Myrtle Beach, SC
Four days out, I think the larger civic question is no longer whether one payment was ugly, one records practice was sloppy, one call was inappropriate, or one valuation was aggressive. Those are finite questions. The larger question is whether the country created a special legal climate for one former president and then called every objection to that climate an attempt to erase the underlying facts.
That is why I put equal application first. But I want to state the scoreboard before I do, because I have spent this Forum insisting that arguments must survive their own best counterexample, and here the counterexample is the entire record.
Four criminal prosecutions. One produced a verdict. The classified-documents case was dismissed in July on the ground that the prosecutor was unlawfully appointed, and it sits at the Eleventh Circuit. The January 6 case was gutted by the Supreme Court’s immunity decision in July, has no trial date, and is back before a district judge working out what survives. The Georgia racketeering case has been stalled since spring while the district attorney’s own disqualification goes to the Georgia Court of Appeals, with argument set for December. The Manhattan conviction has not been sentenced and cannot be appealed until it is; sentencing was moved to the twenty-sixth of this month, after the voting.
And the one genuine attempt to remove him from a ballot — Colorado, under the disqualification clause — was rejected by the Supreme Court in March, unanimously.
So the honest description of the past three years is not that the legal system successfully disqualified a candidate. It is that a great deal was attempted and most of it did not work, largely because courts stopped it.
I work in marketing, and I know what happens when separate messages reinforce one another. One lawsuit remains a lawsuit. One indictment remains an indictment. But when the same name accumulates labels across civil, criminal, and regulatory forums, the public stops experiencing them as separate legal objects and starts experiencing a campaign of meaning. Fraudster. Felon. Racketeer. Document offender. Election conspirator.
Those nouns moved faster than the legal distinctions beneath them, and three of the five have no operative judgment behind them at all as of this morning.
That is where the equal-application argument actually lands, and it is a stronger place than where it started. The danger was never that the method worked. The danger is that it was attempted — that a county prosecutor, a state attorney general, two federal cases, and a state disqualification suit all converged on one candidate inside eighteen months, and that the answer to every process objection was a slogan.
Nullification would mean pretending a jury never voted or a judge never ruled. That is not the argument. I can acknowledge the Manhattan verdict and still ask whether the theory that produced it was ordinary enough to deserve confidence. I can acknowledge that classified records were found and still ask why the enforcement temperature differed across three men. Those are comparisons, and equal application depends on comparison.
What I would not now claim is that the courts were the instrument. On this record they were mostly the brake. The Supreme Court kept him on the ballot and gave him substantial immunity. A district judge threw out an entire prosecution. An appeals court cut a gag order to two categories and a bond requirement by more than half. A state appellate court is examining whether a district attorney should have been disqualified. A trial judge moved a sentencing off the election calendar.
That record is inconvenient for the strongest version of the lawfare thesis, and it is the reason the thesis should be stated as a warning about method rather than as a description of a completed operation.
The exhibit is finite. The method reproduces. A payment theory cannot prosecute the next president by itself; a norm of politically selective charging can. That is the risk I would rank first, and it survives the fact that this particular attempt mostly failed.
Novi, MI
A plant can have a defective part and a defective inspection system at the same time. Fixing the part does not excuse the inspection system, and finding bias in the inspection system does not make the part good.
That is the frame, and before I use it I want the yield numbers on the table, because a process argument without output data is an opinion.
Four criminal prosecutions initiated. One reached a jury. One was dismissed outright in July on the constitutionality of the prosecutor’s appointment and is on appeal. One was substantially dismantled by the Supreme Court’s immunity ruling and has no trial date. One has been frozen since spring pending an appellate ruling on whether the district attorney should be removed. The single conviction is unsentenced and unappealable until later this month.
Twenty-five percent of the prosecutions produced a verdict. Zero have produced a final judgment. That is the actual yield.
I raise it because the version of the lawfare argument I hear most often describes a machine that ran efficiently, and the data describe a machine that jammed at nearly every station. Any quality engineer who ignores that has stopped measuring and started advocating.
So what does the data support?
It supports a process-drift finding rather than an outcome finding. The question was never whether every underlying allegation was fabricated. Records handling matters. Business records matter. Calls to public officials matter. The question is whether one neutral system was inspecting a class of conduct or whether the tolerance changed with the caption.
