The Coming Constitutional Storm

But whatever the baseline for Supreme Court confirmation acrimony might now be, the fight for Kennedy’s vacant seat will surpass it. President Donald Trump is replacing the Court’s longtime “swing justice,” which will shift the Court’s ideological center of gravity on at least some significant issues. Even in tranquil times, this nomination would be a political knife fight. In the most venomous political year of recent memory, this nomination could become even bloodier, capable of exceeding in rancor even the attempted “high-tech lynching” of Clarence Thomas and the successful character assassination of Robert Bork.

Indeed, President Trump had not even announced his nomination of Brett Kavanaugh when the New York Times called upon “Democrats and progressives to take a page from ‘The Godfather’ and go to the mattresses on” the not-yet-announced nomination. Perhaps the Times deserves credit for understatement; mob warfare may understate the partisan furies that will be unleashed upon Kavanaugh.

Still, we should not let the worst aspects of the confirmation process overshadow its best aspects. Each Supreme Court vacancy is an occasion for the nominee, the president, and the senators to look squarely at the Court’s role in our constitutional republic. And though each confirmation hearing tends to focus greatly on familiar disputes—as always, Roe v. Wade—the senators inevitably invoke current controversies in their rhetorical questions to the nominee. Thus, each nomination is a product of its time—the John Roberts and Samuel Alito hearings raised post-9/11 questions of presidential powers in wartime; the Elena Kagan and Sonia Sotomayor hearings raised questions of campaign-finance reform; the Neil Gorsuch hearing raised questions of religious liberty and, as usual, of Donald Trump. In that respect, Supreme Court confirmation hearings remind us that our momentary political debates ride atop deeper, timeless constitutional debates; a Supreme Court nomination provides the opportunity for the senators and all of us to focus on those deeper themes.

And while we may complain that judges play an outsized role in American politics, the fact remains that we put them at the center of our constitutional debates. For even those of us who bristle at the notion of “judicial supremacy” tend to invoke judicial authority in support of our own constitutional arguments—we always have. Nearly two centuries ago, when Alexis de Tocqueville studied American democracy, he observed that there is “no political event in which [one] does not hear the authority of the judge invoked.” Yes, nearly all political questions are eventually resolved into judicial questions, as Tocqueville famously observed. But even before those questions reach the courts, we invoke the authoritative views of judges. Which, in turn, vests those judges with political authority.

And so the public will watch as the president and Senate choose the next justice who will, with his colleagues on the Court, have the final word on nearly all of the most important constitutional issues of our time. We can hope that the president and senators approach the task with the high-mindedness that it deserves. But in this constitutional moment, as the Senate weighs the president’s choice, we are little more than spectators . . .

. . . Unlike the other part of our constitutional moment, in which we won’t have the luxury of watching the game from the sidelines. Rather, the American people themselves will be responsible for rendering the final judgment.

For over a year, we have watched three constitutional debates slowly unfold. First, there is the investigation by Robert Mueller, the Justice Department’s special counsel, raising difficult questions of whether the prosecutor can subpoena or even indict the president. Second, there is President Trump’s invocation and use of his constitutional power to pardon, raising unsettling questions of whether the president can use the pardon power tactically to thwart an investigation of himself or even to pardon himself. And third, there is the specter of impeachment, raised by congressmen and activists who hope that Democrats will win control of the House of Representatives this fall.

Each of these issues—prosecution, pardon, and impeachment—is constitutional but not legalistic. Each involves legal authorities that are not defined precisely and therefore turns significantly on the relevant actors’ prudential and ethical judgments, not just their legal judgments.

More and more, we hear these debates described in terms of “constitutional crisis.” That term means different things to different people but at its core it reflects public unease with a conflict among government officers or institutions that lacks an obvious mechanism for resolution—which means, more precisely, a conflict that will not be resolved by the courts with the acquiescence of the warring parties.

But what that really means, in turn, is that it is a conflict that can be resolved only after the fact, at the ballot box, by us—a rare opportunity for the people themselves to make decisive constitutional judgments on questions of consequence. If we see such an opportunity as a constitutional “crisis,” then this says less about the warring government factions than it does about ourselves and our capacity to make collective, deliberate constitutional judgments instead of leaving them to federal judges to decide for us.

“I have the absolute right to PARDON myself,” the president tweeted on June 4. Then he added, “but why would I do that when I have done nothing wrong?”

His caveat offered little assurance to his critics, who from the outset of his presidency have worried that the president will use the pardon power tactically—either to advance his political agenda or to insulate himself against investigations by pardoning people close to him—or perhaps even take the unprecedented step of pardoning himself.

