Yesterday, Chief Justice John Roberts delivered what conservative originalists have long been rooting for: overturning Humphrey’s Executor v. United States. That 90-year-old decision, which the Roberts Court has gradually been chipping away at, held that Congress could create independent commissions—such as the Federal Trade Commission—whose members are appointed by the president but protected from no-cause presidential removal. According to Roberts’s opinion in Trump v. Slaughter, this limitation on the president’s power is unconstitutional.It is worth revisiting Humphrey’s, which didn’t deserve the burial Roberts gave it. Imperfect in ways, it offered a better understanding of the separation of powers—one consistent with constitutional history, originalist understandings, and earlier Court precedent—than what Roberts offers in Slaughter.[Duncan Hosie: A false pretense of judicial modesty]Humphrey’s was a unanimous opinion written by Justice George Sutherland, no progressive squish. Sutherland was one of the so-called four horsemen of the Supreme Court, justices who often found New Deal legislation unconstitutional. Sutherland tended to couch his judicial opinions in historical and originalist terms: “The whole aim of construction, as applied to a provision of the Constitution, is to discover the meaning, to ascertain and give effect to the intent of its framers and the people who adopted it.” In Humphrey’s, he turned to founding-era history, including the writing of thinkers such as James Wilson, to understand the separation of powers and the contours of the president’s removal power.This is all to say that Humphrey’s was not a progressive opinion that played fast and loose with the text, structure, or history of the Constitution. In fact, at the time, progressives were often the ones advocating for sweeping presidential-removal power.Yet for Roberts, Humphrey’s is at odds with Chief Justice William Taft’s 1926 opinion Myers v. United States, which held that Congress could not limit presidential removal of executive officers. Taft’s opinion rested on the Decision of 1789, when Congress established the first three executive departments—State, War, and Treasury—and debated whether the principal officers (the secretary of foreign affairs, in this case) were removable at will by the president or whether Senate approval was required for removal just as it was for appointment. Although the constitutional text is silent on this question, Taft argued that history from the Decision of 1789 confirmed that executive officers were removable by the president alone. But four of the justices who signed Humphrey’s had also signed Myers, including Sutherland. They didn’t just suddenly forget what they’d written in Myers; Humphrey’s accepts that the Constitution bestows presidential-removal power of principal officers such as the secretary of state, it just holds that this power does not sweep so widely as to include at-will removal of all presidential appointees. This is a distinction that Roberts has now totally elided.At the core of Sutherland’s opinion in Humphrey’s is the insistence that independent agencies are ultimately exercising legislative power, so Congress can attach conditions on the removal of such administrators without encroaching on executive power. As Sutherland wrote, “The Federal Trade Commission is an administrative body created by Congress to carry into effect legislative policies” and thus could not “in any proper sense be characterized as an arm or an eye of the executive.”Sutherland’s logic was clear: The power to regulate trade (or interest rates, for that matter) clearly belongs to Congress, and when it tasks agencies with administering its powers, Congress has discretion on the contours of those agencies. In Humphrey’s, the Court saw a distinction between an administrative function (something the executive branch stewards) and true “executive power” in the constitutional sense. And this is precisely what distinguishes the constitutional question in Humphrey’s from the question in Myers. By contrast, Roberts’s opinion in Slaughter collapses this distinction. For Roberts, to faithfully execute the law, the president must have the power to remove officers of independent commissions. If the president cannot remove such officers, the president cannot act to carry out the law or be held accountable for his actions.The truth is, this is a difficult constitutional question without clear guidance from text. From the beginning, political actors and scholars have turned to history and theory to answer this question, but history does not provide clear answers; it only highlights that these constitutional questions have been the subject of politics from the beginning.In overruling Humphrey’s, the Roberts Court is solidifying a formalistic version of the separation of powers that narrows congressional power while expanding presidential power. Its logic points to the idea that all administration functions—not just independent agencies but possibly also the professional civil service—are an inherent part of the executive power and must be brought under the control of the president.[George Thomas: The other fear of the founders]Yet even Roberts cannot swallow the direction of his reasoning. That is why he exempted the Federal Reserve Board from the implications of his decision. In making this exception, Roberts notes that monetary policy at the core of the national bank’s mission is not subject “to plenary Presidential control.” True. But this is just as true of the regulation of commerce or trade, so why doesn’t this logic apply to the Federal Trade Commission?Humphrey’s was a sensible understanding of the separation of powers. It was certainly preferable to Roberts’s reasoning in Trump v. Slaughter, which threatens to turn the administration of policy into the president’s personal power. The Roberts Court likes to think of itself as originalist, but it is imposing a distinctly modern and stilted understanding of the separation of powers onto the country, while denying Congress power that it has exercised since the founding. That’s not originalism, no matter what the Court says.