Separation of Powers and Our Constitutional Freedoms

Harvard University Professor Cass R. Sunstein is one of the nation’s leading administrative and constitutional law professors, having written scores of books and articles on each topic. His most recent work—Separation of Powers: How to Preserve Liberty in Troubled Times—is a combination of the two types of publications. Like his earlier work, Sunstein’s newest book is a valuable contribution to the law’s treatment of the intersection of the two subjects he knows best. It contains something of benefit to everyone, from first-year law students to grizzled veterans of the post-New Deal wars over the role that the Fourth Branch of government should play in contemporary life.  

Separation of Powers opens with a novel approach to discussing our separation of powers doctrine. The traditional approach is to walk through the powers that the Framers vested in each branch, how they have been exercised, and how each branch tends to trespass on the authority given to the other two. Sunstein, however, flips that sequence on its head (or takes it in reverse). He assumes that readers are familiar with the orthodox, civics description of what each branch of the federal government may do and refocuses the reader’s attention on what they may not. Given the subtitle of his book, which identifies his book’s mission as a guide about how to “preserve liberty in troubled times,” Sunstein’s reorientation is a clever effort to put front and center his principal concern and remedy for those encroachments. 

That strategy is consistent with three of his subsidiary goals. One is to emphasize how the limitations found in the Constitution’s text, read wholistically, were intentionally formulated to empower each branch to govern the new republic in a cooperative manner, while also corralling any attempt by that newly created federal government to become just another overweening central government like the one that the colonists had shucked off in the Revolutionary War. Another goal is to describe, as any well-educated reader of the New York Times is now familiar with, how President Donald Trump has sought to strongarm or threaten each of the other two branches into doing whatever he would like them to do by firing executive branch officials who displease him, supporting challengers to incumbent members of Congress who do not back his play, and castigating members of the federal judiciary as being fifth columnists or traitors to his cause. The third goal is to make those first two points while simultaneously and studiously avoiding directly saying or clearly implying that the president is a “fascist,” a charge that has been periodically leveled against Trump. (Yet, opening his book with a discussion of how “Nazi legal theorist Carl Schmitt” saw “what happened in Germany on June 30, 1934,” the “Night of the Long Knives,” is a bit of a give-away. There also must be something in the water in Cambridge because he is not alone in finding worthwhile references to the views of a Nazi sympathizer. But I digress.)  

Governance without power, even if it only comes in the form of a bully pulpit, is little more than the advice-giving found in academic literature (like this book review). That is why the Framers assigned a “Power” to each branch of the new central government in Section 1 of Articles I, II, and III. The Articles of Confederation omitted distinct executive and judicial branches, creating only a legislative branch entrusted with limited authority to legislate and even less power to enforce its legislation or adjudicate disputes arising under it. Experience under that Parliament-lite structure had not gone well, however, because each “State” (no doubt a well-chosen term) saw itself as an independent, sovereign, juridical entity and was reluctant to abandon the self-governing authority that it had secured through force of arms. The Framers’ genius was to create a federal system that vertically separated governing authority between the national government and the states and horizontally separated the powers of the three branches of the central government.  

Until the Great Depression. That changed everything. 

The Depression hit the nation—and the world—like a George Foreman uppercut. Blaming then-President Herbert Hoover for its plight, the nation elected and thrice re-elected Franklin Roosevelt, who pursued more muscular versions of Hoover’s already-statist interventionist approach. Roosevelt’s New Deal, inspired by Progressivism’s mindset, used a host of new “expert” administrative agencies (e.g., Federal Deposit Insurance Corporation, Securities and Exchange Commission) to exercise substantive control over private enterprise in novel ways. The New Deal did not end the Depression—in fact, FDR’s policies prolonged it—but the New Deal agencies outlived the war. A second wave of government regulation, spurred by 1970s-era concerns over environmental harm and workplace safety, created additional agencies to regulate the economy, this time for non-economic purposes (e.g., the Environmental Protection Agency, the Occupational Safety and Health Administration).  

Yet, just as a pendulum spends half of its life swinging in the opposite direction, since Ronald Reagan became president in 1981, the nation has become increasingly fond of a more deregulatory mindset. Some agencies (e.g., Interstate Commerce Commission, Voice of America) have altogether disappeared, while others (e.g., Equal Employment Opportunity Commission) have had their wings clipped, in one way or another. 