So far President Trump has granted seven pardons. The fifth was of Dinesh D’Souza, the political activist who broke federal campaign-finance laws by funneling $20,000 in campaign contributions through other people to a Republican Senate candidate. Trump pardoned him just in time for the release of D’Souza’s latest film, Death of a Nation, which compares Trump to Lincoln. But Trump’s first and most controversial pardon was of Joe Arpaio, former sheriff of Arizona’s Maricopa County, whose longtime record of aggressive and abusive actions against Hispanics and immigrants led to his federal conviction for contempt of court in July 2017. A month after Arpaio’s conviction, President Trump pardoned him.

Trump’s pardons of Arpaio and D’Souza were unsettling; their convictions and sentences were well justified, and their pardons seemed to reflect sheer politicking. But the pardons also raised concerns that President Trump was pardoning political allies to send a message to friends and former associates, such as his campaign chairman Paul Manafort, to “hang tough” in their own criminal proceedings rather than turning against the president.

President Trump’s own legal team has done little to dispel such concerns. When asked in June whether the president could someday pardon Manafort if he were to be convicted in the course of the Robert Mueller probe, Rudy Giuliani told CNN, “When it’s over, hey, he’s the president of the United States, he retains his pardon power, nobody’s taking that away from him. . . . I couldn’t, and I don’t want to take any prerogatives away from him.” This evidently reflected the legal team’s longtime position: In June, the New York Times released a January 2018 memo from Trump’s lawyers to Mueller in which the lawyers asserted that the president “could, if he wished, terminate the inquiry [into former national security adviser Michael Flynn], or even exercise his power to pardon if he so desired.”

It is a worrisome situation. That, however, does not mean it is a situation easily or properly resolved by courts. But many of the president’s critics leap to that conclusion, framing their own responses with arguments that the president can’t lawfully use pardons in these ways and that courts must intervene to stop him. After the Arpaio pardon, law professor Laurence Tribe argued in the Washington Post that federal judges should simply nullify the pardon, because “pardoning Arpaio for his willful disobedience of a court order to stop violating Arizonans’ constitutional rights” was itself a violation of the Fifth Amendment’s right not to be deprived of “life, liberty, or property, without due process of law.”

Tribe’s argument got no traction in the real world. Parties filed briefs in federal court, arguing that the pardon was unconstitutional, but Judge Susan Bolton, who had herself convicted Arpaio for contempt of court, promptly dismissed Arpaio’s criminal prosecution because of the pardon.

Other critics, who focus on the possibility of Trump’s pardoning people being investigated by Mueller, raise other legal arguments against the pardons. The most nuanced and careful version of this argument was advanced by Benjamin Wittes at the prominent Lawfare blog. Wittes examined the federal obstruction-of-justice statutes and concluded, “It’s not clear to me why a facially valid action taken in the service of managing the Executive Branch, taken with specific intent to commit a crime in order to influence a judicial proceeding, can never violate statutes that endeavor to protect the judicial function.” Wittes carefully avoids a categorical conclusion that Trump pardons would constitute obstruction of justice. But even his argument that a presidential pardon could be criminal is, itself, a controversial proposition.

Others are less nuanced. Jed Shugerman and Ethan Leib, law professors at Fordham, point to the president’s constitutional duty to “take Care that the Laws be faithfully executed” and assert categorically, “If the president pardons his associates primarily out of a motivation to protect himself, those pardons would also be invalid as disloyal, and federal courts should probably allow those prosecutions to proceed notwithstanding the pardon.”

Shugerman and Leib offer no support for this assertion, other than the inferences that they draw from the “take care” clause, because none exists. It is one thing to argue that the president is duty-bound under the clause not to use his pardon power in service of lawbreaking; it is quite another thing to assert categorically that pardons granted for improper reasons would be null and void and that the courts “probably” would enforce this reading of the Constitution. Shugerman and Leib, like so many of the president’s critics, see a constitutional issue and leap confidently toward a judicial solution.

In that respect, they exemplify an unfortunate habit of too many Americans (and far too many lawyers). But their confidence notwithstanding, arguments over the president’s constitutional power to pardon are not susceptible of such easy judicial resolution.

The Constitution’s provision committing the pardon power to the president is written with one and only one exception: “The President . . . shall have Power to grant Reprieves and Pardons for Offences against the United States, except in Cases of Impeachment.” There is no exception for self-interested pardons or corrupt pardons. And the fact that the Framers expressly included an exception—for impeachments—shows that they knew how to include exceptions. As the Framers chose not to add another express exception for self-interested pardons or corrupt pardons, their words are best interpreted as intending not to include any implicit limits along those lines.

Moreover, to the extent that the Supreme Court has grappled with the pardon power, it has poured cold water on the suggestion that Congress can pass laws restraining the president’s constitutional pardon power. “To the executive alone is intrusted the power of pardon,” the Court concluded in United States v. Klein (1871), “and it is granted without limit.”

