Getty / The Atlantic

In the 240 years since America’s founding, no former president has been indicted for criminal conduct. This isn’t because they were angels—far from it. And it isn’t because post-term indictment is not legally allowed. Instead, it is because Americans don’t like the idea of criminalizing politics. Both parties and the public see the prospect of post-term immunity as a guarantee that the country’s politics will remain civil and that power will transition peacefully from one party to the other. That is what drove President Gerald Ford to pardon Richard Nixon. And it’s one reason why the Office of the Independent Counsel decided not to indict former President Bill Clinton.

The presidency of Donald J. Trump has upended those calculations, and the resistance to post-term investigation may now come at too great a cost. When he leaves office, whether in January or four years later, the next administration or one of the states can and should investigate citizen Donald Trump—a former president whose legal status will be no different from that of any other American. The risk of politicization of such an investigation is far outweighed by the danger posed by failing to uphold our nation’s values. To protect future presidents from retributive investigations once they leave office, however, any investigation should be limited to Trump’s conduct before and after his presidency, not his behavior while he was president. If the findings of such an investigation justify it, prosecutors should indict the former president for violations of criminal law.

I come to this view reluctantly. The risks in the approach are both real and substantial. But after having served as a prosecutor in the Department of Justice, as a senior counsel in the Whitewater investigation of Clinton, and as a Bush appointee at the Department of Homeland Security, I’ve come to recognize that challenging, balanced judgments of the sort necessary today are sometimes forced on us by circumstances beyond our control. Hard choices do, sometimes, make bad law, but they cannot always be avoided. To decline to investigate Trump’s alleged criminality after he has left office is itself a choice—and it’s the wrong one.

The biggest danger of countenancing the investigation of ex-presidents is also the most obvious: an ever-escalating cycle of retribution. One can easily imagine a losing president resisting the call to leave the White House at least in part because he feared subsequent prosecution, or a winning president prosecuting her opponents over normal political differences. Indicting one former president risks making a habit of doing so, and reducing America to little more than a revolving-door banana republic. That’s why, for example, former Attorney General Eric Holder has reacted with grave concern to calls for Trump’s post-presidency prosecution. As Holder might put it—with substantial justification—if you thought “Lock her up” was the wrong thing to say about Hillary Clinton, you shouldn’t support a “Lock him up” perspective on Trump.

But a reluctance to prosecute does not mean there should be a prohibition against doing so. The idea of absolute presidential impunity from prosecution for all time and for all actions is just a re-instantiation of the kingly prerogative—“The king can do no wrong”—that was one of several reasons America had a revolution. Should a president who committed murder before his election that was only discovered once he was in office be immune from prosecution after impeachment and removal? Surely not.

And yet the promise not to prosecute after a term ends is part of the price we pay for the routine peaceful transition of power. One can readily imagine, for example, the violent reaction of some presidential supporters to even the hint of a possible criminal investigation.

This is true even in normal times, but it is all the more true during periods of deep political hostility. The prosecution of Trump after he leaves office, as the conservative journalist Jonathan V. Last recently wrote in his newsletter, The Triad, is of secondary importance to the more important value of preserving the nation: “Buttressing the rule of law today won’t matter if we descend into widespread, open civic unrest that undermines the legitimacy of the political system itself. That would be a generational, ongoing crisis. And once the toothpaste is all the way out of that tube, then there is no going back until the people who have decided to be against the system die off.” That’s a pretty grim prospect, and if that were the choice, it might be wise to buy civil peace with the coin of prosecutorial deferral.

But is that the standard we aspire to? Do we think so little of our civil society that we set rules of behavior based on fear of mob rule? America is often said to be a nation of ideals, not of cultural groups. It exists as a collection of aspirational principles—equality of opportunity, freedom of expression, and, ultimately, the rule of law. If we discard those ideas to save the nation, have we actually saved the nation? If we truly believe in those principles, then, without prejudging the result, it would be a dereliction of duty for the next president—or for any state with cause to investigate—to refrain from examining the potentially illegal actions of former President Trump just because of his previous title. As Teddy Roosevelt famously said, “No man [should be] above the law and no man [should be] below it; nor do we ask any man’s permission when we ask him to obey it.” To categorically say otherwise is to undermine the foundation of American democracy.

