Mark Tushnet | Vox

Is it anti-Catholic to ask Amy Coney Barrett how her religion affects her decisions?

2020-10-11T17:17:39-04:00

Judge Amy Coney Barrett meets with Sen. Josh Hawley (R-MO) (not pictured) at the Capitol on October 1. | Demetrius Freeman/Getty Images

Twenty-two years ago Amy Coney Barrett, then a law clerk soon to enter the legal academy, co-authored an article called “Catholic Judges in Capital Cases.” The topics raised were intensely debated in 2017 when she was nominated to serve on the federal bench, and are being debated again now that she has been nominated to replace Ruth Bader Ginsburg on the Supreme Court.

Barrett wrote the article with John Garvey, her soon-to-be colleague at Notre Dame’s law school (he’s now the president of Catholic University), and in it they concluded that “Catholic judges (if they are faithful to the teaching of their church) are morally precluded from enforcing the death penalty.” The article is a model of serious scholarship — analytically precise, willing to take clear positions on important and controversial questions.

It became the focus of some quite misleading discussion during Barrett’s 2017 confirmation hearings, and is what led Sen. Dianne Feinstein (D-CA) to make her now notorious comment, “The dogma lives loudly within you” — a comment that has led some of Barrett’s supporters to characterize opposition to her nomination as religiously motivated. That controversy aside, does the article raise questions that might fairly be explored in a nomination process — without lapsing into anti-Catholicism? I believe so.

Maybe Barrett’s views on the matters discussed in the article — which I first analyzed for Vox in 2018 — have changed since 1998, but asking her whether she still agrees with the article’s analysis seems well within bounds.

Barrett explored how Catholic judges should handle cases involving capital punishment

There’s some underbrush to clear away before we get to the heart of the question. First, the article deals with the obligations, in connection with capital punishment, of a person the authors describe as an “orthodox Catholic.” The authors equate “orthodoxy” with being “faithful to the teaching of the church on the subject.” A fair amount of the article deals with the church’s teaching on capital punishment, which is complex and not unequivocal. To somewhat oversimplify their careful analysis: They agree with the view that the church’s teaching strongly implies that administering the death penalty in the United States is immoral.

On October 3, Pope Francis issued an encyclical strengthening the teaching and making it clear that capital punishment in the United States today is a serious evil.

The matter of current interest is the obligation of an orthodox Catholic in connection with abortion, where the church’s teaching is clear and unequivocal; that removes one layer of complexity.

Second, the article argues quite correctly that the mere fact that a judge is a Catholic is not the basis for reasonable concern about his or her impartiality in death penalty cases, because the range of views among Catholics on the death penalty is rather wide, and even orthodox Catholics might not fully understand the church’s teaching. The case is different, I suggest, when one knows — or has quite strong reasons to believe — that a particular judge has orthodox Catholic views on the matter of abortion.

And that certainly seems to be the case with Barrett, whose religious convictions appear to be one of the primary selling points in her favor. They’re one reason she has such strong support among social conservatives, which in turn brought her to Trump’s attention.

She drew a distinction between two kinds of “cooperation with evil,” one worse than the other

Now, to the law review article. It draws upon long-standing Catholic moral teaching to distinguish between “formal” and “material” cooperation with evil. The authors define formal cooperation (the more direct kind) this way: “A person formally cooperates with another person’s immoral act when he shares in the immoral intention of the other.” In the article, the key example of formal cooperation is actually imposing a death sentence (as a judge might when presiding over a criminal trial).

Material cooperation, in contrast, occurs when an act “has the effect of helping a wrongdoer, where the cooperator does not share in the wrongdoer’s immoral intention.”

We’ve recently seen arguments based on the idea of material cooperation, when orthodox Catholics object to being forced to provide, or even subsidize, what they regard as abortion-related and contraception-related medical care: In their view, doing so amounts to material cooperation with evil. (What the law should do about that view is, of course, a different question, and I’m not concerned with it here.)

Barrett and Garvey’s article identifies some examples of formal cooperation in connection with the death penalty, such as actually administering the drugs at an execution and signing the document authorizing an execution — something a judge at a trial does. But the article devotes most of its analysis to questions about material cooperation, the morally grayer type.

