Gastonia, NC
The easiest constitutional principle in America is the one that binds the other side. The useful test is whether the sentence survives a change in the proper noun.
Before running the test, I want to correct the tense in the question, because it is doing a great deal of work and it is wrong. This is not a hypothetical. It is a campaign promise, made repeatedly, by a candidate who may win in fifteen days.
At the conservative conference in March of last year he told the audience, in terms, that he was their retribution. In November he said he would appoint a real special prosecutor to go after the President and his family. He has said since that he would have his opponents investigated, and he has named some of them. Whatever one makes of how much of that is rally language, it is not a thought experiment supplied by this Forum. One of the two people who may take office in January has stated the intention publicly and more than once.
That changes what we are actually discussing. We are not asking whether defenders of the current docket would rediscover process in some imagined future. We are asking whether they will do it in about eleven weeks.
My prediction is that many of them would, and quickly. I do not say that as an accusation of personal hypocrisy, because political language is often sincere in both directions. People genuinely experience a procedure as routine when it reaches someone they distrust and as abusive when it reaches someone they trust. That is exactly why constitutional rules cannot depend on sincerity.
I read copy for a living, and the best test I know is whether a sentence can be reused without rewriting its verbs. “Venue is proper because the charged conduct occurred there.” Fine — would you say it in the mirror case? “A gag order is justified because trial integrity outweighs some public commentary.” Fine — would you apply it to your own candidate? “A special counsel is an ordinary mechanism for politically sensitive prosecutions.” Fine — would you still call it ordinary when the appointing administration belongs to the other party?
If the sentence changes with the name, it was never a rule. It was advocacy.
That is what has troubled me about the phrase “rule of law” over the past three years. Too often it functioned as a closer rather than an argument. Once somebody said no one is above the law, every further question about timing, venue, charging discretion, appointment structure, or cumulative effect was expected to sound like excuse-making.
A transferable rule of law would work differently. It would invite scrutiny precisely because the defendant is important. It would ask whether prosecutors are using familiar statutes in familiar ways, whether timing is necessary or advantageous, whether the venue is fair as well as proper, whether the prosecutor’s authority is clear, and whether speech restrictions are tied to threats rather than criticism.
Those are not Trump rules. They are anti-retaliation rules, and the people who need them next may not be the people who need them now.
Here is why I think the coming test is harder than the one just past. The phrase “they started it” will be available to everybody, and it will be available with specifics. A Republican Justice Department will be able to point to a county prosecutor who elevated a stale misdemeanor, a state attorney general who campaigned on a defendant, and a special counsel appointed under a structure a federal judge found unlawful. Those are not fabricated grievances. Some of them this Forum has spent eighteen questions documenting.
That is precisely what makes restraint difficult and necessary at the same time. A grievance with evidence behind it is the most dangerous kind, because it makes retaliation feel like correction.
So the work has to be done now, in the two weeks that remain, and it has to be done by people whose side may be about to hold the pen. If you believe the prosecutions of the last three years were ordinary, write the governing principles prospectively and accept that they will constrain the next Republican Attorney General. If you believe they were lawfare, say now — before your side has the offices — which of those tools you will decline to use.
Most people become less enthusiastic when the rule is stated that way. That hesitation is useful information. It does not prove any prior prosecution was unlawful. It shows that “rule of law” can become a partisan permission structure unless the underlying rules survive reciprocity.
A principle that disarms only the other side is not constitutional restraint. It is campaign copy with a law-book cover.
Owatonna, MN
A fence is easy to admire when it runs along the other man’s pasture. The constitutional test is whether you still defend it when it reaches your own land.
I have a correction to offer before I answer, and it is the sort that changes the shape of the question rather than a detail inside it. The experiment this question describes as hypothetical has already been run, and it ran for four years.
In May of 2019 the Attorney General assigned John Durham to investigate the origins of the Russia investigation — that is, to examine the conduct of the previous Democratic administration’s law enforcement and intelligence agencies. In October of 2020, with an election three weeks away, he elevated Durham to special counsel under the same departmental regulations that later produced Jack Smith, and did it in a way designed to let the appointment survive a change of administration. Durham operated until May of 2023.
