Presidential Immunity: Preliminary and Tentative Thoughts

Mark Tushnet

Some time after September 11, 2001, I started writing about the constitutional dimensions of emergency powers under the US Constitution. I gave up the project after the Bush administration retreated from its most advanced—and interesting—positions. Before I did, I had begun to develop the view that the best way to understand emergency powers, both descriptively and normatively, was captured by terms like “extra-legal” or “extra-constitutional.” Law professor Oren Gross had already used the term (here and here), as had political scientist Benjamin Kleinerman, whose book The Discretionary Presidency: The Peril and Promise of Executive Power I finally got around to reading in my crusade to rid myself of the unread books I’d accumulated over the decades.

Trained in law, I was interested in the institutional implications of describing something in those or similar terms. We all seemed to agree that they implied some sort of retrospective evaluation through some sort of political rather than juridical process. One possibility, to which I was attracted, but now am not, was that the retrospective political process was ordinary politics: A president would act in an emergency and voters would later approve or punish him/her and their party in subsequent elections. Relying on Locke and Madison, Kleinerman correctly points out that ordinary politics might not be sufficient because ordinary politics includes too much “mere” approval or disapproval of outcomes, too little (if any) component of constitutional evaluation. Kleinerman argues that the retrospective evaluation requires that the President “prove” (his term) that the actions taken were truly necessary to preserve the nation—but (perhaps because he’s not a lawyer) he doesn’t spell out the institutional form for making that proof.

Other institutions for retrospective evaluation might be impeachment and Truth and Reconciliation-like commissions, the former explicitly constitutionalized (but perhaps too difficult to use given partisan divisions centered not on constitutional concerns but, again, on approval or disapproval of the merits of the actions taken), and the latters’ ad hoc nature perhaps giving them constitution-like status. (After January 6, 2021, I did suggest the use of such a commission but it turned out that partisanship prevented the creation of one--I initially had this as "Republican partisanship" but I know that Republicans say that the partisanship originated in Nancy Pelosi's rejection of Republican "nominees" for the Select Committee; this "you did it first" back-and-forth is a characteristic of constitutional harfball that I identified in my initial presentation of that idea.)

Trump v. United States brought to my mind my earlier thinking about emergency powers—though it bears emphasizing that that setting had a triggering condition (“emergency”) absent from Trump v. United States. What follows are truly tentative and preliminary thoughts, inconsistent with a tweet I posted invoking Wittgenstein (“Whereof one cannot speak, thereof one must be silent”).

Maybe we should think of immunized presidential action as “extra-criminal.” The starting point would be that presidential action immunized from criminal (or civil) liability remains criminal though unprosecutable. A second preliminary point is that the immunity is the President’s alone (or so it appears). That is, presidential subordinates aren’t immunized directly. So Trump shooting someone on Fifth Avenue is different from Trump directing his chief military aide to do so.

Assuming that Trump is immunized for shooting and for giving the aide the order (not inevitably true depending on circumstances), the aide would be criminally liable absent some defense. The obvious defense is compliance with a superior’s orders, but in the military that’s not available for “manifestly unlawful” orders (and outside the military there’s no such defense, I believe)—and Trump’s order would be manifestly unlawful even if he couldn’t be held liable for giving it. (As the scenario has developed on the internet, that’s why the president’s pardon power comes up: “Go shoot my political opponent. I’ll issue you a pardon immediately upon your completion of the task—or here’s a pardon written out that you can carry in your pocket when you do it.” [I think there’s a serious question about whether a president’s pardon power extends to future actions and, as the military aide’s lawyer I’ll tell him/her that the prospective pardon might be valid but might not be—and that the president might or might not follow through on the promise to pardon.])

The emergency-powers problem differs from this one in another way. In that setting there are (usually) no ways of obtaining a determination, prior to the action being taken, by some institution other than the presidency that there really is an emergency. That’s the burden of the generally though not universally agreed-upon proposition that emergencies take such variegated forms that there’s no realistic way of specifying beforehand what counts as an emergency, at least not in terms that significantly constrain presidential discretion. (One of my favorite examples is the provision in the ICCPR referring to an emergency that “threatens the life of the nation.”) The criminal setting is different because we ordinarily have an institution—the criminal process itself and the jury—to certify that the action was indeed unlawful.

My guess is that retrospective evaluation of an action as criminal will be significantly more difficult than retrospective evaluation of actions taken in an emergency because of the absence of such a certification. I suppose we could think of this as a situation in which, faute de mieux, the people are allowed to “take the [criminal] law in their own hands” through some form of collective action, though of course that phrase has a badly damaged history. (And, notably, in some real-world lynchings in the US West those who took the law in their own hands did so in a situation in which they believe that the institutional certification provided by the criminal process is unavailable because the process couldn’t be deployed until, months later, a judge would be available.) And, equally of course, the suggestion is not that lynching those immunized by Trump v. United States is the correct form of retrospective evaluation. (Given the way social media work these days, I suppose the preceding sentence should be in ALL CAPS.) Following the thoughts about emergency powers, we need some institution different from ordinary politics for performing the retrospective evaluation. At present I’m at a loss to figure out what such an institution would look like.