On that, the record is genuinely mixed and I would rather say so than pretend.
The documents comparison remains the strongest exhibit for drift: three men, three temperatures, and the escalation ladder published only in February, after the outcomes. The Manhattan elevation remains the strongest exhibit for a theory built around a defendant, because the office never produced a comparable. Those are real gauge problems.
Against them: the courts corrected repeatedly and in most instances against the government. That is the inspection system catching its own errors, which is what an inspection system is for. A plant whose downstream checks reject three of four bad lots is not a plant without controls. It is a plant with expensive controls that worked late.
Late is the operative word, and it is where the cost sits. Each of those corrections arrived after the political effect had been spent. The immunity ruling came thirteen months after the indictment. The appointment ruling came thirteen months after that indictment. The gag narrowing came seven weeks after the order. The disqualification appeal is still pending fourteen months in. In manufacturing we call that inspection at final assembly rather than at the station, and it is the most expensive place to catch anything.
That is the finding I would file. Not that the system produced unjust convictions — it produced almost no convictions. That the system permitted extraordinary charging decisions to run for a year or more before any control engaged, in a period when the political value of the charging decision was highest.
The exceptions are the danger, not the outcomes. Too many parts of this required “this case is different.” Exceptions accumulate until the exception becomes the specification, and the next operator inherits the tool whether or not it worked the first time. A failed prosecution still teaches the next prosecutor that the theory can be filed, survive a motion to dismiss, and dominate a news cycle for eighteen months.
That is why I keep two ledgers. One records the underlying case: what happened, what the statute says, what a jury found, what remains open. The other records equal application: comparables, temperature, timing, venue, charging history, and whether the rule survives a change of names. The first can contain adverse facts without settling the second. The second can reveal a dangerous pattern without erasing the first.
The second ledger carries the larger long-run risk, and it does so even though — maybe especially because — the first one shows so little.
Orange County, CA
Orange County finance people are used to ranking risks that coexist. A borrower can have a genuine covenant breach and the lender can still be administering the credit unfairly. One does not cancel the other.
That is the ranking exercise, and it requires an accurate mark on the position first.
Here is the mark as of this morning. Of four criminal prosecutions, one produced a verdict that has not been reduced to judgment. One was dismissed in July and is on appeal. One lost most of its substance to the immunity ruling and has no trial date. One is frozen pending an appellate decision on the prosecutor’s disqualification. On the civil side, both Carroll judgments and the state fraud judgment are on appeal, with bonds posted; the Appellate Division heard argument on the fraud judgment in September.
And the single most consequential legal question of the cycle — whether he could be kept off ballots under the disqualification clause — was resolved in March by a unanimous Supreme Court in his favor.
That last one matters more than anything else in this file and it gets almost no attention from people making my argument. Colorado’s supreme court held he was disqualified. The Supreme Court of the United States reversed, nine to nothing, and held that states cannot enforce that clause against federal candidates. If the thesis is that the legal system was used to disqualify a candidate, the actual disqualification attempt was rejected without a single dissenting vote, eight months before the election.
An honest risk assessment has to price that in.
So my ranking is not that the method succeeded. It is that the method was attempted at scale and that the country’s protection against it turned out to be a handful of appellate rulings arriving late. That is a thin margin, and thin margins are what risk officers worry about even when the trade came out fine.
Think about how much of this depended on things that could have gone the other way. Trump v. Anderson could have been five to four. The immunity decision could have drawn the official-unofficial line differently. Judge Cannon could have ruled the other way on the appointment, as every prior court had. The Georgia Court of Appeals could have declined review. Any of those flips and the picture in this cycle looks entirely different, with the same underlying conduct and the same statutes.
That is what an unpriced systemic risk looks like: an outcome that depended on a sequence of discretionary judicial calls rather than on a rule anyone could have stated in advance.
The nullification objection is too binary. It assumes that if a jury returned a verdict, institutional criticism must be an attempt to erase it. That is not how risk review works. A judgment can be legally operative and part of a troubling pattern. An indictment can contain real evidence and reflect selective appetite. A civil remedy can rest on law and still raise proportionality questions.