That is the mirror. A Republican Justice Department, using the special counsel mechanism, investigating a departing Democratic administration, with an appointment timed just before an election.
The results are the most instructive part and almost nobody remembers them. Durham brought three prosecutions. An FBI lawyer pleaded guilty to altering an email and received probation. Michael Sussmann was tried on a false-statement charge and acquitted in under six hours in May of 2022. Igor Danchenko was tried and acquitted in October of that year. The final report was heavily critical of the FBI and produced no further charges.
Four years, three defendants, two acquittals, one probation sentence.
I spent a career in financial planning, and the discipline that stays with me is stress testing. A portfolio that works only in one market regime is not robust; a constitutional rule that feels acceptable only when your coalition holds the offices is the same fragile structure. Well — here is a completed stress test that nobody is citing, and it tells us three useful things.
First, the tool crosses party lines already. This is not something a future Republican administration would have to invent. It was used in the other direction within the last five years, under the same regulations, and it was defended by the same kinds of arguments now being made about the current docket.
Second, Democrats at the time made exactly the objections this question predicts they would rediscover. They said the appointment was politically timed, that the theories were thin, that the prosecutor was working backward from a conclusion, and that the process was being used to legitimize a political narrative. Some of those complaints look better after two acquittals than they did when they were made.
Third, and this is the part that should sober everybody, the tool produced almost nothing. Four years of investigation with subpoena power and grand juries yielded two losses at trial. That is not an argument that the tool is harmless — the process itself is punishment and the reputational effect is real. It is an argument that reciprocal prosecution is a poor investment even for the side that undertakes it.
So my answer is that I expect rediscovery, and that I have already watched it happen once in the other direction, which makes me less impressed with everyone’s principles and more interested in written rules.
Those rules have to be decided before the emotional facts arrive, and there are fifteen days left in which nobody knows who will hold the offices. That is the last honest moment. Should former presidents face criminal prosecution for official acts, or should immunity draw a strong boundary? If special counsels are going to prosecute politically sensitive defendants, should Congress establish the office rather than leaving it to regulations that have now been challenged successfully once? Should a local prosecutor be able to construct a felony whose political meaning comes from federal election law? How narrow should gag orders be when the defendant’s criticism is also campaign speech?
Those questions are easier to answer honestly while the next defendant’s name is unknown, and they will be impossible to answer honestly in January.
A principle becomes credible when it imposes costs on your own coalition. That is the test most political movements avoid, and the Durham file suggests both movements have avoided it at least once already.
Prescott Valley, AZ
A civic rule that changes when the name changes is not much of a rule. That is the lesson I would want students to take from the last three years, and I would teach it with a case study rather than a hypothetical, because we have one.
Put the current docket aside and put this on the board instead.
An Attorney General appointed by one party assigns a prosecutor to investigate how the previous administration, of the other party, conducted a major investigation. He runs for four years. Three weeks before a presidential election the Attorney General converts him into a special counsel under departmental regulations, specifically so the appointment will survive if the administration changes. The prosecutor brings three cases. One defendant pleads to altering a document and gets probation. Two go to trial and are acquitted, one of them in under six hours.
Now ask the class the questions.
Was the appointment timing appropriate? Was the theory of the cases sound? Was the prosecutor working from evidence or from a conclusion? Were the objections raised at the time — and they were raised, loudly, by the party whose officials were being investigated — legitimate process concerns or partisan noise?
Only after the class has answered do you tell them the names: Barr, Durham, Sussmann, Danchenko. And then you ask whether any answer changes.
That is the exercise, and it is more useful than the version with Trump in it, because the students will already have opinions about Trump and almost none about this.
I taught English and journalism for a long time and the habit that survives is refusing to teach a rule I cannot state without a proper noun. So here are the five habits I would want students to carry out of this period.