A post-term investigation would be on solid legal footing. Post-term immunity is fundamentally inconsistent with the ground that is offered by the Department of Justice for immunity from prosecution while a president is serving. The DOJ has long been of the view that sitting presidents cannot be criminally charged. It justifies that position in two ways.

First, it looks to practical questions of implementation. The DOJ has argued, broadly, that the possibility of an indictment and criminal prosecution of a sitting president would “undermine the capacity of the executive branch to perform its constitutionally assigned functions.” It is difficult, they say, to imagine a president running a government while sitting in jail. While other, lesser officials have successfully continued in office from prison (the example of James Michael Curley, who served as the mayor of Boston while in prison for mail fraud, springs to mind), it is not unreasonable to think that doing so would be impossible for the president of the United States.

In addition to the practical difficulties created by a requirement for the president’s physical presence at a trial or in jail, the DOJ has relied on the intangible but significant effects that an indictment and trial could have on presidential power. As Nixon’s DOJ put it in a memorandum prepared in 1973, “The President is the symbolic head of the Nation. To wound him by a criminal proceeding is to hamstring the operation of the whole governmental apparatus, both in foreign and domestic affairs.”

Notably, for our purposes, both the analysis and the import of the DOJ’s views are limited to a time when the president is still in office. After the president’s term is over, there is no longer the practical problem of running a government. Nor are there the same sorts of intangible effects on presidential symbolism; he is, after all, no longer the “head of the Nation.”

All of this is precisely why the DOJ has long justified its term-based immunity argument by contending that a president would be subject to prosecution “after he left office” (albeit while noting the possibility that a lapse in the statute of limitations might create a gap in criminality). In other words, in the department’s view, it is the office itself that commands the immunity, not the person. It would be strange and ironic if the argument for immunity during a term of office were somehow converted into a prohibition on post-term indictment as well.

It is likely that even the DOJ would argue against this sort of impunity. To do so would be, in effect, to recant much of what they said in 1973 and later repeated in 2000, with respect to Bill Clinton. As a formal matter, no legal barrier to post-term indictment exists.

Thus the ultimate question is not whether a former president can be investigated but whether one should be. What is best for our country? How can the country uphold the rule of law and the idea that no one is above the law, without driving itself into civil discord and risking fatal polarization?

There are no easy answers here. The best one can offer is a discretionary judgment that has some convincing rationale and offers a plausible way forward. In my view, the outlines of this are clear: It would be too great an affront to law for a president to have perpetual immunity. At the same time, the risks of polarization from criminalizing decisions that were made by the president during the course of a presidency is substantial. Alternatives, such as impeachment and loss of an election, exist that can address those wrongs.

Hence, let us try to thread the needle: Forgo the prospect of prosecution for actions undertaken while in office, but recognize that crimes a president commits while a regular citizen should not be excused just because he or she has served as the president of the United States.

This is not, by any means, a perfect solution. In our parade of horribles, there might be edge cases of conduct that occurred while the president was in office that would be so egregious we would want them to be criminally addressed. If, say, a hypothetical future president committed murder while in office, we would hope that a post-term prosecution for that offense would be permissible.

This example suggests that a ban on temporally based prosecution may be too broad and would, if strictly interpreted, revive the kingly prerogative against which we rebelled. On the other hand, any bright-line temporal rule that we adopt as a prudential matter has the virtue of being easy to administer and of avoiding post-term disputes about the level of egregiousness necessary for certain conduct to be prosecuted.

As a theoretical matter, the discretionary policy of not prosecuting an ex-president for acts committed while in office (especially those involving even tangentially the execution of his official duties) would have to yield in extreme cases. And while we cannot, with precision, define what those extreme cases might be, one hopes we would know them when we saw them.

Thankfully, we have yet to confront this degree of egregious behavior. For now, it is sufficient to articulate a general rule: A president should not be prosecuted after he leaves office for actions that occurred while he was the head of state, but he should remain subject to investigation for actions that occurred before or after his term.

To say anything else would be an affront. The powerful should be held to account. For society to function, all Americans must believe that crime doesn’t pay and that everyone is equal before the law. To avoid strife, we may exempt a president from criminal investigation for his political actions (however heinous and criminal they may be), but if we go further, and extend to him the kingly prerogative of impunity for his lifetime, we go a long way to destroying the faith in the rule of law that undergirds democracy.