Catholic teaching holds that formal cooperation is always immoral, but an appellate judge will almost never be faced with the possibility of formally cooperating with abortion. He or she will almost never have to sign a document authorizing an abortion for example. At most, the judge would say that the woman can have an abortion if she wants to, and that’s material, not formal, cooperation.

The article asserts, again in a way that’s consistent with Catholic teaching, that the morality of material cooperation depends upon the outcome of a “moral balancing test — weighing the importance of doing the act against the gravity of the evil, its proximity, the certainty that one’s act will contribute to it, and” — very important, I think — “the danger of scandal to others.”

Barrett asked what conclusions (moral and otherwise) reasonable observers would draw from a judge’s action

“Scandal to others” occurs when those “others” see the believer’s actions and infer that the “wrongdoing is not so wrong,” thereby “provid[ing] material for rationalization and self-deception by people tempted to undertake the same sort of wrong.”

The authors are uncertain whether an appellate judge’s sitting on a capital case amounts to material cooperation with evil because at least sometimes that judge will have to say that the law required a death sentence be imposed. But they conclude that sitting on a capital case concerning habeas corpus does not amount to material cooperation. That’s because in a habeas corpus case the issue is whether a trial judge did something that was within the range authorized by law — not whether the law required the death penalty.

So, for them, an orthodox Catholic judge should recuse himself or herself from imposing the death penalty but need not opt for recusal in habeas cases. They leave open the question about participating in direct appeals, where formal cooperation will be rare and the balancing required by the standard for material cooperation might allow participation.

They then examine whether the federal recusal statute requires in death penalty cases the recusal of a Catholic judge, merely by virtue of his or her church membership. They conclude that it does not.

The question of the “appearance of impartiality”

So what are the implications of the analysis for a judge who one knows is or has strong reasons to believe is an orthodox Catholic on the issue of abortion? The relevant recusal statute includes a provision dealing with “appearance of impartiality” and requires recusal if the judge’s impartiality “might reasonably be questioned.” The authors discuss this provision only in connection with a motion to recuse based solely on the judge’s general religious affiliation — identifying as Catholic, Jew, Mormon, or the like.

But, as I noted above, the case is different where the motion rests on assertions about a particular judge’s religious convictions rather than mere affiliations. Is it reasonable to believe that an orthodox Catholic would be subject to reasonable concerns about his or her impartiality in connection with actions that constitute material cooperation with evil?

We have to begin by observing that such a judge could conclude from his or her examination of the relevant purely legal materials that Roe v. Wade was mistaken as a matter of law and should be overruled. After all, many non-Catholic conservatives — and indeed some quite nonreligious conservatives and even some people who identify as liberals — think exactly that.

An orthodox Catholic judge’s views about material cooperation with evil might therefore play no role in his or her decision. The difficulty, though, is that someone observing the judge’s action — the reasonable person whose inferences matter when the “appearance of impartiality” standard is applied — can’t tell the difference between a decision to overrule Roe because the decision was not firmly rooted in the relevant purely legal materials, and a decision to do so because any other course would amount to material cooperation with evil, from the judge’s perspective.

Barrett and Garvey say a judge who makes a decision based on religious views when he or she believes the purely legal materials dictate a different result is guilty of “cheating.” They condemn cheating; their position is that, rather than cheat, the judge should recuse himself or herself. But, if an orthodox Catholic judge does not opt for recusal, could an observer take the possibility of cheating to indicate that the judge’s impartiality might reasonably be questioned?

When an orthodox Catholic judge sits on such a case, he or she may have antecedently concluded that affirming Roe v. Wade based on the purely legal materials would not amount to giving material assistance to the evil of abortion. But suppose the observer — again, the person we have to think about in connection with the “appearance of partiality” statute — concludes that conscientious application of the moral balancing test does mean that reaffirming Roe v. Wade would constitute giving material assistance:

The evil, from the orthodox Catholic’s point of view, is extremely evil (maybe even the ultimate evil), and though the evil is not at all proximate to the judge’s decision, that it will occur is nearly certain.

And the issue of scandal may be quite serious here: Others who see an orthodox Catholic judge affirming Roe v. Wade might conclude that abortion is not all that wrong. After all, if even such a judge can go along with a decision allowing abortions, how wrong could it be?

Is all of that enough to support the observer’s inference that the judge cannot be impartial on the relevant questions (or at least that the judge’s participation raises reasonable questions about his or her impartiality)?