The systemic exposure is what outlasts the cycle. A hush-money theory ends with this defendant. A precedent about how local prosecutors can elevate political cases remains available. A classified-documents dispute ends. A discretionary two-temperature records culture remains. A particular racketeering indictment ends, and it ended badly for the office that brought it. The template for treating a national political coalition as an enterprise remains on the shelf.
Those are future liabilities and they do not depend on the current cases succeeding. In fact a failed attempt may be more dangerous as precedent than a successful one, because the failure gets attributed to the particular facts rather than to the method.
So I would rank the method first, and I would state the reason precisely: not because it worked, but because it nearly did, and because nothing in the current structure would stop the next attempt from being better executed.
Bismarck, ND
A real bad fact can sit inside a bad process. Those are two different files, and I have said all year that the process file is the more dangerous one. Four days out, I still think so, and I want to say what the process file actually shows rather than what I assumed it would show.
It shows a lot of attempts and very few completions.
One conviction, unsentenced. One case dismissed on the prosecutor’s appointment. One case gutted by the immunity ruling with no trial date. One case frozen while an appellate court decides whether the district attorney can stay on it. A disqualification effort rejected nine to nothing in March. Two civil judgments under appeal with bonds posted, one of them cut by more than half by an appellate court that acted in ten days.
Compliance work does not let me file a finding of systematic capture when the record shows the system correcting itself four or five times, mostly against the government.
So here is my test, and I will hold myself to the answer.
Republican administration. Republican state attorney general. Republican county prosecutors. Departing Democratic president, running again. Four criminal cases, a half-billion-dollar civil judgment, a disqualification suit, gag orders in two jurisdictions. And then: the Supreme Court keeps her on the ballot unanimously, gives her substantial immunity from prosecution for official acts, a district judge throws out one of the federal cases entirely, an appeals court cuts a gag order to two categories, and a state appellate court moves to disqualify one of the prosecutors.
Would I say the rule of law had collapsed? I would say it had been badly stress-tested and had mostly held, and that the near misses should terrify everyone. That is my answer in this direction too.
What remains in the process file is the thing courts cannot fix on appeal, which is timing. Every correction came late. Thirteen months from indictment to the immunity ruling. Thirteen months from indictment to the appointment ruling. Fourteen months and counting on the Georgia disqualification. Seven weeks on the gag order. The corrections were right and they arrived after the political value of the original decision had been spent.
That is the compliance failure, and it is not a failure of outcomes. It is a failure of sequencing. In my work, a control that engages after the transaction has cleared is documentation, not control.
I would keep two columns and I would not let either erase the other. Column one: what did the defendant do and what did the court decide? Column two: was the government applying one rule across comparable people, and did the checks engage in time to matter?
Column two is where I still find trouble. Too many special explanations attached to one name — this case is different, that venue is necessary, this timing is normal, this appointment is valid, these records facts are distinguishable. Some of those explanations were right. Several were rejected by courts. A compliance system should not need a new explanation every time the same person enters the room.
And I want to say the thing that follows from all of it, because it is the point of having run the test. If the process was unequal, the remedy is not to give the next Democratic former president the same treatment. That would prove the method spread rather than that it was corrected. Equal application does not mean everyone eventually gets the same political punishment. It means everyone gets the same procedural standard, written before the case rather than explained after it.
Write the standards now, while nobody knows who holds the offices in January. That is the only moment when the writing is honest.
Clermont, FL
A theme park can have a real broken rail and a bad inspection culture at the same time. Fixing the rail does not excuse a system that inspects only one ride, and exposing selective inspections does not make the rail sound.
That is the cleanest description I have of where the country stands four days out, and I want to add the operating numbers, because the version of this argument I have been making needs them.
The rides mostly did not run. Four criminal cases: one verdict, one dismissed in July, one dismantled by the immunity ruling, one frozen since spring. The disqualification effort failed unanimously at the Supreme Court in March. Two civil judgments on appeal. The one conviction has not been sentenced and cannot be appealed until the twenty-sixth.
Small-business owners recognize this shape. It is what happens when a manager brings five disciplinary actions against the same employee and four of them get thrown out on review. The employee may still have done something wrong in one of them. What everybody now watches is the manager.
That is where I think the country is. The underlying allegations were finite. The process became a culture, and the culture is what reproduces.