In Klein, the Court was rejecting Congress’s efforts to limit the broad use of the pardon power for Confederates, first by President Abraham Lincoln and then by President Andrew Johnson. The Court reiterated its position a century later: “A fair reading of” English and American constitutional history, the Court explained in Schick v. Reed (1974), “compels the conclusion that the power flows from the Constitution alone, not from any legislative enactments, and that it cannot be modified, abridged, or diminished by the Congress.”

So even if Congress had actually intended for the federal obstruction statutes to limit the president’s pardon power—which is probably not the best interpretation of those statutes—it lacks the power to impose such limits. At least according to the Court.

And, perhaps more important, according to our Constitution’s most famous expositor. In Federalist 74, Alexander Hamilton writes that “the benign prerogative of pardoning should be as little as possible fettered or embarrassed.” And he emphasizes the Framers’ wisdom in subjecting the pardon power to one and only one exception—namely, impeachments, for which the president cannot grant pardons. (Indeed, when Hamilton notes that the only other exception that some had sought to add to the Constitution was one for treason, he further undercuts any argument that courts should step in to nullify pardons that are self-interested or corrupt, or that “obstruct justice.”)

Yet the most important part of Federalist 74 is not its defense of the broad pardon power. It is Hamilton’s justification for entrusting that broad and largely unfettered power to the president. As he explains, the pardon power is entrusted to the president because we can expect the president to temper the criminal laws’ categorical prohibitions with case-by-case mercy. “The criminal code of every country partakes so much of necessary severity, that without an easy access to exceptions in favor of unfortunate guilt, justice would wear a countenance too sanguinary and cruel,” Hamilton writes. By providing this tool of mercy to a single man, who is in turn accountable to the people at large, the Constitution relies not on judicial management but presidential judgment; not on the president’s accountability to judges but his accountability to the people. “The reflection that the fate of a fellow-creature depended on his sole fiat,” Hamilton argues, “would naturally inspire scrupulousness and caution; the dread of being accused of weakness or connivance, would beget equal circumspection.”

In other words, Hamilton expected that presidents would exercise the pardon power with an eye both to the interests of the criminal and the expectations of the general public. This is an argument not from checks and balances but from presidential character and self-restraint. It reflects Hamilton’s earlier argument, in Federalist 68, that our Constitution “affords a moral certainty, that the office of President will never fall to the lot of any man who is not in an eminent degree endowed with the requisite qualifications”—not men with “[t]alents for low intrigue, and the little arts of popularity,” but rather “other talents, and a different kind of merit, to establish him in the esteem and confidence of the” people, such that “there will be a constant probability of seeing the station filled by characters pre-eminent for ability and virtue.”

Most important, this argument about presidential character is also, by its own terms, an argument about the people’s character—or, at least, the character of a coalition of voters sufficient to elect a president.

And that is where the question of President Trump’s pardons should (and almost surely will) fall: to the people themselves. If President Trump uses his constitutional prerogative in the service of rewarding political allies or in service of protecting himself by protecting those who might otherwise testify against him, then his conduct will be judged by the voters. Courts are ill-suited to the task of deciding whether President Trump’s use of the pardon power is “constitutional” in legalistic terms—which are, of course, the only constitutional terms that courts can apply. But the people are necessarily well-suited to the task of deciding whether Trump’s use of the pardon power is “constitutional” in broader terms of ethical and moral value judgments. Indeed, these are constitutional judgments that only the people, in the long run, can decisively adjudicate: through elections and the lessons that history draws from those elections.

This point is especially acute on the question of presidential self-pardon. As Harvard’s Jack Goldsmith observed recently at Lawfare, “No president has ever tried to self-pardon, constitutional text does not speak overtly to the issue and there is no judicial precedent on point.” Despite what confident advocates might assert, “there is no obvious right answer on the validity of self-pardons, and if Trump becomes the first president to pardon himself, a court will probably not provide an answer.”

Such skepticism of the possibility of a judicial resolution surely dissatisfies the president’s critics but it may also dissatisfy much of the public at large. Whatever one thinks of this president, most neutral observers would be disconcerted by a president’s using the pardon power to immunize himself against the law and thus to put himself above the law. So there is a good reason no president has ever attempted it. But that is not to say there is good reason for courts to attempt to nullify presidential self-pardons.

Legal scholars and lawyers can conjure legalistic arguments against self-pardon—such as the general notion that “no one may be a judge in his own case,” as the Justice Department’s Office of Legal Counsel stressed in a cursory 1974 opinion arguing against presidential self-pardon. But that general notion is itself not actual law but a value judgment being grafted on to the Constitution’s actual text—and thus it provides little or no basis on which the Court can adjudicate cases while retaining its political legitimacy. Perhaps a single judge would be willing to go beyond the constitutional text and invoke this nontextual principle against the president (either out of conviction or partisanship)—but it seems unlikely, and justifiably so.