The principal legal development has been the Supreme Court’s chipping away of its New Deal-era ruling in Humphrey’s Executor v. United States, which held that Congress may require the president to establish “cause” to fire a federal official. In decisions such as Free Enterprise Fund v. Public Company Accounting Oversight Board, Seila Law LLC v. Consumer Financial Protection Bureau, and Collins v. Yellin, the Court has consistently recognized that the ability to dismiss executive officials is an inherent component of the president’s Article II “executive Power,” and in Trump v. Slaughter, the Court will decide this Term whether to put Humphrey’s Executor out of its misery. In cases like Northern Pipeline Construction Company v. Marathon Pipe Line Company, Stern v. Marshall, and SEC v. Jarkesy, the Supreme Court has been increasingly protective of the federal courts’ constitutional adjudicatory role (and the civil jury’s related factfinding role) in resolving legal disputes instead of leaving that chore to administrative agencies. If the current trajectory of U.S. Supreme Court caselaw remains undisturbed, we might, as some have predicted, approach the classic separation of powers model that the Framers adopted, and that existed before the New Deal birthed what Justice Robert Jackson presciently described as “a veritable fourth branch of Government” that has “deranged our three-branch legal theories.” The Supreme Court has not (at least, not yet) resurrected the pre-New Deal Nondelegation Doctrine of Panama Refining Co. v. Ryan and A.L.A. Schechter Poultry v. United States, which barred Congress from directing federal agencies to “make America a better place” without much further guidance, but the Court’s new Major Questions Doctrine goes a long way in that direction (about which I will say more below). 

My favorite aspect of Sunstein‘s book is his discussion of what has come to be known as the Major Questions Doctrine. I particularly enjoyed his treatment of the “Tastes great! vs. Less filling!” debate between Justices Neil Gorsuch and Amy Coney Barrett over the sub-issue of whether that doctrine stems from separation of powers principles (the former) or is merely a canon of statutory interpretation (the latter). Justified either way, the doctrine most often comes into play when agency officials (greatly aided by computer-based research into potentially helpful statutory terms) read modestly worded statutes to grant them extraordinary power to regulate subjects that lie (at best) at the periphery of what a specific law could reasonably be said to reach. (Does the Center for Disease Control and Prevention have the authority to impose an eviction moratorium under a statute empowering the agency to “prevent the introduction, transmission, or spread of communicable diseases,” the issue that arose during the Covid-19 pandemic in Alabama Association of Realtors v. HHS?)  

The doctrine also can come into play when the government seeks to give a criminal statute a breadth exceeding the average, everyday understanding of its text, as illustrated by the dictionary meaning of its terms. (In an old case, is an “airplane” a “motor vehicle,” the issue that arose in McBoyle v. United States, a prosecution under the National Motor Vehicle Theft Act for stealing an aircraft?) While the Gorsuch-Barrett debate does have something of a purely academic air to it, since the two theories likely produce the same result more often than Ivory Soap is pure (“99 and 44/100%!”), Sunstein does the law professoriate proud by dissecting the two theories and placing them in the context of the broader debate over the proper construction of statutes implemented by the executive branch or used as the basis for a criminal charge. 

I have some small-scale disagreements with Sunstein. For example, at one point he concludes that “the separation of powers has nothing to do with checks and balances” and, instead, “is genuinely about separation as such.” That seems a little like saying that baseball has nothing to do with balls and strikes but is about the strike zone as such. But that is a minor point.  

I have a larger disagreement with Sunstein over his treatment of the Supreme Court’s decision in Trump v. United States. There, the Court granted the President immunity from criminal prosecution for any acts he (or she) takes within his exclusive constitutional authority (e.g., granting clemency) and at least presumptive immunity for all other official acts (e.g., entering into an executive agreement with a foreign head of state). Sunstein believes that the Trump decision is wrong and mangled the proper way to analyze an immunity issue. I agree with his bottom line, but not with his reasoning.  

Sunstein reasons that “the right question to ask” is whether criminal prosecution of a sitting president might “compromise his ability to perform his constitutional functions.” He implied that, perhaps in some circumstances, a reasonable person could answer that question in the affirmative. (That was the conclusion D.C. Circuit Judge Robert Bork reached, though Sunstein doesn’t mention him.) I disagree. The answer to that question is beside the point.  

Start with the constitutional text. It does explicitly address the question of which, if any, central government officials enjoy a constitutional immunity from criminal prosecution. The Speech or Debate Clause, Article I, § 6, cl. 1, grants that immunity only to “Senators and Representatives” and only for “any Speech or Debate in either House.” Clause 2 of that provision also makes it clear that no one may simultaneously be President and a member of Congress, so that possibility cannot arise. Plus, the Twenty-Fifth Amendment established a procedure that the Vice President and a majority of the cabinet can use to raise the issue of whether “the President is unable to discharge the powers and duties of his office,” a procedure that ultimately places that decision in the hands of Congress, which, by a two-thirds vote of both houses, can find that the President cannot for some reason, medical or otherwise, discharge the responsibilities of his office. The Constitution, therefore, leaves the issue of a potentially compromised President to the political process for it to make the necessary judgment in a particular case.  

Congress can also make that judgment across the board by passing a statute granting the President such immunity, or postponing any criminal prosecution until after he leaves office (which, of course, might only encourage a President to pardon himself, which I have argued he may). Every rationale that the Supreme Court majority found persuasive in the Trump case supplies a powerful rationale for granting a president such immunity by statute. Any such law would equally benefit Presidents belonging to either major political party, which in theory might reduce some of the political rancor this issue would raise. A permanent or temporary immunity also would be a sensible way to address this subject. But it is difficult to see how those policy-based arguments for compromise overcome the clear constitutional text. Its terms demonstrably show that the Framers granted a constitutionally based immunity to members of Congress only for their remarks in the Capitol, and to no one else for any other purpose. That should have been Q.E.D. for purposes of presidential constitutional immunity. 