Barrett and Garvey’s article has a footnote dealing explicitly with abortion, but its implications for today are not entirely clear. The text to which it is attached reads, “If one cannot in conscience affirm a death sentence the proper response is to recuse oneself.” Quoting another scholar, the footnote reads, “‘Where there is no honest, legitimate alternative for deciding the case but to follow positive law supporting the right to commit an abortion,’ the judge should recuse himself. The abortion case is a bit easier.”

The implication seems clear but not entirely helpful: An orthodox Catholic judge should recuse himself or herself if the legal materials required affirming Roe. If an orthodox Catholic judge opts instead to cast a vote to overrule Roe, though, either the judge has decided that the purely legal materials support that vote or is willing to cheat. And the problem again is that the observer concerned with impartiality can’t tell which inference to draw.

Sen. Feinstein handled the issue very clumsily, but that doesn’t mean it’s not a legitimate one

The discussion of the article during Judge Barrett’s 2017 confirmation hearing was execrable. Either Sen. Feinstein’s staffers didn’t prepare her well or she disregarded or failed to understand their advice about what precise questions were raised by the article. The article’s arguments are complex and subtle, and the confirmation process is ill-suited to addressing arguments of that sort — particularly when what’s at issue are the arguments’ implications for highly charged matters not directly addressed in the original article.

Still, the article raises questions that can be asked in a confirmation hearing without devolving into anti-Catholicism. The questions should be about how an orthodox Catholic judge’s religious commitments interact with his or her legal ones — which are exactly the questions the article itself raises. If it’s not anti-Catholic to have written the article, which it surely is not, it can’t be anti-Catholic to explore the implications of its arguments for other matters.

Yet, the issues are complex and delicate, and we should be skeptical about the possibility that a nomination hearing is a good place to explore them. With limited time to frame questions, Democratic senators would probably be well-advised to avoid asking about recusal in abortion cases.

Mark Tushnet is the William Nelson Cromwell professor of law at Harvard Law School.

What the Constitution says Berkeley can do when controversial speakers come knocking

2017-09-23T12:45:29-04:00

Protesters march off the UC Berkeley campus on February 1, 2017. A scheduled speech by the controversial author Milo Yiannopoulos was canceled after protesters and police skirmished. | Elijah Nouvelage/Getty Images

As of Saturday, it was unclear whether Free Speech week would be happening. Still, UC Berkeley, the proposed site, is bracing for the event:

“This coming week is “Free Speech Week” at the University of California Berkeley. Conservative speakers, some very incendiary, are scheduled to appear in public spaces — the storied Sproul Plaza, the nearby Mario Savio Steps — offering their views on feminism, Islam, and more.

Maybe the organizers really want to gather an audience that will listen to what the speakers have to say — the list seems quite fluid, but the names of alt-right provocateur Milo Yiannopoulos, anti-Islamic polemicist Pamela Geller, and former White House strategist Steve Bannon have been bandied about — and exit with changed minds. Mostly, though, they want to lay down a marker at what they regard as a center of intolerance for conservative views. And they probably expect some disruptions that will, they hope, discredit their opponents.

There’s a lot of misunderstanding about what public universities and cities can do when faced with requests to use their facilities for demonstrations, particularly when officials reasonably think the events might turn violent.

People offer simple-minded answers to what are, in fact, difficult questions: “Time, place, and manner regulations are okay — but not restrictions based on speech content.” Or: “You can’t give the counterdemonstrators a heckler’s veto” (that is, you can’t let them make it impossible for the demonstrators to get their message across to people who want to hear it). The former statement is true, but it’s too short on details to be a useful guide to action. The latter is debatable.

Amid the confusion, there’s room for people to cry “censorship” when none has occurred. My prediction: The odds are slim that there will be real violations of the speakers’ constitutional rights next week, but chances are high that conservatives will say that the university has violated the First Amendment, confirming its reputation as a hotbed of liberal intolerance.

Universities have considerable leeway in deciding when, and where, speakers may appear

When you walk through the First Amendment rules, things get complicated very fast. Sophisticated lawyers like those Berkeley and larger cities have can usually work their way through the constitutional maze, though sometimes even they will be blindsided by unexpected developments. Smaller campuses and towns are more likely to misstep — largely unintentionally, I think — and provide fuel for conservative attacks on the purported “suppression” of free speech.