The more likely and more appropriate scenario is the one envisioned by The Federalist—namely, that the hardest questions of presidential pardoning power will and should be adjudicated, in the end, by the people themselves in subsequent elections. Or, in the meantime, by Congress, through impeachment.

Which leads us to the next part of our complicated constitutional moment.

“If a president pardoned himself,” a member of Congress recently observed, “the [House] Judiciary Committee would probably be bound to hold an impeachment inquiry on that and decide what to do based on the testimony that was presented at the inquiry.”

These were not the words of a partisan Democrat. They were the words of Rep. Jim Sensenbrenner, a Republican, in June 2018. Sensenbrenner is no stranger to impeachments—in 1999, he helped lead the House Republicans’ prosecution in the Senate impeachment trial of President Bill Clinton. And he’s not alone. President Trump’s friend and adviser Chris Christie agrees. “If the president were to pardon himself, he’ll get impeached,” the former governor told ABC’s George Stephanopoulos in June.

Many Democrats agree, to say the least. And pro-impeachment Democrats are not limiting themselves to just this one argument. From nearly the outset of President Trump’s term in office, Democrats in and out of Congress have called for impeachment. In May 2017, just four months into the Trump presidency, Rep. Al Green called for Trump’s impeachment: “President Trump has committed an act for which he should be charged by the U.S. House of Representatives,” Green announced in a press release. “The act is the obstruction of a lawful investigation of the President’s campaign ties to Russian influence in his 2016 Presidential Election.”

In December 2017, Rep. Green drafted formal articles of impeachment against Trump, expanding the justification to include everything from the president’s executive order limiting entry into the United States by persons from specified nations, to the president’s weak response to the Charlottesville white-supremacy riot, to his criticism of former NFL quarterback Colin Kaepernick. (The House voted overwhelmingly to table Green’s resolution, 364 to 58.)

Meanwhile, California billionaire Tom Steyer has undertaken a national media campaign demanding that Congress impeach the president. His website—NeedToImpeach.org—lists eight “impeachable offenses” he says Trump has committed, ranging from obstruction of the FBI’s investigation of Michael Flynn, to violation of the Constitution’s foreign emoluments clause (by making money from foreign officials staying at his hotels), to “conspiring with others” to facilitate Russian interference with our election, to even “undermining freedom of the press.”

As with the pardon power, the Constitution’s impeachment power is phrased broadly: “The President . . . shall be removed from Office on Impeachment for, and Conviction of, Treason, Bribery, or other high Crimes and Misdemeanors.” The first two terms—“treason” and “bribery”—are familiar and reasonably specific. But we cannot say the same for “high crimes and misdemeanors.” As Cass Sunstein writes in Impeachment: A Citizen’s Guide (2017), the phrase is “opaque”—it “does not have a self-evident meaning.”

Madison’s notes from the 1787 constitutional convention recount a brief exchange that offers some guidance, but not much. In September of that year, as the convention came to a close, George Mason invoked Warren Hastings’s scandalous misconduct in managing the East India Company, for which Hastings had been arrested and faced an impeachment trial only months earlier. Reminding his fellow delegates that Hastings’s transgressions fell short of “treason,” Mason objected to the initial version’s limit of impeachment to only cases of “treason” and “bribery,” which “will not reach many great and dangerous offenses.” Instead, he urged the convention to add “maladministration” to the list of impeachable offenses.

Madison objected, warning “so vague a term” would effectively turn the new constitutional system into a parliamentary one, by rendering the presidency “equivalent to a tenure during pleasure of the Senate.” Mason withdrew “maladministration” and instead proposed (per Madison’s notes) “other high crimes & misdemeanors.”

Hamilton added further insight in Federalist 65 on the Senate’s power to try impeachments. Writing, “The subjects of its jurisdiction are those offenses which proceed from the misconduct of public men, or, in other words, from the abuse or violation of some public trust,” Hamilton argued that they “are of a nature which may with peculiar propriety be denominated political, as they relate chiefly to injuries done immediately to the society itself.”

In his own history of impeachments (written before he presided over President Clinton’s), Chief Justice William Rehnquist characterized the Framers’ choice of broad constitutional words as committing the issue to the discernment of posterity. “The framers were sufficiently practical to know that no charter of government could possibly anticipate every future contingency,” Rehnquist wrote in Grand Inquests (1992), “and they therefore left considerable room for ‘play in the joints.’ Nor did they try to foresee exactly how each of the many powers and checks and balances they conferred and established would work out in particular situations.”

Rather, Rehnquist wrote, such questions were “of necessity left to future generations.” It played out in the unsuccessful impeachments of Justice Samuel Chase (in 1805) and President Andrew Johnson (1868), the two main subjects of Rehnquist’s study, and in the impeachments and intended impeachments of other officers, judges, and even Presidents Nixon and Clinton. (Rehnquist himself did believe that, in Nixon’s case, “the counts relating to obstruction of justice and to the unlawful use of executive power were of the kind that would surely have justified removal from office.”)

