I have been pro-trans for as long as the question has existed, and I put that on the record first because the rest of this essay leans on it, and because 2018 was the year it was put to the test. During those years, at an event I attended, a young woman took me aside to explain, with care, which boundaries I ought to respect as an ally in light of my status as a white heterosexual male. The conversation was not hostile in any ordinary sense. It was patient, with the air of somebody setting out the rules of a house I had been permitted to enter. I listened, said very little, and kept quiet about what I actually made of being credentialed before I was allowed to agree, because saying it would have gotten me chased out of my job. That was the culture, plainly, and anybody who worked in Silicon Valley then and tells you otherwise is describing a different place.
My objection was never to the cause. It was to the instruction, which ran in precisely the wrong direction: a movement with a public still to persuade was spending its energy auditing the people who already agreed. The arguments of those years ended on two sentences, “it’s not my job to educate you” and “check your privilege,” each received as a mic drop. You drop the mic when you have decided the audience does not matter. The gesture announces completion to the people behind you, and it tells the person in front of you that the conversation was never for them.
I was not pushed rightward by any of it, and my sympathy for trans people did not go then and has not gone since. Others around me, I watched go the other way: people who had started out in favour of gender-neutral bathrooms, people with no particular stake who had been inclined to decency, and who radicalised against the whole subject. They did not turn into bigots. They turned into people who had learned the subject was a trap, and once people believe a subject is a trap they stop discussing it and start resenting it.
The movements that won were not polite. Suffragists marched and went to jail, and the militant wing in Britain broke windows; civil rights marchers sat at counters they had been told to leave and filled the jails of Birmingham; gay liberation started with a bar fight against the police. None of it was gentle, and gentleness is not what I am asking for. What the movements had in common was confrontation addressed outward, staged for a public they intended to move, with every tactic measured by whether it moved them.
Take Proposition 6 in California, 1978, a measure to bar gay men and lesbians from teaching in the state’s public schools, one that led in the early polls. The opposition carried the slogan “Come out! Come out! Wherever you are!” from door to door in towns across the state, and Harvey Milk debated John Briggs in town halls and on television. What Milk asked of gay Californians was explicit about what it would cost, namely that they come out to family, coworkers and neighbours, because “you will hurt them if you come out, but think of how they will hurt you if they vote for Briggs.” The result was a defeat of 58 to 42, and the measure could not even win Briggs’s own Orange County. Here is a movement for which the unpersuaded were the entire object, shouldering a burden no one today would call fair, and shouldering it regardless.
A generation on, the marriage campaign made the same instinct into a method. Proposition 8 passed in 2008, after which Freedom to Marry re-examined more than eighty sets of polling data from state campaigns and found that talking about rights and benefits was not persuading conflicted voters, whatever its success with people already won over. In Oregon, 74% of voters told pollsters that couples like themselves married for love and commitment, whilst a plurality supposed that gay couples married for rights and benefits, which was the campaign’s own message coming back at it. The message was rebuilt around love, commitment and the Golden Rule, and the campaign devised what it called a journey narrative to show voters that one could change one’s mind on the question whilst staying faithful to what one already valued. Marc Solomon, describing the overhaul, adds that the campaigns had been doing too much finger-pointing and too much accusing people of bigotry, and that this stopped. Four states voted marriage equality in at the ballot box in 2012; no ballot had ever done so before.
Behind “love is love” was a movement that had studied, in detail, how its opponents thought, and had made its case to them. Within a few years the public face of LGBT politics had passed from “love is love” to aggressive language policing. No single actor was the movement, and I am not alleging that the marriage campaigners turned into the language police. What shifted was which wing set the public tone, and the wing that took over had drawn the opposite lesson from the one that had just prevailed.
Here, more than anywhere else in this piece, the target has to be named exactly. Trans people are not it; I take their dignity as given. Trans rights are not it; I support them. The target is the online enforcement culture which, from the first Trump term through the Biden years, read any departure from a linguistic orthodoxy as proof of bigotry and directed its considerable anger at allies who had phrased something wrong. To be set upon you did not have to be anti-trans. You had only to be slightly off the approved wording, and the wording changed often enough that staying current was a full-time occupation. In my adult life I have seen no political formation more counterproductive, and I say that as a person who wanted the cause to win.