With those concerns somewhat resolved, how do we decide whether to investigate former President Trump?

In many ways, the investigation of an ex-president should be no different from that of anyone else. As in other cases, a prosecutor would conduct interviews, subpoena documents, serve search warrants, convene a grand jury, and, in the end, if appropriate, ask the grand jury to return an indictment. There may be plea agreements, or trials, and then convictions, appeals, and, ultimately, perhaps, a prison sentence.

In that context, a prosecutor would typically ask two interrelated questions: First, is there sufficient admissible evidence of criminality that could sustain a conviction on the crimes to be charged? If not, the prosecutor should let the matter drop.

In Trump’s case, it seems clear that multiple credible criminal investigations are warranted. While not all of them may prove well grounded, the existing public record of well-documented allegations of criminal misconduct provides plentiful predication for opening an inquiry. This record includes but is not limited to a New York Times investigation that has described potential tax and mortgage fraud by Trump and the Trump organization; a narrower investigation of a series of transactions in the run-up to the 2016 election that has suggested the possibility of both tax fraud and campaign-finance fraud; a claim in Bloomberg that Trump may have committed insurance fraud; the uncovering of evidence by ProPublica of Trump’s alleged mortgage and tax fraud; the allegations of Trump’s niece, Mary Trump, that the Trump family committed fraud in the probate of her father’s will; and multiple alleged incidents of sexual assault (to the extent not barred by a statute of limitations).

There are other investigations for which there is likewise predication, but that, as a matter of prudence, we ought to forgo because they involve actions the president took while in office. For example, more than 1,000 prosecutors have concluded that the Mueller report uncovered ample evidence of Trump’s criminal obstruction of justice, and, additionally, former Trump staffers have reported the president’s corrupt offer of a pardon for illegal conduct that advanced his political interests.

One cannot, of course, know what an investigation of the allegations of pre-term criminal conduct might ultimately uncover (and, indeed, at least one, and possibly three, investigations are ongoing). But were Trump just an average citizen, there would be a basis to open up an inquiry into his behavior.

Which brings us to the second question: If prosecutors (at least those in the federal system, with which I am more familiar) conclude that there is sufficient evidence to prosecute, they will ask if reasons of public policy exist that suggest that the prosecution should not be brought. Typical reasons might be that it’s a small enough infraction that it’s not worth their time, that they don’t have enough resources, or that the prosecution won’t have any deterrence value.

The Principles of Federal Prosecution are intended to guide prosecutors in the exercise of their discretion, and offers nine (admittedly flexible) factors for assessment and consideration.

The first of these, which asks what the current federal priorities are, is not specific to any individual. It allows, for example, for an administration to say that it is focusing on drug crimes or for another to devote resources to fighting child pornography or white-collar crime.

The remaining factors, however, deal with the specifics of the offense and the nature of the defendant. How serious is the crime? How culpable is the accused in the scheme, and what is his role? What is his criminal history? And, more generally, what would be the deterrent value of the prosecution?

Here, it is fair to say that any balance we can strike at this stage, before all the facts are known, strongly suggests that an investigation of former President Trump would be consistent with these principles and that they would not bar an investigation of his conduct were he just a typical citizen. Trump’s pre-term conduct (if it is proved) would indicate a long-standing scheme of fraud (akin to that perpetrated by Bernie Madoff, for example) and significant financial abuse—exactly that sort of pattern of conduct and severity of offense that, in normal cases, would demand the investment of federal resources. If Americans are to have any confidence in the concept of the rule of law and equality before the law, and if the Principles of Federal Prosecution are to be applied in a neutral manner, the same result must obtain here.

Focusing exclusively on potentially illegal conduct that occurs outside the presidency is unlikely to solve the problems that lie ahead. Trump’s supporters will not be mollified by the distinction. And leaving unaddressed criminal activity that occurred during the presidential term may be too high a price to pay. But this sort of uncomfortable compromise is the only way to maintain accountability for crimes without making political differences a criminal offense. At least, I hope that is so.

The real shame, of course, is that we even have to contemplate this issue at all. Three times in the past half century, Americans have had to ask whether a president should be prosecuted after he leaves office. Perhaps the better solution would be to be more careful in the person we elect.

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.