Legal scholars have attempted to translate “high crimes and misdemeanors” into a more precise framework. In Impeachment: A Handbook (1974), for example, Yale’s Charles Black argued that the constitutional term must not be taken literally to include only actual “crimes,” because it would fail to include presidential impropriety that does not violate criminal laws—e.g., a hypothetical president who announces “that he would under no circumstances appoint any Roman Catholic to office,” in clear violation of the Constitution’s prohibition of religious tests for office. Nor, according to Black, can impeachment be triggered by any and all criminal violations—e.g., it would be “preposterous” to suggest that a president could be impeached for merely “assist[ing] a young White House intern in concealing the latter’s possession of three ounces of marijuana,” even though the president would have been “guilty of ‘obstruction of justice.’ ” Black argued that impeachment should cover only “serious offenses against the nation or its governmental and political processes, obviously wrong, in themselves, to any person of honor”—or, “those offenses which are rather obviously wrong, whether or not ‘criminal,’ and which so seriously threaten the order of political society as to make pestilent and dangerous the continuance in power of their perpetrator.”

Black’s account of the Constitution’s impeachment standard may or may not make sense—but it certainly is not “law” in any sense. It is the best approximation of a constitutional scholar.

There are others. Laurence Tribe and Joshua Matz, in To End a Presidency (2018), argue that impeachable offenses should “involve corruption, betrayal, or an abuse of power that subverts core tenets of the U.S. governmental system. They require proof of intentional, evil deeds that risk grave injury to the nation,” and must be “so plainly wrong by current standards that no reasonable official could honestly profess surprise at being impeached.” Again, this may well be a sensible standard.

But to demand legal precision on this point is, in fact, to demand too much. This was Hamilton’s key warning in Federalist 65. In explaining why the trial of impeachments was committed to the Senate instead of the Supreme Court, Hamilton stressed that the trial of impeachments would not be a purely legalistic affair—it “can never be tied down by such strict rules, either in the delineation of the offense by the prosecutors, or in the construction of it by the judges, as in common cases [in courts].” And, he further added, “There will be no jury to stand between the judges who are to pronounce the sentence of the law, and the party who is to receive or suffer it,” which is to say that the Senate itself must play both roles, reaching decisions that grapple (as judge) with considerations of law, and (as jury) with considerations of justice and mercy. Altogether, Hamilton concludes, the Senate must wield a truly “awful discretion.”

It wields that discretion, ultimately, subject to the judgments of the people. This, too, was a point that Hamilton emphasized. Impeachment would be initiated by the House, the people’s “immediate representatives.” The Senate, like Britain’s House of Lords, would “decide upon it.”

Hamilton urged that the Senate, unlike the Court, would be capable of “reconciling the people to a decision” at odds with a prosecution brought by their immediate representatives in the House. That is the key point: The legitimacy of an impeachment and trial would ultimately be judged by the people themselves. The people, not judges or congressmen or senators or law professors, are the ones ultimately responsible for concluding what constitutes “high crimes and misdemeanors,” as the public eventually judges the actions of the House and Senate.

This does not mean that impeachment does not call for legal judgments or that it is an act of sheer politics. Rather, impeachment does call for legal judgments, but not legalistic judgments. It calls for a decision of what “high crimes and misdemeanors” rightly means—a decision made not by lawyers and judges, but by the House and Senate, and then by the people to whom they answer.

But in all likelihood, the House and Senate will not even begin to grapple seriously with these issues until another man concludes his own assigned task.

Of those of all the players on the present constitutional stage, Robert Mueller’s directions might seem to be the most straightforward. His authority as special counsel derives exclusively from the Justice Department’s regulations providing for the designation of special counsels, set forth in Title 28 of the Code of Federal Regulations. Specifically, Section 600.4, which provides:

The jurisdiction of a Special Counsel shall be established by the Attorney General. The Special Counsel will be provided with a specific factual statement of the matter to be investigated. The jurisdiction of a Special Counsel shall also include the authority to investigate and prosecute federal crimes committed in the course of, and with intent to interfere with, the Special Counsel’s investigation, such as perjury, obstruction of justice, destruction of evidence, and intimidation of witnesses; and to conduct appeals arising out of the matter being investigated and/or prosecuted.

Pursuant to that regulation, acting attorney general Rod Rosenstein defined Mueller’s jurisdiction as follows:

Rosenstein’s order to Mueller adds, “If the Special Counsel believes it is necessary and appropriate, the Special Counsel is authorized to prosecute federal crimes arising from the investigation of these matters.”

On its face, Rosenstein’s order seems much more precise, and thus much simpler to administer, than the constitutional provisions authorizing the president to grant pardons and the House and Senate to conduct impeachment proceedings. But this is illusory. Rosenstein’s specification of what Mueller may do does not dictate what Mueller must do or should do.

