What Money Can’t Buy

I. Six Articles, One Question

The architects wrote it down. Mann’s balance wheel. Morrill’s land. Cubberley’s factory. Dewey’s prediction, published in the year the Smith-Hughes Act was being drafted. The architecture was built openly, in print, by named people whose own annual reports and federal statutes are still on the shelf at any research library in the country. The mythology that the system was built for democratic flourishing is a retrofit. The system was built for sorting, for labor discipline, for the Protestant moral reform of immigrant children, and for the production of a docile and literate workforce on terms set by the men who owned the mills. Mann reached for a steam-engine governor as the metaphor for his life’s work because he wanted his readers to understand the work as a part inside a machine. The architecture is still running. (Article 1.)

The betrayals were named, voted on, written into court opinions. The Freedmen’s Bureau defunded in 1872 after building 1,500 schools serving 100,000 Black students. Plessy in 1896. Cumming in 1899 — the same Justice Harlan who had written the color-blind dissent three years earlier, now writing for a unanimous Court that the only Black high school in a Georgia county could be closed while the white one stayed open. The HOLC residential security maps of 1933 to 1940, encoding race as risk in 239 cities. The GI Bill written race-neutral and administered through the redlined plumbing — two of three thousand two hundred and twenty-nine Mississippi VA loans to Black borrowers in 1947. Brown in 1954 and Prince Edward County’s five-year school closure rather than integrate. Swann in 1971, written by Chief Justice Burger for a unanimous Court, authorizing busing and the brief moment integration actually worked. Milliken in 1974, written by the same Chief Justice Burger, three years later, 5–4, holding that desegregation remedies could not cross district lines. The pipeline from 1872 to 1974 to the 2024 per-pupil spending gap runs through named decisions made by named people whose votes are in the record. (Article 2.)

The abundance was real, briefly, for a specific cohort under conditions that cannot return. Public four-year college tuition in 1964–65 was two hundred and fifty-six dollars. Two hundred and five hours of summer minimum-wage work covered the year of tuition with change to spare. The same arithmetic in 2022 required one thousand five hundred and ten hours. The ratio is the entire story. The mythology survived the conditions. The conditions did not. When Boomers tell their grandchildren they worked their way through college, they are not describing a virtue the grandchildren lack. They are describing a ratio that no longer exists. The arithmetic does not care who tries hardest. It only adds. (Article 3.)

The reproduction has been continuous since 1636. Lowell’s 1922 letter is in the Harvard archives. The classification system he invented — J1, J2, J3 — is still in the admissions files, not folklore. The personal essay, the “character” interview, the photograph requirement, the geographic distribution requirement to dilute the urban Northeast, the legacy preference invented in the same crisis for the same purpose — all of it engineered between 1922 and 1926 by a Harvard president who wanted to exclude Jews. The 33.6 percent legacy admit rate is in Peter Arcidiacono’s expert report from the federal litigation. Eighty-six percent for recruited athletes, 42 percent for the Dean’s Interest List, 30 percent of the admitted class drawn from 5 percent of the applicant pool. Rick Singer’s “side door” was a cheaper unlicensed franchise of the front door elite universities run openly. Stony Brook produces nearly four times the upward mobility per student that Harvard does, not because Stony Brook is a better school but because Harvard does not admit poor kids in the first place. The mechanism requires the opacity. The credential’s value depends on the buyer not seeing how the credential was distributed. (Article 4.)

The mythology is morally corrosive even when it works. The rhetoric of rising tells losers they deserved their fate and tells winners they earned what was given. Both halves rest on the same premise — that the sorting tracks moral desert — and the premise is the pathology. Even a perfectly fair sorting machine produces hubris on one side and humiliation on the other when both sides believe the sorting reflects who they are. Meritocratic hubris is the inability to see one’s position as gift, accident, or contingency. The dignity of work, on Sandel’s account, is what the rhetoric of rising has been hollowing out for forty years — not by malice, but by the cumulative effect of telling the non-credentialed that their loss in the sorting is the measure of their worth. (Article 5.)

The class that won the game protects the game. Not the one percent. The broader winners of the meritocratic order — the credentialed professionals who organized their lives around preserving what they were told they earned. Geographic mobility cut in half since the 1980s. Startup formation halved in the same window. Cross-class intermarriage gone. Income segregation rising sharply. Matching technology — the Common App, Zillow, Tinder, Spotify — made the sorting more efficient and mobility harder. Cowen named it. The author of this series writes from inside the apparatus, not above it; the architecture being described is the architecture the writer has used; there is no view from outside. Recognition is the precondition for any change that does not begin with the same complacency under a different name. (Article 6.)

Six articles. One question remains, and this article is about the one question.

What is the series actually arguing?

That sentence is the entire purpose of this final article. Six articles of historical reconstruction and structural analysis are not, by themselves, an argument. They are a description. The argument is the move that gathers the description into a claim, and the claim has to be made plainly, in the open, where the reader can hold it and turn it over and decide whether it is true. The argument is not a policy program. It is not a manifesto. It is not a list of villains or a plan to defeat them. The argument is a distinction. The distinction belongs to Michael Sandel, but the work of putting it on top of six articles of American educational history belongs to this series, and the work of articulating it as an editorial position belongs to this final piece.

The pipe metaphor for this article is the synthesis pipeline — all four temporal arcs assembled in one frame, all five scholars deployed, the carpenter/house distinction developed in full. The four arcs are not separate stories. They are one story told through four temporal windows. The architecture (1840s–1917) explains what was built. The Reconstruction betrayal (1865–1974) explains who was excluded from it. The postwar abundance (1945–1975) explains the brief window during which the architecture appeared to deliver on the mythology, and why that appearance was historically contingent and racially partial. The reproduction engine (1636–2026) explains the parallel system that has operated beside the public architecture for almost four hundred years and that has progressively recaptured the partial opening of the postwar window. The four arcs interlock. The interlocking is the story. This article holds them all in view at once and locates them inside Sandel’s framework of moral corruption and the carpenter/house distinction. Then it draws the distinction explicitly.