Ask whether the rule is general enough to apply to both parties. Distinguish the defendant’s conduct from the government’s procedure. Ask whether timing creates avoidable electoral effects. Identify the legal authority of every institution exercising power. And test the rule against the next defendant rather than only the current one.
Those habits matter more than agreement about any particular case, and the Durham record shows why. When that investigation was running, Republicans described it as accountability and Democrats described it as a politicized fishing expedition. Today the positions on the current docket are precisely reversed, and both sets of arguments are the same arguments. That is not a coincidence and it is not evidence that everyone is lying. It is evidence that people evaluate procedure through the identity of the target, which is exactly the failure a constitution is supposed to prevent.
There is a second lesson in the Durham outcome that I would not skip. Two acquittals. A jury took under six hours in one of them. Whatever one believes about the origins of the underlying investigation, the prosecutions failed, and they failed in front of ordinary citizens who heard the evidence.
That is worth teaching because it complicates the pessimism. The mechanism was available, it was used across party lines, and juries declined to ratify it. A system in which the tool exists but rarely works is not the same as a system in which the tool exists and always works, and students should learn the difference rather than assume the worst version.
But the tool still cost four years, several careers, and an enormous amount of public confidence, and it did that without a single trial conviction. That is the real argument against reciprocity: not that it always succeeds, but that it burns institutional capital whether or not it succeeds.
If commentators who defended the last three years want to prove they believed in rule of law rather than one-way accountability, the moment to say so is now, before the offices change hands. Write clear appointment rules. Write narrow gag standards. Write a uniform records process. Demand neutral venue principles.
Do it before the mirror. Otherwise the rule may be legally sound and will still look like something people found in a drawer when they needed it.
Dayton, OH
A specification is only a specification if the plant uses it on its own parts. That is the process-control version of this question, and I want to answer it with a data point rather than a prediction, because there is one available and it is being ignored.
In 2019 a Republican Attorney General assigned a prosecutor to examine how the prior Democratic administration had conducted a major investigation. In October of 2020 he made that prosecutor a special counsel under the same regulations that produced the current one, three weeks before an election, structured so the appointment would survive the transition. That investigation ran until 2023 and produced two acquittals and one probation sentence.
So the mirror exists and the plant has already run parts on the other line.
What interests me as a process person is not the outcome but a second decision the same Attorney General made in the same period, which almost nobody discusses. Under sustained public pressure from the President who had appointed him, he declined to open a prosecution of the Clinton Foundation. He declined to appoint a special counsel on the President’s election-fraud claims. He said publicly, in December of 2020, that he saw no basis for it. He was fired in effect for those refusals, and he resigned that month.
That is a control functioning. An officer with the authority to act, under enormous political demand to act, declined because the evidence did not support it. Whatever else one thinks of that Attorney General, that is the behavior a specification is supposed to produce, and it happened.
I raise it because the pessimistic version of this question assumes that every officeholder will use every available tool once permission exists. That assumption is not supported by the record. Permission existed. The tool was partially used and partially declined. The parts of it that were used mostly failed.
So my answer is a process answer rather than a moral one. The question is not whether the tool will be used — it has been, by both sides. The question is whether the specification governing its use can be written so it does not depend on the character of whoever holds the office.
That is the whole point of a specification. A plant cannot maintain confidence if inspectors apply one tolerance to Supplier A and another to Supplier B, even when each inspector sincerely believes his judgment is correct. Repeatability requires a rule that different operators can apply to different parts. Right now the country is relying on the personal restraint of individual Attorneys General, and it got lucky once in 2020. Luck is not process control.
Take venue. Whatever standard governs transfer in a nationally consequential case should be available to both parties, and it should be written before anyone knows which district is hostile to whom.
Take timing. Prosecutors cannot suspend every case because a defendant is running. They also should not pretend timing has no political effect. The standard for explaining election-year timing should be identical regardless of target — and note that an appointment three weeks before the 2020 election would fail a strict version of that standard just as clearly as anything in the current docket.
Take special counsels. The office has now been found unlawful once, this July, by a federal judge, and that ruling is on appeal. Whatever the Eleventh Circuit says, Congress could remove the ambiguity, and the reason to do it in the next few weeks rather than the next few years is that nobody yet knows which party will need the answer.
