Executive Branch

Prosecuting Trump: A Reply to Josh Marshall

Jack Goldsmith
Wednesday, August 17, 2022, 11:30 AM

Even if one focuses narrowly on the rule of law, caution is still warranted.

Donald Trump leaves Mar-a-Lago aboard Marine One on Feb. 18, 2019. (White House)

Published by The Lawfare Institute
in Cooperation With

Josh Marshall has a thoughtful response to my Aug. 14 piece on the Mar-a-Lago raid. Below I try to engage with his points.

Where We Agree

Marshall says that I am right to insist that “it is dangerous for a President’s administration to pursue criminal charges against a predecessor.” But he adds that “[w]hat’s lacking in [my] discussion is any real grappling with the opposite side of the equation.” The opposite side of the equation is “the line past which a President’s lawlessness becomes so pervasive, persistent and dangerous that the risks of acting are outweighed by those of failing to do so.” 

I agree that there is such a line. Indeed, I said so in the piece he responds to:

There obviously must be a point where information is so sensitive, and Trump’s disregard for law so extreme, to justify legal process against Trump, even in the current milieu. Otherwise the law is entirely hostage to a former president’s (and his supporters’) self-serving veto—something no legal system can tolerate. 

I then explained how I think one should assess whether that point had been reached with the Mar-a-Lago documents episode, without taking a position.

As for my “grappling with the opposite side of the equation,” I have tried to acknowledge very clearly the consequences of giving Trump a pass for convictable crimes:

If Mr. Garland concludes that Mr. Trump has committed convictable crimes, he would face the third and hardest decision: whether the national interest would be served by prosecuting Mr. Trump. This is not a question that lawyerly analysis alone can resolve. It is a judgment call about the nature, and fate, of our democracy.

A failure to indict Mr. Trump in these circumstances would imply that a president—who cannot be indicted while in office—is literally above the law, in defiance of the very notion of constitutional government. It would encourage lawlessness by future presidents, none more so than Mr. Trump should he win the next election. By contrast, the rule of law would be vindicated by a Trump conviction.

I agree with Marshall about the dangers of not prosecuting Trump for convictable crimes. They are extremely high.     

I also agree with Marshall that Trump’s criminal activity after office warrants less prosecutorial caution than those committed in office. I made this point in “After Trump,” where I “argue[d] for extreme caution in a criminal investigation of a prior president for acts done in office” (emphasis added), and added that for non-official acts outside the presidency, the calculus is very different. I also made the point implicitly in this post, which emphasized that “[o]nce Trump leaves office, his criminal exposure for selling secrets broadens.” 

Grappling With the Dangers of Not Prosecuting Trump

 All that said, Marshall is right about my allocation of grappling: I have devoted more words to the dangers of the Biden administration using criminal process against Trump than to the dangers of it not doing so. In part that is because most of the commentary that I have read focuses on the dangers of not pursuing Trump. And in part it is because I still believe many people unduly discount the difficulties and possible costs of prosecuting Trump. I will not re-catalogue all of the difficulties and costs here. The narrow point I want to make is that the difficulties and costs should substantially inform the rule-of-law concerns that Marshall raises. I can make the point in two steps.

  1. Convictable Offenses 

The first step goes to whether, as Marshall says, “Trump’s violations of law are pervasive.” To the extent that Marshall is referring to criminal law, I am far from convinced, based on the public evidence to date, that Trump has committed convictable crimes. (Recall that the standard to bring a case against Trump is that his acts constitute a federal offense and “the admissible evidence will probably be sufficient to obtain and sustain a conviction.”)

Marshall says: “Quite apart from potentially conspiring with a foreign power, the Mueller Report contains ample and convincing evidence of obstruction of justice.” I am not sure which of Trump’s many acts Marshall has in mind when he talks about “conspiring with a foreign power.” But absent a clearly venal quid pro quo, it will be practically impossible to nail a president—who has enormous constitutional discretion to direct foreign policy and define U.S. foreign policy interests and conduct diplomacy—for “conspiring with a foreign power” while in office. Which is why that issue does not now seem to be on the table.