So let’s walk through those rules.

Suppose a public university gets a “request” from a group that the group wants to invite a speaker on a specific date, to appear at a particular auditorium, or a city gets a request to hold a demonstration in a specific city park. (Concerning universities, the rules only apply to public institutions, as private colleges can set their own policies, although many say they try to do what the First Amendment requires of public institutions.)

The first reaction to the request has to be, “Sure, in principle. But we have to think about some things before we can sign off on this.” If a public university allows student groups to invite outside speakers, it can’t pick and choose based on how offensive the speaker is perceived to be. Likewise, all demonstrators have a (presumptive) right to use the streets, park, or auditorium — which is why they aren’t really making a request.

What sorts of things do officials who get the request have to think about? Suppose the demonstrators say they want to hold a fairly large demonstration on Main Street during rush hour. The officials can say, “Sorry, that’s going to be too disruptive. If you want Main Street, you’ve got to clear out by 3 pm. And if you want rush hour, you have to use a street a couple of blocks over.” Same for using heavily trafficked areas of the university, or — of course — auditoriums in which classes are already scheduled.

Notice two things about this. The officials telling the organizers to move or reschedule their event aren’t basing their decision on what the demonstrators are going to say. In the jargon, their approach is “content neutral.” As far as they’re concerned, they’d say the same thing to Black Lives Matter, the Republican Party, the Democratic Party, and Richard Spencer, the white supremacist leader.

Large universities and cities with lots of experience probably have content-neutral rules in place about when and where demonstrations can occur. Other places might not, but will say that they’ve come up with rules on the spot that are good enough.

Of course, were he denied permission to demonstrate where and when he wants, Spencer might be suspicious about the officials’ claim that they would say no to the Democratic Party too. Fair enough. That’s what hearings before a judge are for; the officials will testify, and a judge will decide whether she believes them.

The second thing about the officials’ response is that if they say no regarding a proposed time and venue, they have to offer a reasonable alternative. Of course the demonstrators will think —correctly, from their point of view — that the alternative isn’t quite as good: Fewer people will see a demonstration at a plaza outside the engineering school than at Sproul Plaza, for example.

Still, if the city’s alternative is a reasonable one, that’s fine by the Constitution. (Again, “reasonable” is up to judges. Every four years the cities where national political conventions are held set aside specific areas for demonstrations against the parties, often not all that close to the convention site, and courts have routinely found these alternatives reasonable.)

Berkeley has the right to set rules that minimize the risk of violence

Avoiding violence has become an important concern for universities and cities. If there’s time, officials can ask organizers to provide a list of invited speakers and a guess about the number of people the organizers expect to attend — not to say yes to one speaker and no to another, but to figure out how many police officers to deploy.

There are two wrinkles here. With events planned in advance, like Free Speech Week, the university can ask for a fair amount of information. But sometimes demonstrations are more spontaneous, as we saw in St. Louis recently, after a white former police officer was acquitted in the fatal shooting of a black man. Advance planning is impossible, and the First Amendment requires cities and universities to be more flexible.

The other wrinkle involves paying for the security. In principle, the First Amendment allows the university to charge the organizers for additional security such as police overtime pay — if they can afford to. It would be silly to say that a city can’t charge Al Gore and his friends for the costs of cleaning up after a demonstration supporting the Paris climate accord. Courts have been pretty careful, though, to insist that cities do a good job of figuring out what those additional costs are, because the courts are correctly concerned that cities will manipulate the charges to make it too expensive for the organizers to hold a controversial demonstration.

If the organizers can’t pay the additional costs, most students of the First Amendment think that the city or the university has to eat the costs. The University of California has decided not to charge the organizers of Free Speech Week anything, more a pricey formula for avoiding litigation than a constitutional requirement.

One implication of security considerations is that someone who wants to use the auditorium on a specific Tuesday in October can’t complain about a violation of constitutional rights if the university says, “Sorry, we can’t organize security fast enough for that, but four weeks later is fine with us.” It doesn’t matter if the new date doesn’t coincide with the speaker’s book tour. (But, again, the speaker can challenge the claim that it’s really too difficult to set up security on her preferred date.)

