions to that rule have been increasingly limited by the decisions (discussed above) holding that Humphrey’s Executor does not apply to many of the “independent” agencies falling under the executive branch. This is a key theme of the Roberts court’s last fifteen years.

But this judicial sense of broad presidential power in the “who” cases differs completely from the court’s sense of presidential power in the “what” cases – that is, the cases on the substantive policymaking decisions by regulatory agencies that act at the president’s behest. On those cases, the main theme of recent years has been the Roberts court’s reduction of presidential power.

That part of the court’s recent history is exemplified by the string of cases in which the court rejected key policies of the given president. During the first Trump administration, for example, the court struck down two of his key policies – repealing an Obama immigration policy, and reinstituting the census’ citizenship question – because the administration failed to comply with the statutory requirements for agency policymaking. And in a long string of recent cases, from the 2021 decision striking down the Trump and Biden administrations’ COVID-era eviction moratorium, to the 2022 decision striking down the Biden administration’s Clean Power Plan, to the 2023 decision striking down the Biden administration’s student loan waiver program, the court repeatedly rejected administrations’ expansive views of their own regulatory powers under various federal laws.

This culminated with the court’s 2024 decision in Loper Bright Enterprises v. Raimondo, renouncing the decades-old doctrine of “Chevron deference,” a doctrine that has long embodied the court’s previous deference to the president’s particular constitutional role in directing agencies. As Justice Antonin Scalia had argued in a seminal 1989 essay, Chevron deference centered the president in regulatory policymaking: “Under our democratic system, policy judgments are not for the courts but for the political branches; Congress having left the policy question open, it must be answered by the Executive.” Accordingly, under Chevron, when a statute was found to be ambiguous, courts were to defer to the executive branch’s interpretation of it.

Now, when the court renounced Chevron deference in Loper Bright, it did not entirely eliminate the policymaking discretion that presidents enjoy under some statutes. But the decision’s overwhelming thrust was to shrink the president’s zone of regulatory discretion on countless subjects, by ordering the courts to stop deferring to an administration’s interpretations of the law.

In short, the court has steadily increased the presidency’s power over agency personnel, but at the same time it has steadily decreased the presidency’s power over agency policymaking.

One might see these disparate lines of cases and say … well, obviously. Or, to elaborate the point: Well, obviously – a president’s power to remove agency heads, and his control of substantive policies, are two different things.

And of course that’s true, in so far as it goes. But for decades, from the mid-1980s to the mid-2010s, Chevron deference and other related doctrines of judicial deference gave the president significant power to engage in substantive policymaking. The Roberts court’s abandonment of that doctrine was a significant reduction in presidential power, at the very same time that the court was increasing presidential power to remove agency heads. (And, notably, Chief Justice Roberts himself wrote the majority opinion in nearly all of the recent merits decisions that I’ve mentioned in this column, on both the president’s removal power and the president’s substantive policies – which is a point worth keeping in mind while we await decisions in the tariff and FTC cases.)

It’s commonplace, even among longtime court-watchers, to simplistically reduce the Roberts court’s decisions to a basic narrative about the court “bless[ing] broad executive power,” or the justices’ “deference to the president.” These often reflect a misunderstanding of the court’s “shadow docket” orders (a subject for another time) – but they also reflect an oversimplistic view of the court’s recent statements on the presidential removal power.

All of which is to say: The court has indeed increased presidential power. And the court has also decreased presidential power. Recognizing the “who-what” duality of the Roberts court’s cases is crucial, not just in the latest cases on the FTC commissioner and IEEPA tariffs, but also in the three years to come, when legal challenges to Trump’s substantive policies will far outnumber legal challenges to Trump’s personnel decisions.