This question of discretion is evident in the explicit authorization for Mueller to bring prosecutions that he believes “necessary and appropriate”—two terms that implicate Mueller’s own constitutional, ethical, and prudential judgments. For all the lawyerly ink being spilled over the question of whether a president “can be” subpoenaed to testify or indicted, none of it can point to the answer of whether a president “should be” subpoenaed or indicted—or, to phrase it in terms of Mueller’s authorization order, whether it would be “appropriate” for a special counsel to subpoena or indict a president.

In some ways, prosecutors must answer “should” questions all the time. They are not unique to the special counsel order; they inhere in the work of all prosecutors. The prosecutor’s job is to enforce laws, but the task itself requires the prosecutor, in choosing how best to enforce the laws, to look beyond the law itself for values to guide his own exercise of the vast discretion afforded to prosecutors.

“Prosecutorial discretion involves carefully weighing the benefits of a prosecution against the evidence needed to convict, the resources of the public fisc, and the public policy of the State,” the Supreme Court emphasized in a 2014 decision. Or as the Court put it in 1985, a prosecutor exercises discretion from case to case with an eye to “[s]uch factors as the strength of the case, the prosecution’s general deterrence value, the Government’s enforcement priorities, and the case’s relationship to the Government’s overall enforcement plan.” These factors “are not readily susceptible to the kind of analysis the courts are competent to undertake.” Simply put, all of these are value judgments, not questions of guilt or innocence, legality or illegality.

No one has captured this better than attorney general (and later Supreme Court justice) Robert Jackson, in a 1940 address to an annual meeting of U.S. attorneys, aptly titled The Federal Prosecutor. “The prosecutor has more control over life, liberty, and reputation than any other person in America,” Jackson observed. “His discretion is tremendous.”

The prosecutor can investigate. And, “if he is that kind of person,” he can ruin a target’s reputation with suggestive public statements. He can pursue the target’s friends and associates; he can haul them before secret grand juries, where he can secure indictments with one-sided accounts of the facts. “While the prosecutor at his best is one of the most beneficent forces in our society,” Jackson explained, “when he acts from malice or other base motives, he is one of the worst.” True, a prosecutor may find some difficulty in winning the conviction and punishment of an innocent man; but there is a lot of ruin that he can bring, short of conviction or indictment.

Precisely because of “this immense power to strike at citizens . . . with all the force of government itself,” Jackson continued, the office of U.S. attorney “from the very beginning has been safeguarded by presidential appointment, requiring confirmation of the Senate of the United States.” And though a U.S. attorney is afforded great leeway in directing the work of his office in his given federal district, there are limits to the amount of discretion that can rightly be vested in his office. “Experience,” Jackson explained, “has demonstrated that some measure of centralized control is necessary.” The attorney general did not say aloud what that “centralized control” was, but he didn’t need to. He was referring to the Justice Department’s Washington headquarters, operated under control of the attorney general himself, accountable to the president—who is in turn accountable to the people. “We must bear in mind,” the U.S. Court of Appeals for the Seventh Circuit once observed, “that the United States Attorney is an officer of the executive branch responsible primarily to the President, and, through him, to the electorate.”

As a matter of constitutional first principles, the federal prosecutor’s ultimate accountability to the president reflects the Constitution’s imposition of a duty on the president to “take care that the laws be faithfully executed,” and his constitutional oath to “faithfully execute the Office of President of the United States.” It also reflects what Hamilton identified in Federalist 70 as the need for “unity” in the executive branch, which would facilitate the executive’s “energy,” which was itself “essential to the steady administration of the laws” and thus “a leading character in the definition of good government.” And in the end, Hamilton famously wrote, the executive branch’s “unity” ensured its “responsibility”—that is, the president’s ultimate and singular accountability to the people as a whole.

For Mueller, however, there is at least some relaxation of the prosecutor’s accountability to the president. Indeed, that is the whole point. As the Clinton Justice Department explained in the Federal Register announcing a new framework in July 1999, the special counsel’s office reflects “a balance between independence and accountability in certain sensitive investigations,” such that the special counsel enjoys “day-to-day independence” from the attorney general and president, “free to structure the investigation as he or she wishes and to exercise independent prosecutorial discretion to decide whether charges should be brought,” subject only to the Justice Department’s generally applicable standards and procedures. The attorney general nonetheless retains “ultimate responsibility,” and thus there remains the “possibility of review of [the special counsel’s] specific decisions” by the attorney general.

