The Numbers Don't Say That

Two men who spent a decade building the anti-woke machine have noticed it produces Nazis. Their proposal is a firewall.

Nineteen percent.

That is the figure sitting in the middle of a New York Times guest essay by Gregory Conti and Aaron Sibarium, offered as proof that young white men have been squeezed out of American professional life. White men were over thirty percent of American medical students in 2014, they write. By 2025 they were nineteen. The essay presents this as the grievance that built the backlash, the honest wound underneath the culture war, the thing that made the whole scorched-earth campaign against DEI a reasonable response by reasonable people to a real injustice.

Here is what actually happened to that number.

Women went from roughly forty-six percent of medical school matriculants to fifty-five point one percent. Male matriculants declined for six consecutive years starting in 2016-17. That is not a race preference story. That is women applying to medical school in larger numbers and getting in, which is the kind of thing that happens when you stop excluding half the population from a profession. No DEI office did that. No diversity statement did that. There is no policy you can repeal that puts those men back, because the thing that displaced them was other people showing up.

The rest of the gap is Asian American applicants, who are not DEI beneficiaries under Conti and Sibarium's own framework and in fact function as the movement's favorite evidence that preferences are unjust. You cannot count them as victims of the system in one paragraph and as displacers of white men in the next, but the number requires you to do exactly that.

And then there is the part that should have killed the essay in the draft stage. At the point of admission, white applicants are accepted at fifty point six percent and Asian applicants at fifty point five, while Black applicants are accepted at thirty-five point nine. After the Supreme Court gutted race-conscious admissions in Students for Fair Admissions, Black matriculation fell eleven point six percent in a single year and Hispanic matriculation fell ten point eight, while white and Asian matriculation rose.

The mechanism they blame was abolished. The share barely moved. It was never the mechanism.

The other two numbers are worse

When your evidence has no denominator, it isn't evidence.

The essay also reports that white men fell from thirty-nine percent of tenure-track humanities positions at Harvard to twenty-one percent, and from fifty-six percent of Amazon middle managers to thirty-four.

Neither carries a source. Not a link, not a citation, not a study name. In a Times op-ed, from a Princeton professor, in an argument that is doing nothing but statistical work.[1]

Set aside the sourcing. Share of a pool is not opportunity. The tenure-track humanities job market did not hold steady over that decade, it fucking cratered, and a smaller slice of a collapsing pie tells you approximately nothing about whether anyone was discriminated against. To make the claim they are making, you would need the denominators, the applicant pools, and the hiring rates. They provide none of it, because providing it would invite the reader to do the arithmetic, and the arithmetic does not go their way.

This is the load-bearing evidence for the proposition that the anti-woke backlash was a proportionate response to real harm. Three numbers. One that dissolves on contact with the AAMC's own data, two with no visible origin.

The stated version, presented fairly

Their argument deserves a real hearing before it gets taken apart.

Conti and Sibarium are not writing a defense of the right. They are writing a warning to it, and the warning is real.

Their case runs like this. Progressive excess in the late 2010s was genuine. The Republican response was organic, tracking public opinion rather than driving it. On trans athletes in particular they have the receipts: a New York Times and Ipsos poll from January 2025 found seventy-nine percent of Americans opposed to trans women competing in women's sports, including sixty-seven percent of Democrats. That is not a Fox News number. That is the Democratic coalition's own number, and the party's refusal to metabolize it was a real failure of nerve.

But now, they say, the backlash is turning into something "darker, dumber and ultimately self-defeating." The right has begun manufacturing its own invented identities and its own historical revisionism. It has stopped policing its perimeter. Tucker Carlson sat down with a Holocaust denier and the Heritage Foundation declined to do anything about it. A DOGE staffer fired for racist posts got reinstated with the Vice President's public backing. A speechwriter fired in 2018 for turning up at a conference full of white nationalists now has a senior job at the State Department.

Their prescription: build a firewall between anti-wokeness, which they consider justified, and actual bigotry, which they do not.

That is a coherent position. It is also, on inspection, a fantasy, and the essay itself contains the evidence for why.

The machine does not have a firewall setting

They want to keep the engine and discard the exhaust.

Their entire argument depends on the woke right being a separable byproduct. Something that got in from outside. The essay blames podcast bros and online streamers routing around official party messaging, plus a lost "immune response" to cranks.

Then it lists its examples: the Vice President of the United States. The president of the Heritage Foundation, which wrote Project 2025. A senior official at the State Department.

That is not the comment section. That is the movement's command structure. They cite the commanding heights as evidence of a problem originating in the periphery, and they never notice, because noticing would require asking a question they cannot afford: what if a political program organized around the proposition that white men are being displaced tends to attract people who are really, really into that proposition?

They offer no answer, because they have no mechanism. "Lost its immune response" is a description wearing a lab coat. What killed the immune system? The essay does not say. The parsimonious answer is that you cannot spend a decade telling people their country is being taken from them and then act surprised when someone shows up with a fuller theory of who is taking it.

They cite the commanding heights as evidence of a problem originating in the periphery.