Consider what a future prosecutor learns from this record. He does not learn that the theories were sound — most of them did not survive. He learns that a novel theory can be filed, survive a motion to dismiss, generate eighteen months of national coverage, produce a booking photograph, and shape a nomination contest, and that the office that filed it pays almost no institutional price when a court eventually rejects it. The Fulton County district attorney is fighting her own disqualification. The special counsel had a case dismissed on his authority to bring it. Neither outcome imposed any cost on the decision to file.
That is the culture problem, and it is worse when the attempts fail than when they succeed, because failure teaches that the attempt is cheap.
The documents comparison remains my clearest illustration of unequal method. Different facts may justify different outcomes, and the report in February named the distinguishing factors honestly. A neutral system should still have had one published decision tree before anyone’s name was attached, and it did not, and the Presidential Records Act still gives the Archives nothing to enforce with.
The Manhattan case raised a different question: whether the same records theory would have been built, elevated, and pursued the same way against an ordinary defendant. The office never produced a comparable and still has not.
The Georgia case raised another: whether a political operation becomes a racketeering enterprise because some participants allegedly crossed lines. That question is now on hold while an appellate court considers whether the prosecutor should be on the case at all.
Together those make a management culture. Individually they make three dockets.
The nullification charge is too easy. It turns every process argument into “he lost and now he is complaining.” He surely disliked losing. That tells us nothing about whether the process was equal. A difficult employee can complain selfishly and still identify a real policy problem, and a responsible owner investigates the policy anyway.
But I will give the other side its due, because a one-sided finding is worthless. The courts did their job repeatedly here. The Supreme Court kept him on the ballot when a state tried to remove him. Appellate courts corrected a bond, a gag order, and an immunity question. A district judge threw out a prosecution. If someone had told me in 2022 that this would be the record, I would have called it a system working under strain rather than a system captured.
What I would still fix is the culture that made the attempts cheap and the corrections slow. One records standard. Clear appointment authority. Narrow gag rules. Real comparables required when an old or unusual statute is used against a political figure. Those reduce future abuse without erasing a single past judgment.
Once employees believe management has one policy for favored people and another for disfavored people, even correct future discipline becomes hard to defend. That is the cost that outlasts every one of these cases.
Jacksonville, FL
Ports can seize contraband from one container and still have a larger problem if the inspection tariff changes with the shipper’s name. The container can be real. The tariff can still be unfair.
That is the distinction, and after three years of arguing it I want to put the manifest on the table before I draw the conclusion.
Four criminal shipments. One cleared to a jury and produced a verdict that is not yet a judgment. One was rejected in July because the inspector turned out to lack authority, and that ruling is on appeal. One lost most of its cargo to the immunity decision and has no berth assignment. One has been tied up since spring while an appellate court decides whether the harbormaster should be replaced. The attempt to bar the ship from the harbor entirely was rejected unanimously in March.
That is not a record of a tariff being enforced ruthlessly. It is a record of a great many filings and very few completed transits.
I have been describing this as a harbor problem all year, and I still think that is right, but the shape of the problem is different from the one I was describing. The trouble is not that the system delivered unjust outcomes. It mostly did not deliver outcomes at all. The trouble is that the routing decisions were made first and audited last.
In claims work you learn to distinguish a wrong payment from a bad handling pattern. A file can be payable and the adjuster can still have handled it improperly, and auditing the handling pattern does not cancel the claim. That is the frame that survives everything here. The Manhattan jury’s verdict is a verdict. The classified records were real records. A call was made to Georgia officials about Georgia’s count. None of that is erased by asking whether the routing rule was neutral.
And the audit, when it finally ran, mostly found for the shipper. That is worth saying by somebody who spent this Forum arguing the other way.
What remains is the timing, and timing is the whole of my complaint now. The appellate corrections were right and they were late. Thirteen months to the immunity ruling. Thirteen months to the appointment ruling. Seven weeks to narrow a gag order. Fourteen months and counting on a disqualification. Every one of those decisions arrived after the political effect of the original had been fully spent — after the booking photograph, after the arraignments, after the nomination contest.
In port terms, that is inspecting the load after it has been delivered and the customer has already told everyone what was in it.
