The Game Theory of a Trump Subpoena

Time for some game theory.

The Washington Post reported Monday evening that Special Counsel Robert Mueller’s office has threatened to subpoena the president’s testimony in the Russia probe:

In a tense meeting in early March with special counsel Robert S. Mueller III, President Trump’s lawyers insisted he had no obligation to talk with federal investigators probing Russia’s interference in the 2016 presidential campaign.

But Mueller responded that he had another option if Trump declined: He could issue a subpoena for the president to appear before a grand jury, according to four people familiar with the encounter.

That Mueller’s people dangled the possibility of a subpoena is not a particular surprise. The threat of a grand jury subpoena is Mueller’s ace in the hole in his interactions with Trump—the reason Trump’s lawyers are negotiating at all over the possibility of an interview.

What is going on now between the Mueller camp and the Trump camp is a complicated multi-level game of chicken that is best understood not in terms of law but as a structured game. How it plays out will depend on a key variable into which there is little publicly available insight: How much does Mueller actually want an interview with Trump?

For what it’s worth, here’s how I analyze the game.

Support Lawfare

Mueller clearly wants an interview with Trump; we know this because he has asked for one. He wants it either because he is wrapping up his investigation and wants the president on the record on a number of points or because he genuinely needs information only Trump can provide to round out his evidence and make responsible judgments about how to proceed.

There are two ways to get an interview from an investigative subject: by consent or by seeking to compel the subject by issuing a grand jury subpoena. The former course requires the president’s agreement. The latter course requires the cooperation of the federal district court here in Washington.

It is probable—but not 100 percent certain—that Mueller can obtain the court’s cooperation. The reason it is probable is U.S. v. Nixon, in which the Supreme Court allowed the administration of justice and the needs of law enforcement to overcome a presidential assertion of executive privilege with respect to the Nixon tapes. The reason it is not 100 percent certain is that Nixon is not directly on point in a few discrete respects: It relates to physical material (tapes), not testimony, for example, and it relates to a trial subpoena, not a grand jury subpoena. So the president would have some room to argue that this case—in which a court would be forcing him to testify—is different. There is thus some litigation risk for Mueller if he actually issues the subpoena. First, there is the substantive risk that the case is litigated and the special counsel loses. (Steve Vladeck and I are writing a separate piece fleshing out the law on this point.)

Second, even if Mueller were to ultimately prevail on the question, litigation still carries costs. Issuing a subpoena would require months of litigation if the president resisted it. The fight would likely go all the way to the Supreme Court. Even expedited, it would slow things down—and Mueller does appear to be interested in speed.

The bottom line is that a subpoena is a weapon the special counsel has in his arsenal but does not want to use. Its use involves, first, a long delay and, second, a possibility (albeit a small one) of a loss that would set him back and set (from his point of view) a bad precedent. Mueller, in short, is using the threat of a subpoena to coax what he actually wants—which is a consensual interview. This kind of staring contest is not without precedent: A similar one took place between Bill Clinton and Kenneth Starr. Clinton ultimately backed down and appeared before Starr’s grand jury.

Now look at the matter from the perspective of Trump’s legal team—at least if it were acting rationally. (This may be condition contrary to fact, but let’s run with it for now.) It is a simply terrible idea for Trump to sit for an interview. He’s a liar; he speaks both impulsively and compulsively, and he probably has some legal exposure if he tells the truth. So the interview is a lose-lose proposition. This is particularly the case if Trump's lawyers believe that Mueller has nothing serious on Trump personally without an interview but also that the president may lie if he sits for one—and that his greatest criminal exposure involves not what he has already done but the lies he is likely to tell.

Normally, in such situations, the subject asserts his Fifth Amendment right not to incriminate himself—the right whose whole purpose is to relieve the so-called “cruel trilemma” of self-incrimination, perjury, or contempt. But while the president has a right to assert the Fifth, he can’t easily do so without serious political damage. So the alternative for his legal team is to bet that Mueller won’t actually pull the trigger and issue the subpoena, either out of fear of litigation defeat or out of desire not to delay. Rather than assert the Fifth, the Trump team is playing chicken with Mueller.

The problem with this strategy for the president is that the risks associated with losing a litigation are dramatically higher for him than they are for the special counsel. Remember that the law favors Mueller. What’s more, if Mueller issues the subpoena, Trump resists it and the court enforces the subpoena, Trump loses all of his negotiating leverage. Gone is the chance of negotiating a congenial environment for an interview, limiting the time or having his attorneys present. The normal grand jury witness, after all, has to go into the grand jury room alone.

Moreover, time may be an annoyance to Mueller, but it could be much worse than that for the president. A responsible lawyer for Trump has to protect him from the possibility of having to walk into the grand jury room alone, say, a week before the midterm elections. Time is not on Mueller’s side, but it’s really not on Trump’s side.

In other words, refusing consent for an interview may solve Trump’s cruel-trilemma problem in the immediate term but it’s also a roll of the dice in which the downside risk is unacceptably bad and not all that improbable.

So the question is: Who blinks first? Does the Trump team decide to negotiate the terms of an interview rather than risk litigation it is likely to lose? Or does Mueller back down and not follow through on the subpoena threat—thus letting Trump off the hook?

The key variable here is how much Mueller actually needs the interview. If it’s truly important for the investigation, he is much more likely to take his chances in litigation, accepting delay in exchange for the high chance of getting what he wants. By contrast, if he doesn’t truly need the interview, he might well decide to proceed without it. He’s no worse off, after all, than he would be if Trump asserted the Fifth—which the president might do at the end of a long litigation anyway.

To boil it down, here’s what can be said about the possibility of a Trump interview:

First, the logic of the situation favors a negotiated outcome. Neither side has an interest in protracted litigation, and both sides have litigation risk. As the probable outcome of any such litigation is a Mueller win, the situation favors negotiated resolution on terms broadly favorable to Mueller.

Second, if Mueller really needs the interview, he’s going to get it. The litigation risk for Trump is dramatically higher than the risk for the special counsel’s office. That means that the moment Trump’s team truly believes that the alternative to a voluntary interview is losing in litigation, the Trump team will negotiate terms for an interview—or it will take the political hit of asserting the Fifth.

Third, conversely, if Mueller does not really need the interview, he may well blink and let Trump get away with not sitting for one. Indeed, if neither an interview nor a subpoena materializes, that is a sure sign that Mueller is content to proceed without hearing from Trump. That could mean either that his case is weak or that it is strong. It indicates that his ultimate decisions as to how to proceed do not depend on what Trump has to say.

Fourth, in any negotiated interview, the more accommodations Mueller is prepared to make to Trump, the less crucial we can assume the interview is to his investigation. If the interview is a make-or-break thing for Mueller, he will hold the line and make sure Trump sits for live questions. If he allows written answers, that’s a sign the stakes are lower for him, for one reason or another. In my judgment, the mere fact that he was willing to telegraph so much subject matter to Trump’s lawyers suggests that he doesn’t really need Trump’s input that much; if the interview were truly high stakes for Mueller, I doubt he would give the Trump team the opportunity to script answers for him.

Finally, if a subpoena issues, that is not the endpoint of the game but an escalation of it—a sign that neither side has blinked yet. The essential logic of the situation persists until there is an actual court order changing the negotiating status quo or a negotiated interview to avert the litigation. An accommodation is possible even after litigation has begun.