The timeline runs backwards

The disease predates the infection they blame it on.

The essay dates wokeness's peak to 2022 and the collapse of gatekeeping to sometime after 2021. The causal arrow requires the right's racial radicalism to come after progressive excess.

It doesn't. Charlottesville was 2017. Birtherism was 2011. And their own showcase example of a healthy immune system, the 2018 firing of Darren Beattie for attending a conference attended by white nationalists, is proof that the material was already sitting inside the coalition in 2016, close enough to the President to write his speeches.

The counterfactual is worse. They argue a truce was nearly possible, citing Trump's Pride Month proclamation and his criticism of bathroom bills. That same administration rescinded federal guidance protecting trans students in February 2017, announced the military ban that July, argued in Bostock that federal civil rights law did not cover trans workers, and stripped health care protections in 2020.

Picking the one conciliatory press release out of an administration litigating the opposite direction for four straight years is not historical analysis. It's a fucking clip job.

What they are actually objecting to

Count the adjectives.

Read what the essay says against the woke right. It is self-defeating. It could cause defections. It has not yet cost the GOP what wokeness cost the Democrats, but it might. Every single objection is electoral.

Now read what it says about DEI. Antiwhite discrimination. Grievances. Diminished opportunity. Injustice.

One is a polling problem. The other is a moral wrong. The asymmetry is the entire argument, and it is visible without any interpretation at all, just counting.

The closing move makes it explicit. Most voters worried about DEI did not sign up for white nationalism, they write, just as most voters worried about Trump did not sign up for open borders. This asks you to accept white nationalism and open borders as matched coalition excesses, which requires believing that open borders was ever an actual policy of an actual administration rather than a phrase invented to lose an argument. They smuggle a Republican talking point in as neutral description, in the Times, under a Princeton byline, and the copy desk waved it through.

The bio is the story

This is what credential laundering looks like in production.

The Times identifies Conti as an associate professor of politics at Princeton. True and useless.

He is also a senior fellow at the Manhattan Institute, editor-at-large of Compact, and, per Princeton's own James Madison Program, a former senior fellow at the National Association of Scholars, an organization whose reason for existing is fighting DEI in higher education.

So the byline reads "Princeton professor and journalist express reluctant concern." The reality is "Manhattan Institute fellow and Free Beacon staff writer file a strategy memo." The argument's whole rhetorical engine runs on being mistaken for the first thing. The Times supplied the disguise for free, on the same page that has spent a decade letting Bretbug Stephens do the concerned-adult voice about other people's excesses.

The part I can't let the left off for either

Two things can be true, and one of them is embarrassing.

The seventy-nine percent number is real, and the sixty-seven percent of Democrats inside it is the part my side should sit with. That was a coalition telling its own leadership something, and the leadership decided the coalition needed to be educated. You can believe trans people deserve full civil equality, which I do, and still recognize that answering a fairness question about elite athletic competition with a lecture about who deserves to exist was a choice, and a losing one, and that the people who made it have still not admitted it was a choice.

That is the honest version of the failure, and Conti and Sibarium land on it before immediately using it as a permission slip for bathroom bills in twenty-one states, which polls nothing like the sports question and which Data for Progress found Americans much more divided on once you ask who should be writing the rules.

And on antisemitism, the asymmetry has to be said plainly. Progressive antisemitism is real. I have written about it and I am not going to soften it now: there is a chunk of the left that has decided a certain kind of conspiracy grammar becomes acceptable when the conspiracy is about Israel, and the movement's internal policing on it has been cowardly. Sibarium's own outlet has built a substantial identity on documenting it.

But Conti and Sibarium list antisemitism as a symptom of the woke right and then walk directly past what that means. Nobody on the left has an institution comparable to Heritage publicly declining to disassociate from a Holocaust denier on the grounds that disassociation would be cancel culture. The left's version gets punished, eventually, badly, in public. The right's version gets tenure at the think tank that wrote the governing agenda. Those are not the same magnitude and pretending otherwise is not evenhandedness, it is arithmetic performed with your thumb on the scale.

The mechanism is still running

There is no prescription here, because they didn't offer a real one.

What Conti and Sibarium want is the machine without its output. They want the DEI demolition, the campus purges, the federal contracting rollback, the twenty-one bathroom bills, and they want it all delivered without producing the constituency that shows up to collect. They have written eleven hundred words explaining that this is possible and provided three statistics, one of which is false and two of which have no source.

The firewall they propose requires abandoning the "no enemies to the right" doctrine, which they concede is not going to happen, for a reason they state with unusual candor: it has not cost anybody an election yet.

That is the tell, and it is the whole piece. The bigotry is not a problem to them until it starts losing. Until then it is a manageable externality, an unfortunate emission from an otherwise sound process, the price of a project they still think was worth it.

They built the thing. They are standing next to it. They have noticed it is making a noise.

Like what you've read, but not able to switch to paid? Why not buy me a coffee instead! All money goes to help defray the expense of running the site.

1 - I tried to validate these stats, and I found nothing.