The replication risk is what I would rank first, and it does not depend on any of these cases succeeding. A hush-money count ends. A records case ends. A civil judgment ends or changes on appeal. A prosecution culture becomes a manual, and the manual now contains a fully documented demonstration that a county office can file a national political theory, hold it together for eighteen months, and shape a presidential nomination, with essentially no institutional consequence when a court later says the office should not have been on the case.
That is the harbor rule that changed, and it changed even though the cargo mostly never moved.
So the fix is not to nullify anything. It is to stop making the attempt cheap. One documents ladder. Clear appointment authority written by Congress rather than assembled from four general statutes. Defensible venue standards. Narrow speech restraints. Comparables required when an unusual theory is used against a political figure. Faster appellate review where the timing effect is irreversible, which is the only one of these that would have made a real difference in the last three years.
Post the rules before the next ship arrives. That is how equal application becomes a tariff instead of a slogan.
Las Vegas, NV
A bad hand and a bad table can exist at the same time. Saying the table is bad does not erase the cards.
Four days out, I still think the table is the bigger institutional problem. I also think anybody making that argument has to read the results before he makes it, and the results are not what our side has been claiming.
The house lost most of the hands.
One criminal conviction, unsentenced. One case thrown out in July because the dealer was not properly licensed, now on appeal. One case stripped down by the Supreme Court’s immunity ruling with no date on the board. One case frozen since spring while an appellate court decides whether the pit boss stays. The effort to bar him from the table entirely — Colorado — was reversed nine to nothing in March. Two civil judgments on appeal, one with the bond cut by more than half in ten days.
If you had told me two years ago that this would be the tally, I would not have believed it, and I have spent this Forum predicting worse.
So let me be exact about what I still think is wrong, because “the house tried and failed” is not the same as “the house is clean.”
The table rules were made up as the hands were dealt. Documents got one temperature here and another there, and the ladder was published in February, after the outcomes. A local records case became a felony through a theory the office never showed it had used on anybody else. A political operation became a racketeering enterprise. Gag fights became campaign management. Each hand had its own explanation, and the same name had every explanation.
And here is the part that matters most from where I sit, which is operations. The corrections all came from upstairs, and they came slowly. The floor kept dealing for a year or more before anybody from the office walked down and stopped it. In my business that is the definition of weak supervision — not that the bad hand went unreversed, but that it ran long enough to change the night.
The nullification argument tries to make table review sound like sore losing. It is not. A jury can convict, a judge can enter a judgment, and the house can still be running one player under a different rule. Both are true, and the appellate record here shows the second one more clearly than the first.
What worries me for the next player is what everybody just learned. Not that the theories worked — most did not. That the attempt is cheap. A prosecutor who files a novel political theory gets eighteen months of national attention, a booking photograph, a nomination contest reshaped, and no institutional consequence when a court later says she should not have been on the case at all. That is the lesson sitting on the felt, and the next person to pick it up will run it better.
Players learn the real rules by watching what happens to other players, not by reading the plaque on the wall. Citizens do the same with law. They watch who gets searched, who gets charged, who gets a novel theory, who gets a quiet declination, who gets a gag, who gets a giant civil number. Those observations become the practical rule whether institutions like it or not.
So post better rules before the next player sits down. Comparables. Clear escalation standards. Clear appointment authority. Narrow speech restrictions. Neutral venue doctrine. Faster review where the clock is the whole game.
Boring is good, because a predictable table lets people argue about the hand. An unpredictable one makes every hand political, and that is where this ends up if nobody changes the rule while the memory is fresh.
Knoxville, TN
A bad component can be real and the design can still be the larger failure. That is the ranking, and I want to give it with the inspection results in hand rather than from the drawing.
Four criminal prosecutions. One verdict, unsentenced. One dismissed in July on the constitutionality of the prosecutor’s appointment, on appeal. One substantially dismantled by the immunity decision, no trial date. One frozen since spring pending an appellate ruling on the district attorney’s disqualification. The disqualification suit that would have removed him from ballots failed unanimously at the Supreme Court in March. Two civil judgments on appeal.
That is a structure that was loaded hard and mostly did not carry. Whatever else this record shows, it does not show a legal system that successfully removed a candidate, and anybody in this Forum who has implied otherwise — including me, in earlier sessions — should say so plainly.