These measures of independence fall short of the now-expired independent counsel statute that governed the controversial investigations of the 1980s and 1990s (including Ken Starr’s Whitewater investigation, culminating in President Clinton’s impeachment) and that passed constitutional muster with the Supreme Court in Morrison v. Olson (1988). Indeed, there are open questions about how “binding” these regulations actually are: Because they were rules applying to agency organization and personnel, they were exempt from the usual notice-and-comment procedural requirements and thus can be rescinded without an opportunity for public comment just as they were promulgated without it. And because they are merely an agency’s own rules, not a law passed by Congress, there is an open question of whether they can bind the president at all.

But no matter how thorough the special counsel’s independence is, the regulations creating a stand-alone office of special counsel for case-specific prosecutions inherently relax at least some of the institutional forces that circumscribe a prosecutor’s exercise of discretion. For example: Because the special counsel is focused on a narrow set of cases, not a broad and active docket of different cases competing for fiscal and personnel resources, he can approach his narrow set of targets with a significantly different mindset. Justice Scalia highlighted this in his famous Morrison dissent:

And so when special counsel Mueller faces the prudential choices that any prosecutor must make—including decisions of whether a particular indictment is “necessary and appropriate,” he and his office will make those decisions without the institutional influences and presidential accountability that normally surround and oversee a prosecution. Again, that’s the point, and the benefits of this independence in cases implicating the executive branch itself are freely conceded. But we must not forget that those benefits come at a cost.

These choices will be made, in the first instance, according to Mueller’s own legal, ethical, and prudential judgments. And while those cheering him on often assume that Mueller can and will press those powers to the greatest possible extent, there is reason to think or hope that he will restrain himself and thus embody the same qualities of character that Hamilton ascribed to the executive in Federalist 68.

For example, many of President Trump’s critics argue that Mueller can subpoena the president to testify, or even indict the president. There is great reason to doubt such assertions: Stuart Taylor and I outlined the possible constitutional limits on Mueller’s subpoena power in these pages (“Privilege and Precedent,” May 11, 2018); the Justice Department’s Office of Legal Counsel outlined the possible constitutional limits on the indictment power in a 2000 opinion memorandum.

But more immediately, there is reason to expect Mueller himself to stop short of testing those constitutional limits. First of all, there is “litigation risk.” But second, and more important, he might simply restrain himself from fully asserting his powers against the president, except where strictly necessary.

Lawrence Walsh, independent counsel for the Iran-Contra investigation, suggested such measures of self-restraint in a 1998 hearing of the House Judiciary Committee. “[W]‌hat restraint should be expected of a prosecutor investigating the president of the United States?” he asked. “Should a president be subjected to the same tactics as a racketeer or a corrupt municipal official?” Walsh, though hardly known for self-restraint, conceded the need for prudence. “In the national interest, the president should probably be entitled to greater protection from longshot efforts to prove a minor crime,” he explained. “Should a president, carrying out the most demanding and important job in the world, have to guard himself against such tactics”—namely, “sting” operations and “cooperating informer[s]”—“when dealing with his confidants—or with his friends?”

Perhaps Mueller is grappling with these questions himself. While Trump opponents around the country seem keen to condemn every new move by the president as another indictable “obstruction of justice,” will Mueller—the man singly responsible for the investigation and eventual prosecutions—be similarly eager to press his powers to the fullest?

No matter what Mueller ultimately decides, we too may have to grapple with these questions. If Mueller attempts to prosecute the president (perhaps leading to an attempted self-pardon), or if Mueller delivers to Congress a report that spurs the House to begin impeachment proceedings, the ultimate judgments will be delivered by the American people through the elections that keep the president and members in Congress or cast them out of office.

These issues of constitutional structure and power, like those of pardon and impeachment, will ultimately be adjudicated by the people themselves, within the constitution’s structure and powers. That is our true constitutional moment.

We rightly think of our courts as the final voice in the interpretation of our Constitution,” Chief Justice Rehnquist observed in the closing lines of Grand Inquests, “and therefore tend to think of constitutional law in terms of cases decided by the courts.” But the actual history of impeachments, he wrote, especially those of Justice Chase and President Johnson, should teach us otherwise: “These two ‘cases’—decided not by the courts but by the United States Senate—surely contributed as much to the maintenance of our tripartite federal system of government as any case decided by any court.”

Rehnquist was right to focus our attention beyond the courts. But we should look further still—beyond the Senate, to the American people who received the Senate’s decisions in those cases and ratified them.

If the American people had loudly rejected the Senate’s choices in the Chase and Johnson impeachments, by punishing the senators in subsequent elections and pressing their replacements to continue the attacks on Federalist judges (in Chase’s time) or to continue the attacks on President Johnson and other impediments to the Radical Republicans’ Reconstruction, history would have drawn very different lessons from the two impeachments.

Instead, after Justice Chase’s failed impeachment, the public did not punish the moderate Republicans, and the lack of appetite to use impeachment as a tool to punish Federalist judges soon became evident, as Richard K. Neumann Jr. found in his 2007 study of impeachments. And this effectively settled the constitutional question; in an 1822 letter, Jefferson would complain that life-appointed judges were “responsible to no authority . . . for impeachment is not even a scarecrow.”