And while Trump did seem in office to engage in “obstruction of justice as a way of life,” in John Bolton’s memorable words, that does not mean that there is “ample and convincing evidence” that he has committed convictable obstruction of justice crimes. (I know this seems inane to many people; more on this issue below.) I have explained why I am skeptical that Trump committed convictable obstruction of justice crimes at great length here and here and here, and will not repeat those arguments. Some of the episodes recounted in Volume II of the Mueller report would face fewer Article II problems than others, but all would have been heavy lifts to convict Trump for obstruction of justice.

And then there is what Marshall describes as “the multifaceted and multi-jurisdictional plot to corrupt and then overturn the 2020 election all culminating in the violent assault on the seat of government.” This was all atrocious, unpatriotic, impeachable, and unforgivable. But I am still not convinced, based on current public evidence, that it meets the standard for federal prosecution. 

The numerous criminal statutes that different people say Trump supposedly violated by his Jan. 6 and related actions makes this claim hard to assess, and the breadth of the statutes invoked are some indication that none of them apply straightforwardly to Trump’s actions. The public hasn’t yet seen a smoking gun, and any prosecution here will face intent and Article II problems. I do not rule out that the Justice Department can make out a winnable case here—we do not know what Merrick Garland knows. I just have not yet seen a convincing analysis, based on the public evidence, that grapples with all of the difficulties.  

(To get a flavor of how hard the legal issues will be, independent of the intent and Article II problems, read this opinion by Judge Rudolph Contreras in a criminal case against Kyle Fitzsimons, who allegedly breached the Capitol on Jan. 6 and “engaged in physical violence” that injured police officers. The case involves one of the statutes most frequently invoked by commentators against Trump—corrupt obstruction of an official proceeding, 18 U.S.C. § 1512. Contreras denied Fitzsimons’s motion to dismiss the charges under § 1512. But as the opinion makes clear, there are uncertainties about whether that statute applies even to someone who actually engaged in violence in the Capitol on Jan. 6—all of which will be contested on appeal. The legal issues will be harder in a prosecution under this statute against a president who did not breach the Capitol or engage in violence.) 

Next is what Marshall says “appears to be a concerted and deliberate effort to maintain possession not only of government property but significant state secrets for reasons unknown.” I agree that post-presidency criminal actions face a lower bar, and I have outlined (in the context of the warrant) what I see as the legal issues here. Not enough information has yet surfaced on this one, in my opinion.  

As for the “evidence that Trump [after he left office] personally contacted potential witnesses against him and tried to ensure their silence with a mix of inducements and threats,” if the evidence ends up supporting that description, the Department of Justice should face none of the legal hurdles to prosecution identified above. We will see.

2. The Danger in a Failed Prosecution

The second factor builds on the first. The absence of an obvious convictable crime (based on public evidence thus far) underscores the dangers of a failed prosecution.  

“A trial of a president following years of battle and ending in acquittal will cause the country tremendous pain for no good end and will likely create numerous unfortunate precedents along the way,” I once wrote. And in assessing the possibility of failure, one must consider “how any charges against Mr. Trump would fare in an adversarial criminal proceeding administered by an independent judge, where Mr. Trump’s lawyers will contest the government’s factual and legal contentions, tell his side of events, raise many defenses and appeal every important adverse legal decision to the Supreme Court.”

In calculating costs and benefits, in other words, one cannot assume that a prosecution will succeed. One must discount all of the dangers of not prosecuting (and the benefits of a prosecution) by the likelihood of the prosecution failing.  

Would a failed prosecution be good on balance for the rule of law, broadly conceived—for the idea that no person, including a former president, is above the law; for the incentives it would create for Trump and others going forward; and for the devastating impact it would have on Justice Department legitimacy, beyond the legitimacy hit the department would suffer for going after the president’s political rival? I don’t think so. To be sure, some people believe, as I once put it, that the rule of law “might be enhanced by a full judicial airing of Mr. Trump’s possible crimes in office, even if it ultimately fails.” But I think that is very unlikely. Since I think that the likelihood of a conviction of Trump based on publicly available evidence is low, I do not think the rule of law would be advanced by prosecuting him.


This reasoning, ultimately, is why I continue to put myself in the caution category, and to emphasize dangers of prosecution. In my view, the appropriateness of prosecuting Trump goes up a lot if and when the Justice Department has clear evidence of a convictable crime. I have not seen it yet. Marshall thinks he has seen it, and that we have reached the point where the “risks of acting are outweighed by those of failing to do so.” That is the essence of our disagreement. I should note, in this regard, that both my argument and his rest on guesses about future consequences based on legal and practical reasoning grounded in very incomplete information. The right posture here is humility about how little we know, and prayer that Merrick Garland makes the right calls.    