Punish imperfect allies and everyone within the enforcement culture’s reach learns to keep quiet, and quiet is not consent. The persuadable middle learned the subject was dangerous to raise, and a subject people are afraid to raise is one they are easily frightened about, since no one they trust will discuss it with them. My own trajectory serves as the control: the same kind of encounter that drove people around me off did not shift me, so I make no claim that the culture manufactured bigots. My claim is that it worked on the persuadable, and that I watched it work.
Idaho became, in March 2020, the first state to bar transgender girls from girls’ school sports, and inside a year a version of its bill was before some thirty legislatures. Gallup put its first question about trans athletes to the public that May and found 62% saying they should play only on teams matching their birth sex; by 2023 it was 69%, with support for gender-identity teams down from 34% to 26%, and the reading this year is 69 to 24. Where that decline originated is what I would most ask activists to study. Among Democrats support went from 55% to 45%, among independents from 33% to 23%, whilst Republicans stood still, with nowhere further to go. Gallup’s question on whether changing one’s gender is morally acceptable moved from 46% acceptable and 51% wrong in 2021 to 38 and 57 this year. Pew’s measure of whether gender is determined by sex at birth climbed from 54% in 2017 to 56% in 2021 and 60% in 2022, the survey in which 64% also favoured laws protecting trans people from discrimination in jobs, housing and public spaces, and by February 2025 66% favoured requiring trans athletes to compete by sex at birth, eight points more than three years before.
Read together, the numbers tell one story. The public had not turned on trans people’s dignity; it had been losing the argument about everything beyond that, and those who moved were not the right. They were Democrats and independents, the persuadable middle the movement existed to keep, and they moved during the years in which the sports question went from hypothetical to ballot issue and the movement told anyone who raised it that raising it was bigotry. That the activists caused the shift on their own I do not claim; the right’s campaign ran through the same years, and sequence is not cause. Still, a question the public was plainly asking was ruled out of order, and a public whose question has been ruled out of order does not stop asking. It stops asking the movement.
On this the right built its closing message. Republicans spent roughly $215 million on anti-trans advertising across the 2024 cycle, by AdImpact’s count, the Trump campaign’s “Kamala is for they/them; President Trump is for you.” among it. NPR, working from the same tracker, put the party’s anti-trans and anti-LGBTQ total at $222 million, against under $29 million in Democratic LGBTQ advertising, with more than $46 million of it on the they/them message in the presidential race alone. The ads were built from Harris’s own statements of 2019. By “free material” I do not mean that activists wrote them. I mean that a political culture which had made the subject untouchable for its own allies left its candidate, five years later, alone with positions no one had been permitted to debate in public, before a public taught to keep quiet rather than taught to agree. The enforcement culture thought it was shielding vulnerable people. In the event it left them to face a $215 million campaign with no one persuaded.
How the left got itself into that state is a question I return to often, and what happened inside college to produce it I cannot tell you. I was not there, and because the theories I have read tend to prove whatever their authors already believed, I will not borrow one to plug the gap. What I can report is how it looked at the receiving end and what it cost the cause. The etiology belongs to those who were in the room.
Years of trying to understand reactionary propaganda, what persuades people and what fails to, have taught me a few sure things about human nature, and one of them is that almost no one will stand being told which words they may use. The literature happens to agree, for what that is worth. Psychological reactance theory, dating to Jack Brehm in 1966, holds that a threat to one’s freedom provokes anger and counterargument, and a recent meta-analysis across dozens of studies found that freedom-threatening language reliably raised anger and reactance and lowered persuasion. Being told which words are permitted threatens about the most personal freedom a person has. No theory is required to know it; having met a person is enough.
A single survey item carries the point and keeps it fair. In August 2017 Cato put questions about political correctness to 2,300 American adults, and 71% said it had done more to silence important discussions than to spare people offence. Half of Democrats agreed. Fifty-eight percent said the political climate kept them from saying what they believed. A law requiring the use of transgender people’s preferred pronouns was opposed 62 to 37, with 60% of moderates against. It is that last item I would ask activists to sit with, because it divides a courtesy most people will extend from a compulsion most people will resist. The enforcement culture erased the division, and the right’s propaganda fed on the erasure. I have argued before that reactionary propaganda works by taking genuine grievances and turning them into ammunition for conclusions reached in advance. This grievance was genuine, and that is what made it serviceable.