So content neutrality and reasonable alternatives are the starting point in discussions about controversial speakers and demonstrations. It would be really nice if the city or the university had rules in place about such matters before the requests came in, so observers could have confidence that the rules weren’t jerry-rigged simply to get rid of this particular demonstration. But the world changes, and lawyers for cities and universities can’t anticipate every variation that might pop up. So it’s not a conclusive argument against moving a speech that the decision didn’t come from some rule that was in place before the request was made. And, again, places with more experience are likely to do better on this score — though even Berkeley had to revise its already reasonably good rules over the summer.

There has already been skirmishing over scheduling and the supplying of speakers’ names at Berkeley: The organizers wanted to rent several auditoriums, but they missed the deadlines the university set. And as of this week, the university said they hadn’t confirmed the exact lineup of speakers, which Berkeley says is important so it can make security plans.

If events on the ground change, Berkeley can change the rules on the fly too

The recent deadly white supremacist rally in Charlottesville, Virginia, revealed another potential complication: Events don’t always play out as predicted. Suppose the speaker says, “We expect that there will be a couple of hundred people at the demonstration.” Taking the applicant at his word, the city replies, “Okay, you can hold the demonstration where you want to.” Then it becomes apparent that lots more people are going to attend. Maybe a few of the new ones are unexpected supporters, but suppose most of them are going to disagree with the speaker.

Content neutrality means that the city can’t make its regulation of the demonstration dependent on distinguishing between people who support the speaker and those who oppose him. The opponents might swell the crowd and listen respectfully, or boo only occasionally. From the city’s point of view, all it cares about is size and its implications for security.

Maybe a small demonstration would be fine in a small park but a larger one wouldn’t. The Constitution lets the city adjust its regulation if circumstances change. (When Charlottesville tried to switch its permission from one park to another, a federal judge held a hearing and found — probably mistakenly, as things turned out — that the city hadn’t shown that circumstances had changed.)

The city can probably tell the demonstrators and the counterprotesters that they can’t carry weapons in or near the demonstration’s site, although that’s more a question of Second Amendment law and state law than First Amendment law.

All this is pretty basic, at least to those familiar with First Amendment law. Problems come up at the next stage, though. Suppose the speaker is going to say things that are going to anger a lot of listeners — whom we now can call counterdemonstrators rather than attendees. We’re now outside of “content neutrality land.” But we have to move carefully here.

What speakers can and cannot say

Begin with two easy cases. If the speaker shouts to his supporters, “Beat that motherfucker up!” the police can move in and arrest him. He’s “inciting imminent lawless action,” as the First Amendment rule puts it. (A federal judge recently held that candidate Donald Trump might have done just that when he said at a rally, “Get ’em out of here” — referring to protesters.)

But if the speaker says only, “Beat up every black man you see after you leave this demonstration,” the police can’t do anything to him. There’s a threat of violence, but it’s not imminent. (The Supreme Court held that the government couldn’t punish a speaker who shouted to a crowd being pushed off the street by the police, “We’ll take the fucking street later,” because the word “later” took the case out of the imminence category.)

The theory is twofold: Someone who’s inclined to go along might change his mind before he runs across a black man. And if someone does beat up a black man, the government should punish the actual assailant, not the speaker.

There’s another case that I think is easy one way, although I know lots of people who think it’s easy in just the opposite way: when a raucous crowd shouts down the speaker. A report from the Brookings Institution last week describes as troubling the fact that a narrow majority of students think that’s okay.

As far as I’m concerned — and, I think, as far as the First Amendment is concerned — it is okay. The jeerers are simply people attending the rally, no different from the supporters who cheer the speaker. It just so happens that the opponents vastly outnumber, or at least outshout, the supporters.

The opponents aren’t the government, so even if they prevent the speaker from getting his message across, that’s just too bad — or it’s speech countering speech.

I suppose you could say that the First Amendment gives the government a duty to make sure that the speaker is able to get his message across. But that’s implausible as a general principle. I have a lot of things I’d like to have lots of people hear, but I can’t dragoon the government into helping me get my message to them. Maybe you can figure out why the government has a duty in the context of demonstrations but not in the context of my political views, but I haven’t yet seen anyone do so effectively.

That’s not to say that shouting down a speaker is a good idea. I think it’s sometimes worth doing, but not often, and maybe universities should have unenforceable “civility” guidelines counseling against it. The First Amendment, though, doesn’t say anything either way about heckling.