Take gag orders. True threats and interference can be restrained; political criticism of the prosecution should be governed by a rule that survives a party swap. The D.C. Circuit already wrote most of that rule in December. Codify it.
People become uncomfortable when the specification is stated that explicitly, and the discomfort is diagnostic. It tells you where politics had been hiding inside supposedly neutral process. That discomfort is available to inspect for exactly two more weeks.
Wheeling, WV
I spent enough years around grievances to know how this goes. Management loves a broad rule until another manager uses it on management’s own people. Then everybody discovers due process.
That is what I expect if the parties reverse, and I want to make one observation about it that nobody in this argument seems willing to make, from either direction.
We have seen this exact promise before and it was not kept.
In 2016 the chant at every rally was lock her up. It was on signs. It was in speeches. It was the closest thing that campaign had to a policy commitment on the subject. And then he won, and within about two weeks he told an interviewer that he did not want to hurt the Clintons, that they had been through enough. No prosecution followed. His own Attorney General later declined to open a Clinton Foundation case and declined to appoint a special counsel on the election claims, under heavy public pressure from the man who had appointed him.
Now the same candidate has spent two years saying he is our retribution and that he will appoint a special prosecutor to go after the President and his family. Fifteen days from an election.
I do not know which of those two patterns is the better predictor. I do know that anybody who tells you with confidence that the mirror prosecutions are coming is ignoring 2017, and anybody who tells you they are only rally talk is ignoring how much angrier the last three years have made everybody. Both readings have evidence.
What I am sure of is the labor lesson, and it applies whichever way this goes. Process is not something you invent after discipline lands on your friend. Either the steward gets notice or he does not. Either the contract permits the penalty or it does not. Either the supervisor had authority or he did not. You settle those questions before the grievance, because after the grievance nobody believes you.
Politics does the opposite. It starts with the person and works backward to the principle. That is why I expect the constitutional vocabulary to come back fast if the caption changes. People who called venue complaints whining will discover hostile jury pools. People who dismissed timing as coincidence will discover election interference. People who called appointment arguments technicalities will become very interested in who authorized the prosecutor.
Some of those arguments will be right. That is what bothers me. They were worth hearing before the name changed too.
Here is the part I would say to my own side, because a one-way sermon is worthless. If Republicans take the last three years as permission to build a mirror stack, the country will be in worse shape than it was. Retaliation does not repair unequal process. It makes unequal process reciprocal, and reciprocal is not the same as fair.
And the Durham record suggests it would not even work. Four years, subpoena power, grand juries, and two acquittals. The prosecutions that get built out of political grievance tend to be the ones juries throw out, because grievance is not evidence and twelve people can usually tell the difference.
So my preference is plain. I would rather see Republicans decline a mirror campaign than celebrate one. I would rather see the people who defended the current docket support limits now that might protect a future Republican president than discover those limits in January.
Do I expect either? Not much. I expect process to come back when the caption changes, and I expect people to speak as though those values had never been dismissed.
Once workers see management discover procedure only when a favored supervisor is in trouble, the next fair decision gets distrusted too. Selective process poisons future process. That is where this is headed, and the only cure is to write the shop rule before the next grievance rather than during it.
Tyler, TX
Insurance teaches you to read the exclusion when you are selling the policy, not after your own claim gets denied. That is the symmetry test, and I want to apply it to a specific power rather than to the general mood, because general moods do not price.
Take the special counsel. Under the current regulations, an Attorney General may appoint a prosecutor from outside government, give him a budget and a mandate, and structure the appointment so it survives a change of administration. Jack Smith holds that office, or held it — a federal judge in July held that the appointment was unlawful and the Eleventh Circuit has it now.
Would I sell that power to the other party?
I do not have to guess, because it was sold to them in October of 2020. The Attorney General converted John Durham into a special counsel three weeks before an election precisely so a Democratic administration could not simply dismiss him. Republicans called that institutional protection. Democrats called it a partisan booby trap. Four years and two acquittals later, both descriptions look partly right.