Gaps in Presidential Accountability

I want to close by addressing a theme that I think informs the pro-prosecution camp’s judgment. The theme is that Trump has done so many flagrant, awful, obviously bad things in and out of office that he must have committed a prosecutable crime. It seems unfathomable that any president or ex-president could act in such persistent defiance of law and legal process, in such blatant violation of the public trust, and not have committed crimes. As Marshall puts it, “[T]he very brazenness of Trump’s indifference to the law has a way of rendering precious or quaint any serious and deliberate discussion of potential consequences.”

To frame the problem as Marshall does here is to neglect everything we have learned about the massive gaps in presidential and post-presidential accountability since January 2017. Time and time again, Trump has done outrageous things with indifference to law. And yet time and time again, what Trump did violated norms but not laws, or Trump operated with a mad-genius capacity to find and exploit legal loopholes, especially remedial loopholes, in legal constraints on the presidency. 

Bob Bauer and I wrote a book on this theme. “Trump has shown that the current array of laws and norms governing the presidency is inadequate to protect institutions vital to American constitutional democracy and to ensure that the president is, and appears to be, constrained by law,” we said, in a non-original judgment. We detailed the main accountability gaps that became apparent during Trump’s candidacy and presidency—on foreign interference in elections, conflicts of interest, tax disclosure, pardon abuse, obstruction of justice, politicized investigations, and much more; and we proposed reforms. Trump has continued to illuminate and exploit accountability gaps since we wrote the book—in his abuse of the transition process, in his actions on and related to Jan. 6, and in his handling of sensitive documents during and since he left the presidency.  

For six years the Justice Department has struggled to use extant legal tools to redress Trump’s bad acts. In part its struggle is a result of Article II of the Constitution. The president is in charge of executing criminal law, and thus it has always been hard to execute criminal law against the president; the president is in charge of diplomacy and, in so many ways, gets to decide what U.S. foreign policy is; and the president creates and controls the classified information system and can do what he wants with it while in office. This massive constitutional allocation of authority to the president is why, ultimately, “our system of government depends on the People electing a President who is generally reasonable, prudent, and responsible.”

But Article II is not a bar to holding a president much more accountable while in office and afterward. Laws, regulations, and norms could go much further to impose constraints on a president and ex-president like Donald Trump. A future Justice Department would have an easier time holding a bad-actor president to account if laws, regulations, and norms were clarified to address the acts that we now understand a president can commit. It would also be and appear more legitimate in doing so. 

Yet Congress and the Biden administration have to date done nothing, really nothing, to move in this direction. The nation has since Jan. 20, 2021, put its head in the sand about the very foreseeable repeats of everything Trump did. Except next time it will be worse, because the next norm- and law-defiant president—who could well be Trump—will be much better prepared to circumvent legal and norms-based limits, and much defter at doing so.

Legal reform cannot by itself save us from this situation—only elections can. But legal reform can help, maybe a lot, by eliminating the legal ambiguity and loopholes that Trump has dined on. It can help constrain a future president’s worst inclinations and, perhaps more consequentially, constrain the subordinates through whom he or she must act.         

There is much to do, but the priorities now should include at least these three reforms. First, Congress should enact Electoral Count Act and presidential transition reform; an outstanding bill in the Senate is a great start. Second, Congress should do a much better job of specifying which criminal laws apply to the president and how—starting, as Bauer and I explained, with obstruction of justice. There has been no real movement in Congress on this issue, and reform here will be hard in the current political environment. Third, the Biden administration can and should take steps to clean up and clarify the presidential role related to classified information. I will have more to say on this issue, hopefully in the coming weeks.

Jack Goldsmith is the Learned Hand Professor at Harvard Law School, co-founder of Lawfare, and a Non-Resident Senior Fellow at the American Enterprise Institute. Before coming to Harvard, Professor Goldsmith served as Assistant Attorney General, Office of Legal Counsel from 2003-2004, and Special Counsel to the Department of Defense from 2002-2003.

Subscribe to Lawfare