The police must target lawbreakers first, but they can shut down a speaker to stop violence

There is one government duty, though: The government has to protect speakers against violence directed at them. This is the real problem when people talk about a “heckler’s veto,” but it’s not a problem about heckling. It’s a problem about violence.

Oddly, the few relevant Supreme Court decisions — relatively old ones — suggest that the government can “protect” speakers against violence by arresting them, not the people who are threatening them. Since they were decided in the middle of the last century, though, a strong consensus among First Amendment scholars has developed that points the other way: If a speaker is likely to say things that will provoke listeners to attack him, the government’s (or Berkeley administrators’) first response has to be to put the police between the speaker and the angry crowd.

The response if violence erupts will be shaped partly by whether the government has properly gauged the size and unruliness of the protest, and sent enough officers.

Even if it has, there are limits to what the police can do. A small public college’s police force might be overwhelmed by a large, angry crowd. So might Berkeley’s. Even getting support from city police and state police agencies might not be enough, since every officer devoted to protecting the speaker is an officer who isn’t patrolling the city to prevent crime or make sure traffic is flowing freely elsewhere in the city.

If violence breaks out, the job of the police is to stop it. And the first way to stop it is by arresting the lawbreakers, whether they are counterdemonstrators or people supporting the speaker.

What if the police are overwhelmed and can’t stop the violence by arresting everyone who’s throwing stones or pointing guns? If it turns out that the police are overwhelmed by the violence, they can shut down the demonstration. If the speaker resists, the police can arrest the speaker as well as violent demonstrators and counterdemonstrators. “Can” here means: without violating the First Amendment. Under the First Amendment, the priority is stopping the violence, not stopping the speakers, which should be viewed in this context as an unwanted side effect.

At some point, however, violent clashes in the street — of the sort that we’ve seen in Berkeley already, and in Charlottesville — threaten to render legalistic discussion of the First Amendment beside the point.

When the arrival of controversial speakers provokes large-scale violent conflict, beyond the capacity of the police to control — and the speakers themselves aren’t even “inciting” this violence — we’re in pretty bad shape as a society. We probably should be worrying about more than free speech.

Mark Tushnet, a leading scholar of constitutional law and legal history, is the William Nelson Cromwell professor of law at Harvard Law School. He contributes to the blog Balkinization.

The Big Idea is Vox’s home for smart discussion of the most important issues and ideas in politics, science, and culture — typically by outside contributors. If you have an idea for a piece, pitch us at thebigidea@vox.com.

The US is terrible at investigating politicians. Blame the Constitution.

2017-05-17T10:53:16-04:00

The former president of Brazil, Luiz Inacio Lula da Silva, faces press questions about a corruption scandal, in March. | Victor Moriyama / Getty

Brazil has Operation Lava Jato (Car Wash), an expansive investigation into bribery and corruption that has dragged dozens of legislators — and former President Lula da Silva — into court. South Africa’s President Jacob Zuma has been ordered to repay hundreds of thousands of dollars spent to upgrade his private residence, supposedly to enhance its security — though the improvements included a swimming pool and a chicken run.

Brazil uses “investigating judges” in way that’s common in nations whose legal systems track those on the continent of Europe. A judge gets hold of a case — in Brazil, the current operation began as a small money laundering investigation — and then can pursue it to the judge’s heart’s content. South Africa uses an independent auditor to follow the money, and a prosecutor with wide powers to investigate high-level corruption who is protected from politics by the nation’s Constitutional Court.

Both systems have their flaws. In both nations, there have been credible claims that the investigations have become weapons in partisan battles — wielded by conservatives against the leftist supporters of former Presidents Lula da Silva and Dilma Rousseff, and by opponents of Zuma’s faction within the governing African National Congress.

Designing investigations into high-level misconduct is extremely difficult. Every nation has tough choices to make, and none has come up with a perfect solution — though it’s clear that the US system is uniquely bad.

The problems Donald Trump has created for himself have put into motion the US version of the investigations in Brazil and South Africa. We are watching our political system contort itself as politicians and bureaucrats search for a credible way to investigate the depth and nature of his campaign’s connection with the government of Russia.