The deeper injury was internal. Once dismissing people earns more standing inside a movement than persuading them, the movement grows skilled at pleasing its own members and loses the ability to recruit anyone else. Approval from allies gets confused with progress, and the confusion can persist for years, because cultural power, for as long as it holds, conceals the difference: people comply, and from inside, compliance looks like agreement. Cultural power is borrowed. When the institutions changed hands, those who had trained only in dismissal found they had nothing, and the coalition they needed had been dismissed one purity test at a time. This is a diagnosis from what I saw rather than a measurement of every activist in America, though I have met few who lived through those years and dispute it.
The usual reply should be given its due, since part of it is true. People who have spent their lives arguing for their own dignity are worn down. Requiring them to go on explaining themselves to hostile strangers, patiently and on demand, is a burden the rest of us never carry, and the person requiring it usually has the luxury of finding the argument interesting. No one owes every abusive stranger an unlimited hearing. Scarce energy belongs where it can make a difference, and some conversations are not worth the trouble.
All true, and none of it entitles a movement to make the refusal to persuade its governing ethic. Well, I am exhausted too. Years of arguing for democracy with people who call it a scam, whilst the thing I argue for is dismantled in front of me, have not made me any less tired. And still I will not tell a right-winger that convincing them democracy is good is not my job, since it is exactly my job: my job as a citizen, the one office every one of us holds by right, and my weariness changes nothing about whose job it is. Anyone who means to fight for political power and social change needs to know how to fucking debate, and “I’m exhausted” is no answer to that requirement, merely a description of how the requirement feels.
No movement that seeks power over its fellow citizens can make the refusal to persuade them its principle and then be surprised when they decline to be governed by it. That is the bargain democracy really offers. Convince enough of my neighbours and you may have power over me, and the convincing is not an insult to your cause but the whole of the bargain. I ask no one to do work I have excused myself from. I ask them to do the work I do, tired, daily, because the alternative is losing, and the people on the other side are not tired in the least.
I owe the reader an accounting for a sentence of my own. In these pages I have said that bridge-building is not my job, that my job is to stay clear about what is happening and what it means, and that someone else can build bridges to the people I have written off. A hostile reader will lay that beside this essay’s title, and he will be right to. So here is the difference.
Moral witness gives its reasons in public. Every essay I write is an argument addressed to anyone who will read it, the people it condemns included, and it explains itself. What witness will not do is soften its verdict on people who already know what they are doing. Megyn Kelly does not require me to explain why calling for a dictatorship to crush liberal culture is sedition; what she requires is that I say it without softening, so the record shows someone did. “It’s not my job to educate you” is another sentence entirely. It gives no reasons at all, and in my experience it was said not to collaborators but to allies, people who had come in good faith and asked why. The first refuses flattery; the second refuses argument. Had I ever told a persuadable person that their confusion was not my problem, I would deserve the very criticism I am making here, and I would expect to get it.
⁂
A part of the left went past dismissal. It grew wary of free speech as such, came to see the First Amendment as an obstacle rather than an inheritance, and some of it said aloud that the amendment was too broad and ought to be narrowed to allow the criminalisation of hate speech. I watched this as a liberal and with something close to horror, because I knew what I was looking at: a loaded gun passed to people who had been waiting for one.
I will not inflate this into a majority of the left, and the evidence does not ask me to. In the same 2017 Cato survey, 52% of Democrats agreed that government should prevent people from engaging in hate speech against certain groups in public. FIRE found in July 2024 that 61% of Democrats at least slightly agreed that the First Amendment “goes too far in the rights it guarantees,” though 52% of Republicans agreed too, which tells you the censorious instinct belongs to neither party alone. This spring Knight and Ipsos asked more than a thousand Americans between 18 and 29 whether hate speech, defined for them as speech attacking people on the basis of race, religion, gender identity or sexual orientation, ought to be legally protected. Forty-two percent of young Democrats said it should, against 70% of young Republicans, and 71% of all young adults said it should not be allowed on college campuses.
Those are attitudes rather than election results, and I will not pretend that a poll explains November 2024. What the numbers record is a forfeiture. The left held a distinction the right could never have claimed, between defending a person’s right to speak and endorsing what they say, and it let the distinction slip. The right’s free-speech activism was never offered in good faith; the speech warriors went silent the instant the state began punishing speech they disliked, and I have said so at length. None of which alters what the left gave them to work with. Responsibility is not a fixed sum where one side’s portion is subtracted from the other’s. The right’s vast propaganda apparatus did what it did, and the people who voted for authoritarianism own that vote. The censorious left handed over the material and has never accounted for it. Both are true, and I refuse to pick between them.
