What is the relationship between the person of the president and the office of the presidency? That question lies at the crux of Trump v. Hawaii, though it remains largely unaddressed in the Court’s opinions. “[W]e must consider not only the statements of a particular President,” Chief Justice Roberts writes for the majority, “but also the authority of the Presidency itself.” But what happens to the authority of the presidency when its current occupant is at war with the norms and institutions of the office? The question is likely to recur, if not define constitutional adjudication in the age of Trump.

In a just-published article, I document those foundational norms and institutions and explore their implications for judicial deference. The “office” of the presidency is comprised of certain features—deliberative practices, substantive commitments, and institutional constraints—that are not specified in the Constitution or statutory text. Rather, the nature of the presidency is determined by norms of constitutional governance that past presidents adopted and that other actors in our constitutional system, over time, helped to entrench. A central claim I make is that the rise of presidential power in our time is in part a result of these norms of limitation: “A President armed with nuclear capabilities, overseeing a sprawling criminal code and a sweeping domestic administrative establishment grew to be tolerated in our culture, or understood by many legal and political elites to be constitutionally legitimate, at least in part because of” self-constraining norms that the presidency itself cemented.

Trump v. Hawaii reveals the practical, doctrinal, and normative difficulties of navigating today’s fraught relationship between the office of the presidency and its current occupant. The Court’s separate opinions advance starkly different visions of legitimate presidential authority, the rule of law, and the role of courts. And each opinion operates under distinct, in some ways surprising, assumptions about presidential control and presidential accountability. Read in this light, the opinions offer different perspectives on an increasingly significant and far-reaching question: the relationship between the authority of the presidential office and this president’s flagrantly anti-constitutional behavior.

The Chief Justice’s majority opinion celebrates the resiliency of the institutions of the presidency. These institutions, for the majority, appear capable of producing legitimate presidential authority irrespective of any discriminatory intent or animus-driven goals of the sitting president. The legitimating work of the institutional presidency—its deliberative processes, empirical assessments, and distinctive institutional competencies—appear disconnected from the goals or objectives of the president himself. The image of the presidency that the majority paints is not one of presidential direction to serve his chosen ends. Rather, it is an institutional infrastructure, seemingly autonomous from presidential interference—or at least insulated from presidential animus. Perhaps surprisingly for Justices otherwise committed to a strongly “unitary” vision of the executive, accountability neither starts nor stops with the president himself. As a result, the legal “authority of the Presidency” can be judged by the anodyne legalese that an interagency process has produced, notwithstanding the avowed animus of the President and his direction to his Cabinet to translate that animus into national security policy.

Justice Breyer’s dissent (joined by Justice Kagan) offers a somewhat different account of legitimate presidential authority and the role of courts. The institutional presidency can operate as a check on the animus-driven objectives of a sitting president. But the question is has it here? This is a factual question for courts to resolve. Implementation of the order’s “elaborate system of exemptions and waivers” by the other institutional actors that comprise the presidency provides relevant evidence. Is that on-paper system of exemptions and waivers “merely ‘window dressing,’” as the available evidence would seem to suggest? Or do these paper powers of the presidential bureaucracy operate as genuine, on-the-ground constraints on this particular president’s desire for a Muslim ban?

Recognition of executive “unitariness”—at least insofar as the president has informal powers to direct his subordinates—comes from Justice Sotomayor’s dissent (joined by Justice Ginsburg). Because the president in practice controls the goals and objectives of the presidency, in part by directing his senior aides and his Cabinet, the person of the president cannot—as a legal matter—hide behind the façade of an orderly office. Post hoc process should not be assumed to cure the presidential policy decision that preceded it. If presidential governance promotes accountability, then the president’s avowed animus has consequences: Deference ordinarily afforded to the office of the presidency is inappropriate when its current inhabitant has expressly repudiated the substantive and structural commitments of that office. Ordinarily, courts do not look behind the face of presidential policy to probe the institutional process that developed it. But a sitting president can forfeit that deference, and this one has, according to their opinion.

Justice Kennedy, perhaps foreshadowing the imminent news of his retirement, writes briefly to paint a final portrait of the presidency—as an office constrained by legal obligations that emanate from the “promise” of the Constitution but that (at least absent a showing of more overriding animus than that already made in the instant case) courts simply cannot be expected to review.

With this tapestry of opinions in mind, when can the institutions of the presidency absolve a president’s otherwise unconstitutional behavior of legal taint? Each of these opinions suggests a different answer. And each approach points to additional (and similarly vexing) moral and doctrinal questions. These are meatier questions than what a brief comment here can adequately tackle. So I’ll just close with some preliminary thoughts.

A first step might be to acknowledge the role of the institutional presidency in making the exercise of presidential authority legitimate or “respect-worthy” for the courts—that is, worthy of judicial deference. This is an issue that each side pressed at different stages in the travel ban litigation—and that each side struggled to articulate in legally meaningful terms.  When the President rushed the first EO out the White House door, those who ran to the courthouse door cried foul—here was a presidential order that lacked all indicia of a legitimate presidential process. The litigation that ensued appears at least partially responsible for spurring the institutional process that followed inside the presidency—a process that the Solicitor General would then argue to the Supreme Court makes the third iteration of the executive order different, in legally salient terms. But just how, Justice Kagan pressed during the oral argument, is that institutional process relevant under the governing precedent? The question warrants careful exploration and doctrinal elaboration. The majority opinion failed to articulate a clear, detailed, or explicit vision, leaving the question open for future courts to decide.

A second step would be to address forthrightly the challenging question of taint, and the circumstances under which the institutional presidency can absolve it. Both dissents started to gesture at this idea, but without directly confronting or resolving it. One possibility would be to adopt a burden-shifting model or some other threshold that triggers additional discovery aimed at determining whether the initial taint infected the institutional process that followed. Surely, the majority in Hawaii is correct that the page-count of one intra-governmental report does not tell us much about the interagency review that preceded the third iteration of the policy. But the dissents have a point as well—that this particular process may be a “sham,” a cover for this particular president’s longstanding and avowed animus. Evidentiary mechanisms may be necessary to better understand either the interagency process that produced the third order, implementation of the waiver system that it adopted, or both.

Trump v. Hawaii might be the Court’s last word on the travel ban. But it is also just a beginning. As ongoing litigation over the President’s transgender service members prohibition suggests, this is a President who makes avowedly animus-based decisions and then turns to the institutions of the presidency to “clean things up” for the courts. So the nexus between the office of the presidency and the conduct of its current inhabitant is one that the courts must continue to unpack. At stake is not only the legitimate exercise of authority by the presidency but also by the courts.

Looking beyond the courts, Trump v. Hawaii brings into view the normative difficulties of a president at war with the very institutions that are helping to legitimate (in the eyes of some), or at least to legalize his policies. Even as the president seeks to tear apart the longstanding norms and institutions of the presidency, other actors inside the executive are working just as hard to protect and repair them. This, then, is the deep irony of the Trump presidency: in working to preserve the norms and institutions of the presidency, those actors are also lending credibility to an office that empowers the very occupant who continues to rail against them.