So the design question changes. It is not whether the system produced unjust outcomes. It is whether the specification permitted extraordinary things to be attempted, and whether the checks engaged in time to matter.
On the first, yes. On the second, no.
Every correction in this record arrived after the load had already been carried by something else. The immunity ruling came thirteen months after the indictment. The appointment ruling came thirteen months after that one. The gag narrowing came seven weeks after the order, and the order had already produced ten contempt findings. The Georgia disqualification has been pending for fourteen months and the case has not moved in either direction. In public works we would call that inspecting after the concrete has cured. The finding is correct and the remedy is expensive.
That is my design finding, and it is more useful than the one I brought into this year. The failure mode is not capture. It is latency.
Latency matters because in an election cycle the political effect is spent long before the legal correction arrives, and the correction cannot be applied retroactively to the effect. A dismissed prosecution does not give back the year in which it dominated coverage. A narrowed gag order does not restore speech not made. A reversed disqualification ruling does not undo the months in which a candidate’s ballot access was genuinely uncertain.
The design correction follows directly. Where the timing effect is irreversible, review has to be faster. That is a solvable engineering problem — the Supreme Court took Trump v. Anderson on an expedited basis and decided it in about a month, which shows the system can move when it decides something is urgent. It took considerably longer on immunity, and the delay was itself a subject of argument.
The other design correction is upstream. Make the attempt more expensive. Require comparables when an unusual statute is used against a political figure. Publish the records ladder. Have Congress write the special-counsel office rather than leaving it assembled from four general statutes that a federal judge has now found insufficient. Codify the gag standard the D.C. Circuit wrote in December.
None of that erases a verdict. It leaves the individual cases where they are, for courts to finish, and changes what the next set of institutions may do.
That is why the nullification accusation fails. You can keep the component report and still redesign the bridge. The component report says a Manhattan jury convicted on thirty-four counts. The design report says a great deal was attempted against one candidate in eighteen months, most of it did not survive judicial review, and the reviews arrived a year late apiece.
Both of those are in the file. The second is the one that governs the next cycle, because specifications persist and individual defects do not. A failed attempt still becomes a citation. A dismissed theory still becomes a template. The next designer opens the manual and finds a route already drawn, and asks why he should exercise unilateral restraint.
That is how design drift becomes architecture, and it is why I would rank the method above any single allegation on the eve of an election that will not settle any of it.
Sheffield, Jamaica
From outside the United States, the striking thing about this record is not the prosecutions. It is the scoreboard, and the scoreboard says something the participants seem unable to say to one another.
Four criminal prosecutions of a former president. One produced a verdict, and that verdict has not been reduced to judgment. One was dismissed because a federal judge held the prosecutor had never been lawfully appointed. One was dismantled by the Supreme Court. One has been frozen for months while an appellate court considers whether the elected prosecutor who brought it should be removed from it. And the single attempt to keep the man off a ballot was rejected by the Supreme Court without a dissenting vote.
I have spent a career studying institutional capacity in small states, which means I have spent a great deal of time watching what happens when courts fail to check a prosecution. This is not that. Whatever else the last three years were, they were a period in which the judiciary repeatedly ruled against the government in cases where the government had enormous political incentive to win, and in which the highest court intervened twice on behalf of a defendant its members were widely assumed to dislike.
By the standard the United States has spent thirty years exporting to my region, that is the machinery working. Slowly, expensively, and late, but working.
I say this as someone who thinks the American correspondents in this Forum have the better of the underlying argument. The county records theory was constructed. The enforcement temperature on classified documents was not visibly governed by any rule written in advance. The racketeering indictment was a county office writing the national account of a presidential election. Those criticisms survive the scoreboard and in some respects the scoreboard confirms them, because a great deal of what was attempted did not withstand examination.
But the conclusion I would draw is different from the one my colleagues draw, and I think it is the more useful one.
The danger in this hemisphere has never been that a former president was convicted. It has been that the conviction arrived before the review and the election arrived before both. Brazil is the case: a leading candidate barred from a national election by a conviction later annulled for want of jurisdiction, the annulment arriving three years after the vote it decided. Peru has cycled through a generation of prosecuted presidents without anyone concluding that corruption declined. Guatemala’s anti-impunity commission produced genuine convictions until it reached a sitting president, who expelled it — which is the other half of the lesson. The method does not merely get copied by the next faction. It gets captured by whoever it threatens.
