In a recent working paper and op-ed, Professors Daniel Hemel and Eric Posner argue that the federal statute banning obstruction of justice applies to the president, and that the president would violate the statute if he intervened in an investigation to advance his personal interests. Hemel and Posner suggest that this can be reconciled with the president’s legitimate authority under Article II by recognizing that it would not be obstruction for the president to take action for legitimate reasons, such as setting law enforcement priorities for reasons other than personal gain. But even though the president normally has the authority to grant pardons or remove federal law enforcement officials from office, it would be a crime for him to do so to corruptly impede an investigation. Such action might expose the president to impeachment and criminal prosecution.

Hemel and Posner’s bold proposal raises a host of difficult constitutional issues, which they address with admirable clarity and creativity. But it also supports a more modest administrative law proposal, which I pose merely as a question: Should the obstruction statute be understood as a structural law, providing a limited grant of independence to federal law enforcement agencies? For example, should we understand the federal obstruction statute to qualify the president’s ability to remove the FBI Director from office at will? On this understanding, former FBI Director Comey might bring a civil action for reinstatement or back pay, instead of waiting for an impeachment and prosecution that might never come.

Unlike criminal prosecution of the president, which could raise questions that have never been addressed by a court, it is well settled that Congress has the power to protect certain officials from arbitrary removal. In Morrison v. Olson, the Supreme Court approved a statutory provision that ensured that an independent counsel charged with investigating high officials could not be removed absent a showing of “good cause.” The constraint on the president here would be far more limited. The president does not need a good cause to remove the FBI Director; the president is free to remove the Director unless he is acting on an exceptionally bad cause.

Admittedly, the independent counsel had a more limited mandate and purview than the FBI Director (at least in theory). But Morrison establishes that the key inquiry is whether the restriction on removal interferes with the president’s legitimate powers and duties under Article II. At least on Hemel and Posner’s reading, the president exceeds his legitimate authority and acts in derogation of his duty when he impedes an FBI investigation for purposes of personal or partisan gain. It is hard to see how a prohibition on removal in those circumstances would interfere with presidential prerogatives.

So assuming that Congress can prevent the president from removing the FBI Director absent a legitimate cause, the question becomes: has Congress already done so by enacting the federal obstruction statute?

If Hemel and Posner are right, and the obstruction statute provides for criminal sanctions for a president who removes the FBI Director under certain circumstances, that statute could also be understood as qualifying the president’s ability to remove the FBI Director at will.

A more typical argument runs in the opposite direction. The conventional wisdom, blessed by the Justice Department’s Office of Legal Counsel, is that the president can remove the FBI Director at will. Indeed, Professor Alan Dershowitz has taken this as an uncontroversial fixed point, and used it to urge that removing the FBI Director cannot constitute obstruction of justice.

And it is true that Congress has not enacted a statute saying, in so many words, that the FBI Director can only be removed for a purpose other than corruptly impeding a proceeding. But it is unclear why that type of express statement would be required. Some might demand a clear statement under the doctrine of constitutional avoidance, but I have argued elsewhere that the doctrine of constitutional avoidance should apply differently where Article II norms are concerned. And even that argument would not be necessary here, given the relatively clear propriety of a statutory provision preventing the president from removing a law enforcement officer absent a legitimate reason.

Courts also do not seem to insist upon express statutory language limiting the president’s removal authority. In Free Enterprise Fund v. Public Company Accounting Oversight Board, a majority of the Supreme Court assumed that SEC Commissioners cannot be removed at will, despite the absence of any express statutory language to that effect (and some substantial arguments to the contrary). The structure and function of the SEC had led courts and commentators to the understanding that Congress intended for the SEC to function as an independent agency. Here, Congress has gone further (at least on Hemel and Posner’s reading): Congress has actually made it a crime to remove FBI Director, except for a legitimate cause. There are many ways for Congress to say no to the president; a criminal statute is surely one of the most emphatic.

If the statute is understood as providing that the president cannot remove the FBI Director except for a non-corrupt purpose (instead of merely providing for criminal sanctions if he corruptly removes the Director), the legal toolkit for curtailing presidential abuses would expand. Instead of waiting for a criminal prosecution of the president, a targeted law enforcement official might refuse to accept the legitimacy of an order, or, if terminated, bring a civil action for reinstatement or back pay.

Such civil actions would allow for a public airing of the president’s motives, and for more limited remedies. A criminal prosecution of the president would be a slow-moving disaster—months or years spent identifying and investigating an issue, months or years spent impeaching the president (if the president’s followers in Congress will allow it), and months or years spent prosecuting the former president (if the president’s successor and courts stacked by the president’s party will allow it), and finally punishment. A civil action would cut through much of this, and would allow for a prompt and legitimate adjudication of the issues, along with a far less devastating remedy.

A court might try to duck a civil suit by citing political question concerns. But dismissing a civil suit seeking a relatively modest administrative law remedy could lead to a far more devastating showdown between the three branches. Requiring governmental actors to go all in—impeachment, prosecution, conviction—or not at all would also weaken Congress’s ability to discipline the president by limiting its less extreme options. Courts should resist the urge to take the easy way out.

Of course, this understanding would only temper certain kinds of presidential obstruction, like removal or interference with federal law enforcement officers. The president could still cause mischief with his other authorities, such as the pardon power. But there are other mechanisms for addressing such abuses. For example, pardons would make congressional investigations more dangerous to the president. They would strip congressional committees of the rationalization that they must be circumspect in their approach to avoid interfering with criminal prosecutions; they would strip potentially guilty parties of the ability to assert their right against self-incrimination, exposing them to contempt charges if they are silent and perjury charges if they lie; and they would leave the president exposed and alone.

The administrative law implications of Hemel and Posner’s arguments might not provide a complete solution, but they could address most of the problem, while causing far less damage to the system.

Aneil Kovvali is an Associate at Wachtell, Lipton, Rosen & Katz. The views expressed in this post are his own, and do not necessarily reflect the views of the firm or its clients.