None of the choices we have will solve everything. Will the new FBI director have the credibility and clout to run the investigation? Should he or she, instead, insist that Attorney General Jeff Sessions appoint a special prosecutor — who would still report to Sessions? If Congress opts instead for an independent commission, it would have to decide how to choose members, and how much power the commission would have to force people to talk if they didn’t want to.

The US system for investigating high-ranking officials seems all but designed to generate political distrust

A well-designed investigative system must ensure that the public has confidence in the investigation and its outcome. But if the Justice Department decides not to bring charges against leading political figures — such as, let’s say, Hillary Clinton — lots of Republicans and lots of Democrats have to be persuaded that it’s not a cover-up. (Remember Michael Flynn leading the “lock her up” chant?)

If the prosecutors go ahead with charges, Republicans and Democrats have to be convinced that it’s not a partisan witch hunt. (Think Kenneth Starr here, who ran across Bill Clinton’s sexual adventures while investigating what turned out to be a nothingburger about the Clintons’ real estate investments.)

To be removed from office, Trump would have to be impeached by a majority of the House of Representatives, and then put on trial in the Senate, where two-thirds would have to vote to convict him. And according to the Constitution, impeachment leads to removal but doesn’t automatically produce a criminal conviction. Formally, a president who is removed from office via impeachment could be charged in a separate criminal proceeding. But if things ever got to that point, we can be pretty sure that the new president would do what Gerald Ford did: issue a pardon to the former president, so that government could get on with the daily work of governing.

Nations with new constitutions have started to write into them provisions dealing with what the South African constitution usefully calls “institutions supporting constitutional democracy” — bodies whose job is to investigate charges of official wrongdoing. They try to ensure public confidence by making the institutions both independent of direct political control and somewhat accountable to the public: independence to guard against cover-ups and accountability to prevent overzealousness.

The designs vary among nations, but the better ones give civil society organizations a role in choosing investigators (unsurprisingly, I’m partial to ones that give law professors and political scientists a seat at the nominating table). They also give the investigators fixed and fairly long terms. The constitution Tunisia adopted after the Arab Spring provides for a “Good Governance and Anti-Corruption Commission,” with several members serving nonrenewable six-year terms.

The institutions supporting democracy are chosen in ways that closely resemble the ways in which members of the nations’ constitutional courts are chosen. Many nations have gravitated toward judicial nominating commissions with some politicians, some judges, and a majority of members from civil society — bar associations, universities, and public interest NGOs.

It’s no surprise that similar nominating commissions are used for both anti-corruption institutions and courts: Courts are the paradigmatic institutions supporting democracy. (There may be a lesson here for how the US selects judges too.)

In contrast to the situation in other countries, the US has a jerry-rigged anti-corruption system

Perhaps because we wrote our Constitution more than two centuries ago, it doesn’t set out principles for investigating high-level corruption. It’s worth quoting Thomas Jefferson: Referring to the Constitution’s drafters, Jefferson wrote in 1816, the drafters’ generation “deserved well of the country. It was very like the present but without the experience of the present; and 40 years of experience in government is worth a century of book reading.” Modern constitutions have drawn upon two centuries of experience. We should do what we can to draw on that experience too.

Because the US Constitution fails to outline agencies that might combat official corruption, we’ve jerry-rigged things instead. In the 1920s, we established a system in which the attorney general named an independent prosecutor. That’s how, a half-century later, Archibald Cox came to head the Watergate investigation.

The Saturday Night Massacre, when President Richard Nixon removed Cox, Attorney General Elliot Richardson, and Deputy Attorney General William Ruckelshaus, led Congress to adopt a statute requiring the attorney general to ask the courts to name a special counsel when credible allegations of high-level wrongdoing surfaced. The Supreme Court upheld that system, but it didn’t work well. The trigger for naming special counsels went off way too often, and special counsels proved to be ungovernable. Kenneth Starr’s investigation of Bill Clinton was a political disaster.

Now we’re back to the older system: When political pressure gets strong enough, the attorney general can name an independent prosecutor outside the ordinary lines of authority for federal prosecutors. That’s not a great design, but at the moment it’s what we’ve got.

The decision to name an independent prosecutor is inevitably made in a highly charged partisan setting, when everybody knows pretty much what the stakes are. We can’t guarantee independence and accountability except by making sure that the person chosen has the personal integrity to generate public confidence in whatever she or he does.