That is the risk I would rank first, and it is a timing risk rather than an outcome risk. Every American correction in this record was right and every one of them was late. Thirteen months from indictment to the immunity ruling. Thirteen months to the appointment ruling. Fourteen months and counting on the Georgia disqualification. In each instance the political value of the original decision had been fully collected before the correction arrived, and no appellate opinion can return it.
There is a cost here that Americans cannot see from inside and that I would ask them to take seriously as they vote.
For three decades, prosecution of a former head of government has been treated in international governance work as a signal — not proof of authoritarianism, but a marker requiring explanation. American institutions helped establish that convention and American diplomats have applied it to governments in my region within my own working lifetime. Having now generated four such prosecutions in eighteen months, two of them by elected local officials, the United States has spent that convention. It will be cited back, and it has already begun to be. A minister in Kingston or Georgetown facing questions about a predecessor’s prosecution now has an answer he did not have in 2020.
Whether that is a fair trade is for Americans to decide on Tuesday. It is a real cost and it will not appear on any docket.
Sydney, Australia
The question asks whether the accumulated method has become a larger threat than any allegation in the files, or whether that claim is an attempt to nullify verdicts a defendant disliked. I would decline the framing, because both halves assume a fact nobody in this discussion has established: that the method worked.
Take the outputs. Four criminal prosecutions, one verdict, no final judgment. One case dismissed on the constitutionality of the prosecutor’s appointment. One reduced to a remnant by the Supreme Court in July. One suspended pending an appellate ruling on the district attorney’s own conduct. A disqualification attempt rejected unanimously in March. Two civil judgments under appeal with security posted, one bond halved by an appellate court inside ten days.
That is a system with a very low completion rate and a high correction rate, and both facts belong in any honest assessment. I have spent this Forum’s American arguments waiting for someone to state the ratio and nobody has.
I work in comparative governance, and the observation I would offer is that the entire eighteen-question argument has been conducted without the data that any comparable jurisdiction would generate as a matter of course.
Every substantive claim in this Forum has been a claim about equal application. Would the same records theory have been built against an ordinary defendant? Would the same enforcement temperature have applied to a different name? Would the same venue, timing, and charging decisions have been made for someone else? Those are empirical questions with empirical answers, and the answers require charging statistics, declination rates, and a published policy against which decisions can be measured.
The United States has none of that for these decisions. There is no published charging policy for federal prosecutions of the kind at issue, no equivalent of the Prosecution Policy of the Commonwealth or the English Code for Crown Prosecutors, and no requirement that a county prosecutor demonstrate she has used a theory before. So the American correspondents have spent three years demanding comparables that the system does not produce, and their opponents have spent three years asserting neutrality that the system cannot evidence. Both sides are reasoning from anecdote because the architecture generates nothing else.
That, rather than any particular indictment, is the structural finding I would take from this period. A country cannot audit for equal application if it does not publish the standard or count the outcomes. And it cannot resolve a national argument about selective prosecution by assertion, which is what both camps have been reduced to.
The correction is unglamorous and entirely achievable. Publish the charging policy. Publish declination statistics in politically sensitive categories. Require an office invoking an unusual statutory theory against an officeholder to identify prior applications. Put the special counsel’s office in a statute rather than assembling it from four general provisions, which is what a federal judge found insufficient in July and what the Eleventh Circuit will now decide. Codify the speech standard the appellate court in Washington wrote in December. None of that is a foreign import; most of it exists in the American states in some form already.
I will state the alliance interest plainly, since I have been asked to answer as an analyst rather than as a partisan. Australia’s security and intelligence arrangements assume American institutional predictability across administrations. What allies price is not whether any individual received justice. It is variance — whether the outputs of American legal institutions can be forecast when the government changes. The last three years produced extremely high variance, and the correcting rulings, however sound, arrived far too late to reduce it.
So my answer is neither of the two on offer. The method did not succeed and it should not be described as though it had. The verdicts are real and should not be waved away. What has been damaged is the country’s capacity to demonstrate to itself that its law is applied equally, and that damage is structural, measurable, and repairable by ordinary administrative means that nobody in either camp is currently proposing.
