The Pardon Power and Anti-Circumvention: A Trump Imprisonment Follow-Up

by Michael C. Dorf

I recently wrote a Verdict column on whether a person imprisoned on conviction of a crime can serve while in prison and an accompanying blog post on whether, if such a president were imprisoned pursuant to a state criminal conviction, he could successfully challenge his custody through a federal habeas corpus petition. Comments I received via email and on social media raised related questions about circumvention. In one example, a president might use the 25th Amendment to circumvent limits on the pardon power. In the other, a president might use the pardon power to circumvent a provision of Section 3 of the Fourteenth Amendment.

Let's start with a very brief recap. Suppose Donald Trump is tried, convicted, and sentenced to prison for a crime but then wins the presidency. Depending on the nature of the crime, he could be subject to impeachment and removal for it. He also could be deemed "unable to discharge the powers and duties of his office" within the meaning of the 25th Amendment. However, as I noted in the column, neither impeachment nor the 25th Amendment is self-executing. To remove an imprisoned Trump from office via either route would require the cooperation of many more Republicans (including cabinet officials Trump will himself have vetted precisely for this purpose with respect to the 25th Amendment) than are likely to come to his aid.

Thus, if Trump is in prison for a state offense--and thus beyond the reach of the presidential pardon power--there he would stay even as president, unless he could bring a successful state court case or a successful federal court habeas challenge of the sort I discussed in my earlier blog post. But what if Trump is in prison on conviction of a federal offense--in particular on what I regard as the most serious potential charges: his efforts to undercut the 2020 election result and to incite an insurrection? Could Trump pardon himself?

As I observed in the Verdict column, the prevailing but by no means unanimous view in the relevant literature is that a president cannot pardon himself. Suppose that the prevailing view is correct. Might the president nonetheless engineer a pardon through what I described in the column as a process of "musical chairs?" Here's how it works:

Trump could resign, leaving the office of the presidency vacant. His Vice President—let’s call her Marjorie—would then become President. Marjorie would then pardon Trump and also name him as her Vice President, which Congress would confirm. At that point, Marjorie would resign the presidency, Trump would again become President, he would nominate Marjorie as his Vice President, and Congress would confirm that choice.

A reader wrote me asking why go through the bother of musical chairs when Trump could simply transmit the required paperwork to Congress to voluntarily invoke Section 3 of the 25th Amendment. He could say honestly that he is incapable of performing the duties of his office because he's in prison. That would make the VP the acting president. She would pardon him. And then because the pardon means he's released from prison, he's capable of discharging the duties of office (or at least as capable as a malignant narcissist can be), thereby ending the period of disability. If this ploy works, then it seems like there's no real force to the notion that a president can't pardon himself.

Hold on, you say. Maybe in general a president can't pardon himself but this ploy works only because Trump has a ready-made basis for invoking the 25th Amendment--namely his imprisonment. Most presidents won't have that, right? Maybe not, but most presidents will be able to manufacture a temporary disability. If a president really needs a pardon, he can simply schedule a colonoscopy or other minimally invasive and arguably medically appropriate procedure that requires sedation, invoke the 25th Amendment for a few hours, and pick up a pardon.

Accordingly, if we think that the president can't pardon himself, we might think that as a corollary, the president can't invoke the 25th Amendment to obtain a pardon from an acting president. Such a principle is not evident on the face of either the pardon power or the 25th Amendment, but inferring it from those provisions is less of a stretch than some of the inferences the Supreme Court has made with respect to other constitutional provisions and structures.

Okay, but what about musical chairs? Should that also be forbidden? As long as we're drawing inferences, we could say a pardon received by a president who is briefly out of office pursuant to the musical chairs scheme is ineffective. Of course, "we" here would ultimately mean SCOTUS, where any case testing this proposition would trigger the justices' ideological druthers pretty strongly. It's hard for me to imagine our current SCOTUS saying that a musical-chair-pardoned Trump was not properly pardoned and must go back to prison.

But wait. Maybe a pardon can get Trump out of prison but doesn't put him in the White House in light of the disability to hold office imposed by Section 3 of the Fourteenth Amendment on previous oath-takers who "have engaged in insurrection or rebellion against the" United States. Trump's (anticipated) conviction for attempted overthrow of the government would trigger that disability, which, by its terms, can be removed "by a vote of two-thirds of each House" of Congress. Whether with or without the musical-chairs interlude, a pardon for Trump would not come with a two-thirds vote in Congress and thus should be ineffective in removing the disability and rendering him eligible to hold federal office.

But wait some more. Section 3 doesn't say that the two-thirds vote in each house of Congress is the exclusive means of removing the stain of insurrection or rebellion. It says such a vote does the job, but maybe a pardon does so as well. As I noted in the Verdict column, United States v. Klein held that a pardon removed a statutory disability with respect to property, and so didn't implicate the 2/3 requirement of Section 3, but it did suggest that the effect of a pardon--including for offenses involving disloyalty to the Union--is to conclusively wipe out any logical inference from the underlying offense. That principle could apply here too.

Because the Fourteenth Amendment was enacted after the original Constitution, we could read the 2/3 requirement of Section 3 as modifying the scope of the pardon power--at least with respect to crimes involving insurrection or rebellion. If so, then the pardon power would be ineffective in restoring eligibility. However, it's also possible to read the 2/3 requirement as attaching only where the disability is established some other way, and the pardon power as wiping out that disability so that Section 3 doesn't come into play in the first place.

Here, as elsewhere, we might recharacterize the circumvention question as one about the scope of each provision's coverage. If one agreed with the conservative Justices in the 2012 Obamacare Case that the Commerce Clause didn't authorize the individual mandate to purchase health insurance, one might try to characterize efforts to do the same thing via the Taxing power as impermissible circumvention. But the better view (I think) is the one taken by CJ Roberts: the Taxing power is a separate power; if the mandate is permissible under it, then it doesn't matter that it's not a permissible exercise of the Commerce power.

I'm not the tax expert among the Dorf on Law contributors, but I'll nonetheless close with an analogy to tax law. Some transactions will be deemed ineffective at tax avoidance where they have no apparent economic purpose and therefore can only be understood as efforts to circumvent tax liability. We might likewise ask whether a scheme has a purpose other than to circumvent some limit. Musical chairs would seem to fail that test, but a pardon that results in the end of imprisonment has a purpose independent of restoring eligibility for former insurrectionists, namely, forgiving a crime and ending imprisonment.

Again, though, that's as a matter of general principle. I make no claims about how the courts would actually rule should one of these very high stakes circumvention cases arise.