In 2018, I am fairly certain, I could not have published this essay without being excommunicated from left-of-centre politics, and I do not mean argued with. I mean removed. That I can publish it now does not show that the culture reformed. It shows that the culture lost its power to enforce, which is a different matter.
Since the re-election that element has fallen silent, and silence is no admission. It turns up on BlueSky now and then, in the familiar register, and when it does I can see that nothing has been taken in. The people who spent years treating the First Amendment as a problem to be managed have not decided they were wrong; they have decided the moment is wrong. My anger at them remains, and its object has to be stated exactly, because the imprecise version is the one the right would rather I used. It is not Gen Z, nor people who are tired, nor people who wanted dignity and said so with heat. It is a political culture that held persuasion beneath it and the Constitution an inconvenience, and that has yet to say so aloud.
I know what this essay could become. One version of it, which I have described elsewhere, goes: Trump is bad, but political correctness run amok produced him, so none of this would have been needed had the left behaved. That argument puts the left’s failures to work excusing the right. This one puts them to work demanding more of the left, and it holds the right answerable for every single thing it did regardless. The difference lies in where the argument points. Mine points at my own side, because my own side is the only one I can still argue with.
⁂
Should someone use a word that offends you, the effective course is to follow up afterwards and tell them why it landed as it did. Invite their empathy. Offer them the chance to be the person they would like to believe they are. A call-out in front of others puts people on the defensive, and a person on the defensive is not listening to you; he is composing his reply. And when someone objects to being called out and the answer they get is “it’s not my job to educate people,” the sentence might as well be “fuck you.” It will be heard as “fuck you.” It will be remembered as “fuck you,” and the next time the subject arises that person will remember who said it and vote accordingly.
I am not alone in thinking this, nor the best credentialed. Loretta J. Ross, a Black radical feminist and MacArthur Fellow who teaches the practice at Smith, calls it “calling in,” and her example is almost exactly the script I would offer: “When you use that word, it didn’t land on me quite correctly. Do you think we can go have coffee and talk about it?” About the alternative she does not mince words. “Calling out is an invitation to a fight,” she says, and calling out is about power, a means of seizing control of a group and ending the conversation. Her book on the subject, Calling In, keeps the public call-out in reserve as a last resort for the powerful and the unrepentant. A student of hers put the whole matter into a sentence I wish I had written: “a callout is not an invitation for growth. It’s the expectation that you’ve already grown.” Ross came to this from inside the movement, after a lifetime in it, and I raise her so that no one mistakes the prescription for a centrist’s advice from outside. She also concedes what I concede: no one is obliged to call in every stranger. The objection is to a culture that made refusal the default and explanation the exception.
How the left got itself into that state is something I think about, and worry about, a lot, and I have told you I do not know the answer. For a diagnosis without an etiology, that is the honest place to end. A formation whose causes are not understood can come back, the people who built it have not accounted for it, and the conditions that produced it, a movement that cut off its own feedback by punishing those who supplied it, are not hard to recreate. The title of this essay, then, is insurance, not nostalgia for a better left.
Whoever wants to change the world should tell me what they are doing to persuade someone who does not already agree with them. Not what they are doing to signal, or to hold the line, or to spare their own people the exhaustion of the argument. What are you doing to convince a person who could go either way? Tired as I am, I think the obligation holds regardless, and so I intend to go on arguing for democracy with people who may never thank me for it, because the alternative to persuasion is not purity. The alternative to persuasion is force, and at force the other side will always beat us.
Go Deeper into the Circus
The Map and the Landscape: A Generation Lost in Space
I used to think that anyone who disagreed with me on economic matters just needed a better education. The numbers were there (GDP was up, the market was doing well) and I was doing quite well in SV. The whole thing was self-evidently working as it should, the system having obviously recognised my own merits. I was qu…
The Broken Windows of the Lost Libertarian Utopia
You can spend your whole life lecturing people about incentives and still need someone to explain why a man might cheat if it pays. Donald Trump’s teleprompter operator was betting on what Donald Trump would say.
This 1000 times over. What’s great is that it gives me the words to frame these distinctions. Thank you thank you thank you.