Rehnquist emphasizes the point. “The importance of these acquittals,” he wrote, “can hardly be overstated.” They firmly established that

“[i]mpeachment would not be a referendum on the public official’s performance in office,” but a more focused and formal inquiry. And for the Senate to remove an impeached judge or president would require not just “any technical violation of the law,” but something much more significant.

Chase’s impeachment trial had further, subtler constitutional ramifications. As Neumann explains, Federalist judges recalibrated their own judicial behavior after surviving the Chase impeachment’s near-death experience (because Chase, if removed from office, would not have been the last of them to go). “Although the judiciary remained vaguely Federalist in its outlook, it abandoned much of the partisan behavior that so incited the Jeffersonians.”

Neumann then helpfully compares this to a more recent episode: “The Chase impeachment, though unsuccessful, may have had an effect on the federal judiciary similar to the one that the 1936 election and Franklin D. Roosevelt’s court-packing proposal may have had on the Supreme Court.” Which is to say, FDR’s court-packing plan, after being overwhelmingly rebuffed by voters in 1936, was a moment that in hindsight entrenched a constitutional norm against partisan gamesmanship with the Court’s own structure.

We can think of other modern examples in the same vein. The Supreme Court declared the independent counsel statute constitutional in Morrison v. Olson, over Scalia’s vocal dissent, but the subsequent experience of independent counsel investigations, culminating with the failed impeachment of President Clinton, seemed to deliver a much more consequential political verdict against the independent counsel. So Congress allowed the statute to expire. The Independent Counsel Act may have been declared “constitutional” by the Supreme Court, but its imprimatur seems much less credible today. Congress, with the assent of the people, effectively sided with Scalia’s dissent.

We do not often think of political showdowns in terms of constitutional “precedents”; we tend to focus on constitutional questions that are “settled” by the Supreme Court. Perhaps we think we lack a vocabulary or framework for this sort of resolution. But this is a mistake. We need to think of these things in the way that James Madison did in Federalist 37, that of “liquidation.”

Madison, writing of the then-proposed Constitution’s vague, ambiguous, and otherwise unclear terms, warned that there is only so much precision to be attained in any written law—at some point, we must rely on lived experience to clarify the rest. “All new laws, though penned with the greatest technical skill, and passed on the fullest and most mature deliberation, are considered as more or less obscure and equivocal,” Madison explained, “until their meaning be liquidated and ascertained by a series of particular discussions and adjudications.”

Here we find the crucial but overlooked key to preserving a genuinely republican Constitution: namely, an acceptance that in some constitutional disputes, the best meaning of constitutional text will be found not in the lawyerly forensics of Supreme Court arguments alone but through the lived experience of politics. We ascertain the best meaning of some constitutional terms through the constitutional processes themselves. And this is a process that ultimately is resolved by the voters, as illustrated in Larry Kramer’s seminal and suitably titled work The People Themselves (2004), as well as by University of Chicago professor Will Baude in a major forthcoming law review article.

Another significant legal scholar, Yale’s Bruce Ackerman, dedicated multiple volumes of study, in We The People, to what he called “constitutional moments.” Ackerman carried the theory too far, arguing that the political settlement of a constitutional dispute, through public debate rising above the usual partisan fray and focusing deliberately on higher constitutional principle, should be accepted as a de facto “amendment” to the Constitution. But the term Ackerman coined—“constitutional moment”—is apt.

It captures well the moment in which political events settle disputed constitutional territory, liquidating ambiguous terms. Including, perhaps, such terms as “high crimes and misdemeanors,” and such questions as whether the president’s power “to grant reprieves and pardons for offences against the United States” is best understood as incorporating principles of justice that stop short of self-pardon; or whether “the executive power” and the obligation to “take care that the laws be faithfully executed,” both committed to the president, should serve to limit prosecution of the president himself. These are all questions that political actors—Congress, the president, and special counsel Mueller—will answer in the first instance but that we the people, in our own debates and subsequent elections, will settle in the end.

The problem with the settlement of constitutional moments, however, is that we do not know precisely when the moment has started, or when it has ended, or what the actual settlement was, until much later. Those are distinctions that history draws in retrospect. If Mueller tries to subpoena the president’s testimony or even prosecute him, or if the president pardons himself, or if the House impeaches the president and the Senate adjudicates it, then history will draw lessons of constitutional import from the political resolution that follows, just as it has done from past impeachments, and court-packing plans, and other constitutional moments. But only in time.

For us the challenge is not to treat these issues as matters left to courts alone but to engage these deliberations as genuine constitutional decision-makers—which we may turn out to be in hindsight, once the constitutional moment has passed.