Nine years of writing about Amy Wax (b. 1953) has produced a large literature and almost no new information. The reason is that both sides stopped asking questions. Her critics ask whether she is a racist, which they have answered, and her defenders ask whether a university may punish a professor for her opinions, which they have answered. Each camp interviews her to confirm what it already holds. She has now given the same six answers for most of a decade, and she gives them well, because she has argued fifteen cases in the Supreme Court and knows how to hold a position under fire.
A question is good when neither camp can predict the answer. By that test, most of what gets asked of Amy Wax is not a question. What follows are the ones that survive it. Several are hard on her. Several are hard on Penn, and on the people who have spent nine years asserting that her central factual claim is false without producing any evidence that it is. A reader who finds only one of those categories useful has come for something other than the truth.
The first question comes out of a single year in her bibliography.
On July 31, 2020, on Glenn Loury’s (b. 1948) podcast, Wax said she would bet that innate factors are at work in group differences. That fall, in the University of Chicago Law Review Online, she wrote that the etiology of the disparities has been attributed to racism, poverty, culture, and innate group differences, and declined to choose among them. Same author, same year, same subject. The law journal brackets what the podcast asserts.
The pattern repeats two years earlier. Her 2018 Georgetown article distinguishes creedal nationalists, who hold that American commitments can be adopted by anyone, from cultural distance nationalists, who doubt it. She writes the second group in the third person from beginning to end. An assumption appears to operate among them. Some of them go further. She never says she is one. She endorses the policy conclusion, low and slow, which she has already told the reader creedal nationalists can reach on their own premises. In July 2019 she stood at the National Conservatism Conference, said the country will be better off with more Whites and fewer non-Whites, and told the press the speech recapitulated the article. It does not. The article attributes that position to people described from outside.
The question is what the bracketing was for. There is an innocent answer and she should be given room to give it: a legal argument that does not require a causal claim should not make one, and a scholar who writes only what his argument needs is doing his job. If that is the answer, it is a good answer, and it means the scholarship is the careful version and the podcast is the loose one, which has consequences for how much weight her admirers should put on the podcast. The other answer is that the journal audience could impose a cost the podcast audience could not, and that the hedges tracked the enforcement. She is the only person who knows which. Nobody has asked.
The second question follows from her own method.
Wax is at her best when she separates a causal claim from the evidence for it. Correlation establishes nothing. A laboratory effect need not explain a population gap. Identifying a channel does not identify a remedy. She has applied this to implicit bias, to stereotype threat, to structural racism, and to the claimed benefits of diversity, and she has been right often enough that her opponents have never answered her on the merits.
Then look at what carries her own arguments. The 2018 article opens by demanding dispassionate analysis, facts, reasons, evidence, and the details of the law, and builds that demand out of four anecdotes: a lunch with a former friend, an exchange between two cable hosts on election night, a newspaper column, and a tweet. It closes by resting a claim about the attitudes of Asian immigrants on three essays that crossed her desk, two of them undergraduate op-eds in a student newspaper. The 2020 essay tells courts to require employers to prove the benefits of a diverse workforce with accepted, well-defined, quantitative metrics, and to refuse deference to assertion. Two pages later it answers the question with Apollo 11, Bletchley Park, and the Antwerp diamond trade, and then states in the same paragraph that a priori generalizations are baseless.
One of those three is wrong. She describes the staff at Bletchley Park as mostly young, male, well-educated Caucasian British citizens. At its January 1945 peak Bletchley employed roughly ten thousand people and about three-quarters of them were women, most of them Wrens operating the Bombe and Colossus machines. She can retreat to the senior cryptanalysts, who were overwhelmingly men. She wrote about the staff.
There may be a principled defense here. A party asking government to coerce, redistribute, or reorganize an institution can be required to carry more than a party asking government to refrain. If that is her rule, the asymmetry is an allocation of burden, and it is the ordinary allocation in American law. So is that the rule, and where has she stated it? Because she has not stated it in the places where she relies on it, and without it the pattern reads as one standard for other people’s evidence and another for her own.
The third question is for Penn.
In September 2017 Wax said she did not think she had ever seen a Black student graduate in the top quarter of the class at Penn Law, and that they were rarely, rarely in the top half. Penn removed her from the required first-year course on March 13, 2018. Dean Theodore Ruger called the statement false. Penn has never released the data.
Consider what that means. A university stated publicly that a professor’s factual claim about its own students was untrue, holds the records that would settle it, and has declined for nine years to produce them. Wax has asked for them repeatedly. Every journalist who has written that her claim was debunked has written a sentence he cannot support, because nobody outside Penn’s registrar knows. The privacy objection is solvable, since aggregate class-rank distributions by race can be released without identifying anyone, and universities publish comparable figures routinely when the figures help them.
So the question runs both ways. To Penn: what do the numbers show, and if they contradict her, what accounts for nine years of not saying so? To Wax: what does she predict they show, stated as a number, and what result would make her say she had it wrong? She has spent a career insisting that empirical claims be treated as empirical. Here is one of her own. She should be willing to be tested on it, and Penn should be willing to run the test.
The fourth question. Wax is a theorist of the power of culture. People imitate, absorb expectation, respond to prestige, internalize convention, take habits from families. Norms grow strong enough to suppress basic appetite. In the immigration work, cultural habit is sticky enough that crossing a border does not erase it, and sticky enough to justify selecting newcomers by country of origin.