I have no major criticism of Sunstein’s book, but I do have a major disappointment. An issue that has arisen since the Covid-19 pandemic swept the globe is the degree of deference that courts should give to an agency’s factual judgments and predictions, particularly those involving medical or scientific issues, especially those that are at the edge of emerging problems. The pandemic challenged the medical, scientific, and legal systems in ways that ordinary problems do not because we knew by January 2020 that the virus could prove fatal, but did not know how often that would be, how we could prevent or minimize that outcome, or what vaccines or pharmaceuticals would be effective responses to the disease. It is not an overstatement to say that, given those uncertainties, there was a certain feeling of panic in the air that seemed to grow over time. People generally trust the judgment of their own physicians, and they were willing to trust the judgment of the national government’s medical and scientific experts about a subject that, to the average person, did not appear to generate the type of political controversies that other types of health care issues, particularly funding proposals, inescapably pose. That was the initial public response. 

Over time, however, we learned that some of the proposals our public health officials offered, such as the six-foot rule, lacked an evidentiary basis. In other cases, officials varied their proposals, such as masking rules, without persuading the public that the government had newly discovered evidence to justify the change in course. Some governments drew distinctions between permissible and impermissible activities, such as California’s decision to allow cannabis retail stores to remain open while churches were nearly shut down, that could not be explained on any ground other than political favoritism.  

That distrust did not disappear after the pandemic was declared over. People no longer trust the judgment of our politically elected or appointed national public health officials. Such distrust is fatal to our willingness to respect their decisions.  

That sentiment runs headlong into the current legal doctrine requiring federal courts to be exceptionally deferential to factual or scientific judgments and predictions made by our public health officials. Consider the Supreme Court’s decision in Baltimore Gas & Electric Power Co. v. Natural Resources Defense Council (BG&E). That case involved the issue whether the Nuclear Regulatory Commission (NRC) may adopt generic rules governing the effect of the permanent storage of nuclear waste for purposes of all nuclear power plant licensing decisions. The Court unanimously answered that question in the affirmative, making clear that “a reviewing court must remember that the Commission is making predictions, within its area of special expertise, at the frontiers of science,” and that “[w]hen examining this kind of scientific determination, as opposed to simple findings of fact, a reviewing court must generally be at its most deferential.” The issues posed by the pandemic involved the same degree of complexity and uncertainty, and the predictions that spurred the government to issue orders as to what our response should be were no less “at the frontiers of science” than the ones that the NRC made in BG&E.  

Whatever standard of review the federal courts will apply to an agency’s scientific judgments is now of little concern to anyone but the courts. The public believes that politics is at the root of every such decision; that officials, whether political appointees or civil servants, who do not toe the line will be fired; and that self-preservation instead of science will be the force behind whatever career officials tell their politically appointed superiors.  

The upshot is this: Before and during the pandemic, the Supreme Court told the federal courts to respect the scientific expertise of ostensibly neutral public officials. But given what we have seen since then, why should agencies be trusted to make solid, science-based judgments when (1) the senior, politically appointed officials at those agencies know that angering the politically based judgments made by other politically appointed officials at the White House, let alone the President, will be fatal to their careers, and (2) federal civil servants now have the same fear. That is not to say that prior administrations did not follow the same course. Perhaps governments have always acted this way and the only novelty is the supreme audacity of the current chief executive’s willingness to say and do whatever he wants in the moment. Perhaps both major parties are guilty of having such leaders. But that just means the old saw “Two wrongs don’t make a right” applies here. 

So what should be the future of the BG&E rule granting deference to an agency’s scientific factual findings and predictions? And what would be the consequences of allowing the federal courts to review de novo those judgments in the same manner that the courts now may review an agency’s legal conclusions, given the Supreme Court’s 2024 decision in Loper Bright Enterprises v. Raimondo? It’s not that “Inquiring minds want to know”; the legal system and the nation need to know why courts should be the only ones to defer to the nominally scientific but essentially political judgments unelected officials make.  

Of course, no one book can address every issue. Maybe Professor Sunstein is working on that issue as I write. The academy, the profession, the courts, and the public would benefit from a thorough airing of this matter, and he could contribute to our understanding and resolution of it. 

Separation of Powers is a valuable addition to the literature discussing the application of the separation of powers doctrine at the intersection of constitutional and administrative law, a subject of ongoing controversy in the academy, the profession, the courts, and the society in which those institutions are found.  

Paul J. Larkin is a Senior Legal Research Fellow in the Meese Institute for the Rule of Law at Advancing American Freedom. I want to thank John G. Malcolm for helpful comments on an earlier iteration of this essay. Any mistakes are mine.