That is a completed transaction and it should be priced into everyone’s view of the current one.
Here is the underwriting question stated properly. Would I accept a Republican Attorney General appointing a special prosecutor to pursue a departing Democratic president under a structure a federal court has already found defective, on a timetable set by an election calendar, with no statutory office behind it? No, I would not, and I would say so in print. And since I would not accept it in that direction, I do not get to accept it in this one.
That is not a statement about Trump. It is a statement about the instrument.
The reason I expect rediscovery is that people routinely price political risk at zero when their own side holds the institution. They assume good faith. They assume discretion will be exercised responsibly. They assume their prosecutor knows the difference between law enforcement and politics. Then control changes and suddenly discretion carries a premium. That is ordinary human behavior and institutions are supposed to manage around it rather than be surprised by it.
The temptation now is for Republicans to say fine, we will use the same tools. I understand the emotion and it is bad risk management for a reason that has nothing to do with fairness.
Once both parties normalize mirror prosecutions, every president leaves office carrying contingent criminal exposure tied to the next election. Every Attorney General inherits a demand to examine the last administration. Every state prosecutor with a creative reading of a state statute becomes a potential national actor. That changes the risk profile of the presidency itself, and it changes what an election is: winning begins to include control over the legal exposure of the people who just lost.
I would not write that policy for a client and I would not want to be the reinsurer on it.
My preferred answer is to write the exclusions now. Not candidate immunity — boundaries. Private crime remains prosecutable. Official conduct gets the protection the Supreme Court defined in July. Special prosecutors get clear statutory authority, which Congress could supply in a month if it wanted to. Gag orders stay close to threats and tampering, which the D.C. Circuit already wrote. Records handling gets one published standard. Venue doctrine takes saturation seriously in nationally consequential cases. Election-year timing gets a written explanation.
Those limits should apply in both directions, which is the only version worth insuring.
The test to run before defending any new power is simple. Would I sell this to the other party? Would I let their prosecutor use this statute, this venue, this timing theory, this gag authority, and this appointment structure against my candidate? If the answer is no, the power is underpriced. That does not prove it is unconstitutional. It proves the institutional downside deserves more attention than it is getting.
Fifteen days from now, somebody is going to hold the pen. This is the last week in which either side can write a rule without knowing whether it helps them.
Long Island, NY
Reciprocity clauses exist because parties are remarkably good at imagining fairness in one direction only. Constitutional procedure requires the same discipline, and the drafting problem here is more concrete than the question suggests.
Start with what has actually been said, because the hypothetical framing understates the situation. The candidate has stated publicly and repeatedly that he intends to appoint a special prosecutor to pursue the sitting President and his family. He said at a conference last year that he is his supporters’ retribution. Whatever weight one gives to campaign rhetoric, this is not a scenario a forum invented; it is a stated intention two weeks before an election.
And the structural version of it has already occurred once, in the other direction. In October of 2020 an Attorney General appointed a special counsel to investigate the prior Democratic administration’s conduct, deliberately structured to survive a change of administration. That investigation produced two acquittals and one probation sentence and ended in 2023.
So the drafting question is not academic. It is: what rule would the country want in place before January that it would accept regardless of who wins?
My background is in legal process, and the most important distinction I know is between a rule and a litigation position. A rule is stated at a level of generality that survives a change in parties. A litigation position is designed to win the present dispute. The last three years produced too many litigation positions spoken as though they were constitutional rules, and the coming three years will produce the same thing in reverse unless something is written down.
Let me be specific about what a two-way rule would have to say, because that is where enthusiasm usually dies.
On appointments: Congress vests appointment of a special prosecutor in the Attorney General by statute, defines the office, specifies supervision and removal, and applies it to any investigation of a current or former president or a major-party nominee, regardless of party. That would resolve the question the Eleventh Circuit is now considering, and it would bind the next Attorney General as firmly as the last.