Race, Wrongs, and Remedies is partly an assault on explanations that make people seem too constrained by culture. Influence is not destiny. People can work, refrain from crime, marry, stay with their children, and resist a local expectation. She objects when social scientists use causal language that makes choice sound mechanical.
Both propositions can hold. The trouble is that each argument needs its proposition at full strength and needs the other one weak. If a destructive American subculture can be escaped through individual resolve, the cultural inheritance of an immigrant population should be escapable too, and the case for selecting by origin weakens. If inherited culture is sticky enough to justify that selection, it constrains the native poor more heavily than her rhetoric of self-help allows, and the moral weight of her prescription shifts. Her answer will run through the distinction between hard struggles and brick walls, which is in the book. The follow-up is where the answer lives: what evidence locates the line, and why does the line fall in a different place for the two populations?
The fifth question is about her prediction record.
In 2020 she argued that shifting affirmative action from remediation to diversity removed its endpoint. Remediation contains a stopping point because the rightful position is a destination. Diversity contains none, because composition is the goal. Sandra Day O’Connor (1930-2023) borrowed a twenty-five-year sunset from a framework her own opinion had discarded. Three years later John Roberts (b. 1955) treated that horizon as the outer limit of an indulgence and ended the practice.
She got the structure right. She also predicted that a significant tightening was improbable and expected Harvard to survive with modest adjustment. She read the incoherence and misread the appetite. So the question is what the error tells her about the model. Her account of elite institutions predicts that they absorb pressure and continue, and here an institution she had written off acted against its own preferences. What in the model produced the wrong answer, and where else does the model make the same mistake?
There is a live test running. Harvard’s class of 2029 came in at 11.5 percent Black, down from 14 percent. The figure that circulates in hereditarian argument, drawn from litigation exhibits, is that colorblind admissions would produce a class under one percent. Either Harvard is still reaching the old outcome through proxies, or the simulation never described what colorblind admissions produce. The gap between those two readings is large and it is now measurable. Anyone who cites the sub-one-percent figure and has not looked at the post-2023 numbers is asserting a prediction that the world has begun to answer.
The sixth question.
The free speech framing holds that Penn punished a professor for opinions expressed outside the classroom. The Penn framing holds that her statements damaged her ability to teach and evaluate students fairly. Both are stated as though the answer follows from the label.
The distinction that would settle it is one nobody makes. There is a difference between a professor’s claim about a subject and a professor’s public characterization of the performance of identifiable people in his own custody. Wax did not say that Black law students in general perform less well. She said that her own students at her own school did, in a class small enough that the students knew who was being described. A professor who says the same thing about a national dataset is doing empirical work. A professor who says it about the twenty people in his seminar is doing something else, and the something else is closer to an evaluative act than to speech about the world.
That distinction cuts against her and it also constrains Penn, because it defines a narrow ground rather than the open-ended one Penn used. If the harm is to identifiable students in her care, the remedy is removal from required first-year courses, which Penn did in 2018 and could defend. The sanctions imposed six years later, the loss of summer pay in perpetuity and of the named chair, reach conduct that has nothing to do with her own students, including the classroom guests she invited and what she said on podcasts.
So the question for her: does she recognize any category of statement about her own students that a university may act on, and where is its boundary? And the question for Penn: if the 2018 removal was the proportionate response, what happened between 2018 and 2024 to justify the rest, given that Penn’s own process took six years, stopped short of termination, and produced a sanction that a university confident in its case would have imposed faster and a university without a case would not have imposed at all.
The seventh question is asked of dissidents constantly and of institutions never, so it should be asked of both.
What has changed Amy Wax’s mind in thirty years, and what would? She has been criticized by people she respects, including Loury, who has pushed back on her in public more than once. Naming a single instance where that changed a position would answer the charge that her method has become a machine for reaching conclusions she already holds. Failing to name one confirms it.
The mirror is exact. What would change Penn’s mind? At no point in six years did the university state what evidence would have exonerated her, or what she could have said that would have ended the matter. A disciplinary process that cannot describe its own acquittal condition is not applying a standard. Her critics are entitled to demand falsifiability from her only if they can supply it themselves, and the record does not show that they can.
The last question is about audience, and it belongs to everyone in this story.
Wax is a Jewish woman who now speaks primarily to a coalition parts of which are openly antisemitic. She has touched this herself, in the exchange with Richard Hanania (b. 1985) about Jewish decline. She has never been asked what she thinks it does to a writer to be needed by an audience, and whether she has noticed her own positions moving toward the ones that audience rewards. The answer will not be a confession. It might be a description, and a description from someone with her analytical equipment would be worth more than another round on immigration.
The same question belongs to her opponents. The people who signed letters against her, who sat on the hearing board, who wrote the columns, also had audiences, and the cost of opposing her was zero while the cost of defending her was a career. An account of her behavior that runs entirely on her incentives and grants everyone else pure motive is a coalition document.
None of these questions has an answer I can predict, which is why they are on the list. The ones that get asked instead, whether she is a racist, whether tenure should protect her, whether the culture has gone too far, have answers everyone in the room already knows before she opens her mouth. Nine years of that has produced a great deal of writing and almost no knowledge. The grade data still sits in a registrar’s office. The Third Circuit has not ruled. She returns to a Penn classroom this fall. Somebody should use the time better than we have.