And, again, in a highly partisan world, such a person might be hard to come by: Republicans will be suspicious of an investigator with Democratic ties, and vice versa. Even worse, the only protection an independent prosecutor has is the honesty of the attorney general and president, who — as a matter of law — can fire such a prosecutor whenever they want. An independent prosecutor might have to keep looking over his shoulder to see how the president is reacting to the investigation.

The best options, given the current situation

Given the system we’re stuck with, how can we deal with this partisan dynamic in the case of Trump? First, find someone who really is a prosecutor. That’s one reason Kenneth Starr went off the rails. A former appeals court judge and solicitor general, he didn’t have the experience a prosecutor has in figuring when it’s worth looking into something and when it’s not. Second, find someone with a reasonably strong reputation among lawyers — not an obscure career assistant US attorney.

Finally, and I write this with some reluctance, find a prosecutor who has some mild Republican affiliations. That might give the investigation some credibility with Republicans inclined to support “their” team by making it less clear who they should root for. But only “mild” Republican affiliations, because the investigation has to be credible to Democrats as well.

One possibility is to look at the list of US attorneys initially appointed by George W. Bush who were held over into the Obama administration for a substantial time. At the moment, the example of Rod Rosenstein — the deputy attorney general who wrote a memo criticizing Comey’s decisions in the Clinton investigation — isn’t encouraging here. But I think there’s some chance that we’ll find out he was a straight shooter in the Comey episode who was manipulated by the president and the attorney general.

My proposal — to look for a Bush appointee — makes me slightly uncomfortable, because it fits a historical pattern that has deep partisan and ideological roots. That pattern emerges from the background when you look at the list of independent counsels and special prosecutors since Watergate. Investigators with Republican affiliations get to look into charges of wrongdoing by Democrats and Republicans, but Democrats don’t get to look into charges about Democrats. The pattern is modest, the numbers aren’t large, the partisan affiliations of several investigators are difficult to determine, and there might be reasons other than partisanship to explain it. But with all that, the pattern is there.

I think the reason is this: For several generations Republicans have said that Democrats don’t really respect the rule of law. Democrats, Republicans say, think that the law is whatever they happen to want — or that the law is “just politics” in another guise. Those arguments have penetrated deeply into public understanding. So, of course, the public impression comes to be that a Democrat can’t be fair when investigating anyone, Democrat or Republican. Republicans, in contrast, are deeply devoted to a politically neutral rule of law, or so they say. So they can investigate anyone.

Democrats have responded, basically, “So’s your old man!” They say that Republicans don’t follow the rule of law either. That argument, though, hasn’t worked. Justice Elena Kagan made the argument that Democrats should make: It’s “law all the way down,” she said at her confirmation hearings. She meant that every good lawyer and judge uses a range of arguments to support her conclusions, and every one of those arguments counts as “law.”

My Democratic-leaning friends scoff at that, but they shouldn’t. It is law all the way down, but we have to come up with ways of showing to the public that our idea of the rule of law is better than the Republicans’.

Nearly every hour brings new developments in the Comey (or Trump) affair, and what I’ve written might be outdated by the time it’s published. But whatever turns the investigation takes, it is clear that political developments will matter more than legal arguments. That’s the way our system was designed.

The key, therefore, is whether the Republican political firewall now protecting Trump breaks. At present, it seems to be holding, with a large majority of Republican voters supporting the president’s decision to fire Comey, and little beyond mild expressions of regret from Congress. My guess is that we won’t know the long-term repercussions of the firing — and, now, the revealing of classified intelligence to Russia — until June 21, after the election for Tom Price’s congressional seat. (Price left the House to become secretary of health and human services.)

If Democrat Jon Ossoff wins, we might see Republicans heading for the hills, jumping ship, or whatever your preferred metaphor is. Instead of supporting the president, they’d start pressing hard to get a serious investigation off the ground. If Ossoff loses, Trump will probably stagger through this episode until the next outrage.

No one can know how the politics of the current affair will play out. It is unfortunate that the way we’ve designed our system for investigating official misconduct makes politics more dominant than it could be. But it’s doubtful we’ll be able to design a better system while we’re in the middle of a crisis.

Mark Tushnet, a leading scholar of constitutional law and legal history, is the William Nelson Cromwell professor of law at Harvard Law School. He contributes to the blog Balkinization.