On venue: in nationally consequential prosecutions where more than one district has a substantial connection, the defense raises transfer at the outset or waives it, the government states its operational basis, and the court decides on a written record before jury selection. That would have helped the current defendant in Washington. It would equally protect a Democratic defendant in a district that voted seventy percent the other way.
On speech: restrain true threats, tampering, targeted intimidation, and disclosure endangering protected participants, and nothing else. The D.C. Circuit wrote that rule in December; codify it so a trial judge applies it in the first instance rather than an appellate court supplying it two months into a campaign.
On official acts: the immunity decision in July did most of this work already, and it will protect a Democratic former president from a Republican Justice Department exactly as it protects the current defendant. That is the clearest example available of a rule that binds both directions, and the people who denounced it should notice which of their principals it will shield next.
On records: one published escalation ladder, applicable to every official covered by the Presidential Records Act.
Each of those constrains the next Republican administration. That is the point, and it is why I doubt any of them will be adopted before January. People like limits when the other party holds power and resist them when the limits might prevent their own future case. That resistance is the most revealing datum in this entire question.
I would use one drafting test. Can the rule be written without the name of any current political figure? If not, it is too case-specific to carry constitutional weight.
Constitutional credibility is built before the emergency, not during it. A rule adopted after the mirror prosecution begins will be attacked as special pleading and the attack will be fair. The same rule adopted in the next two weeks has a chance to become doctrine rather than defense strategy.
Cheyenne, WY
Swap the name and the rule should stay put. That is the whole test, and I want to add one entry to the ledger before I run it, because the argument gets made as though the machinery has only ever pointed one direction and that is not what the books show.
This Justice Department prosecuted a Democratic senator and convicted him in July. It appointed a special counsel who prosecuted the President’s son to a conviction in June and a guilty plea in September. And a Republican special counsel appointed by the same Attorney General investigated the President himself over classified documents and declined to charge in February.
Going the other way, a Republican Attorney General appointed a special counsel in October of 2020 to investigate the prior Democratic administration, and that investigation produced two acquittals. The same Attorney General declined to open a Clinton Foundation case and declined to appoint a special counsel on the election claims, under public pressure from the President who appointed him.
Those are the entries. They do not net to zero and they do not net to a clean story either way. What they show is a system that has run in both directions, produced very little in the way of convictions in the political cases, and left everybody convinced the other side is the aggressor.
I keep books. A rule that changes with the account holder is not a control; it is favoritism. So run the swap.
If Washington is a proper venue for one national political defendant, explain the rule that makes it proper for another. If political saturation matters, say so now. If a special counsel can be appointed this way, write the authority clearly enough for both parties — and note that a federal judge held in July that the current authority is not clear enough, which means somebody is going to have to answer this question regardless of who wins. If gag orders can limit a candidate, define the line before the next candidate. If records get one treatment, publish the ladder.
Same entries, same rules. Different facts can produce different outcomes; that is fine. Names should not.
The worst response to the last three years would be for Republicans to take power and run the same play in reverse. That would not fix the ledger. It would add another bad entry, and on the Durham evidence it would probably add an expensive one that produces nothing.
The better answer is to close the account now. Write limits that bind your own side. That is the part political people hate because it costs a future tool, and costs are how you know the rule is real. A rule that costs only the other party is cheap, and cheap rules do not last.
I expect the future rediscovery to be sincere, which is what makes it dangerous rather than merely irritating. People will genuinely believe the Republican prosecutor is different, the venue is unfair, the timing is political. Some facts may genuinely be different. None of that answers the symmetry question, because the process rule is supposed to be written before the facts arrive. That is what controls do.
Controls work because the same approval rule is applied before anyone knows whether the transaction helps or hurts the department. Politics keeps trying to reverse that sequence: first the desired result, then the rule. That is bad control in a bookkeeping office and it is worse control in a republic.
Write the venue standard before the next venue fight. Write the special-counsel authority before the next special counsel. Write the gag line before the next candidate speaks. Write the records ladder before the next box is found.
There are fifteen days. After that, whoever writes the rule will be writing it knowing exactly whom it protects, and nobody will believe a word of it.