Here is what the rest of the article does, in order. It begins with what the series is not arguing, because the easy misreadings are the ones that have to be cleared first. It develops Sandel’s two objections to market commodification — the fairness objection and the corruption objection — because the corruption objection is the one most readers have never been offered as a separate claim and the one the series rests on. It anchors the corruption objection to the Israeli daycare fine study, which is the empirical demonstration that market norms do not supplement moral norms but replace them. It walks through Walzer’s spheres of justice, which is the framework Sandel inherits and that gives him the vocabulary for what has happened to American education. It develops the carpenter/Latin tutor/admissions consultant distinction in the three escalating cases the series has been building toward. It engages Wooldridge’s steelman for meritocracy honestly, then makes the deeper move that even the least-bad sorting mechanism is corrosive when its winners stop seeing their position as contingent. It states, in a labeled section near the end, what the series is and is not arguing. It closes on Sandel’s conjunction.

The closing move is the one the prior series have prepared the reader for. The conjunction is and. Some goods are corrupted by markets and some goods are fairly sold by them. Education is both kinds of good at once. The task is telling them apart.

II. The Two Objections

The most common misreading of What Money Can’t Buy is that Sandel is making the standard left-liberal complaint about inequality — the complaint that markets allocate goods unfairly because the parties to the exchange are bargaining from unequal positions. That complaint is real and Sandel takes it seriously. It is not the complaint his book is making. The book is making a different one. The second one is the deeper one, and the series has been resting on the second one for six articles without naming it. It needs to be named now.

Sandel distinguishes two distinct moral arguments against bringing something into a market. The distinction matters because the two arguments apply in different cases and have different remedies, and confusing them is the mistake the standard reading makes.

The fairness objection. Markets produce unjust outcomes when the parties to an exchange are bargaining from unequal positions. A peasant may “consent” to sell a kidney, but if he is selling to feed a starving family, his consent is not meaningfully voluntary. Coercion takes a softer form than the boot at the throat. A market in kidneys, in surrogacy, in passport sales, in admissions slots, will be a market in which the rich buy from the poor and the poor sell to the rich, and the poverty of the seller is what makes the sale possible. This is the standard inequality complaint and it is right as far as it goes. The fairness objection can, in principle, be met by equalizing background conditions. Redistribute wealth sufficiently and the sale becomes voluntary in a meaningful sense. The objection is about the conditions of the exchange, not about the nature of the good being exchanged.

The fairness objection is the move that Sandel’s critics on the right complain is class war by other means. It is also the move that Sandel’s readers on the left assume is the whole of his argument. Both readings are wrong about Sandel. Sandel grants the fairness objection, uses it where it applies, and then makes a different argument that does not depend on it.

The corruption objection. This is the deeper argument. Some goods are corrupted — their meaning is changed, their value diminished — by being bought and sold at all, regardless of how fair the bargaining conditions are. A Nobel Prize cannot be purchased, and the reason is not that the purchase would be unfair. The reason is that the thing purchased would cease to be a Nobel Prize. The Nobel is constituted by what it represents: peer recognition of work done for its own sake. A purchased Nobel is no longer a Nobel. The transaction does not just transfer the object; it changes the object into a different object whose name no longer applies. Friendship cannot be purchased; a paid friend is not a friend. Love cannot be purchased; a paid lover is not a lover. The corruption objection holds that market norms, when they enter certain spheres, crowd out the norms that constitute the good in question. And crucially: the corruption can happen even when the exchange is perfectly fair — even when the parties consent freely, even when the prices are transparent, even when the bargaining is between equals.

This is the move the series has been resting on. The corruption objection is the philosophical engine of the carpenter/house distinction that the rest of this article will develop. Without it, the series collapses into the standard inequality complaint, which it is not making. With it, the series has a different and more uncomfortable claim: that some of what American education sells is corrupted by being sold at all, not because the prices are too high or because the buyers are too rich, but because the thing being sold cannot survive being sold without becoming a different thing. The corruption is not a problem of fairness. It is a problem of category.

The reason this distinction is unfamiliar to most readers is that contemporary American political discourse runs almost exclusively on the fairness objection. The progressive case for reform is the case for redistributing access. The conservative case against reform is the case for protecting voluntary exchange. Both sides assume the question is whether the market is operating freely and fairly. Sandel’s contribution is to insist that the question is prior to fairness. The question is whether the good in question is the kind of good that survives commodification at all. Some goods do. Houses do. Cars do. Wheat and sugar and televisions do. These are goods whose meaning is not changed by being priced. Other goods do not. Sandel’s book is a long argument that the second category is larger than American discourse allows, and that some of the most important goods in human life — love, friendship, civic membership, recognition, dignity, education — live in the second category for at least some of their dimensions.

The series is making the same move about education. Education is partly the first kind of good and partly the second kind. The instruction is the first kind. The credential is the second kind. The series’ argument depends on the reader being able to hold both halves of that sentence at once.

III. The Daycare Fine

The most important empirical example Sandel deploys for the corruption objection is a 1998 study of ten daycare centers in Haifa, Israel, conducted by Uri Gneezy and Aldo Rustichini. The story is well known among economists and not nearly well enough known elsewhere. It deserves a paragraph of careful retelling because everything in the rest of this article depends on what it shows.

The daycare centers were having a problem with parents picking up their children late. The teachers were frustrated. The contracts said that the day ended at four. Parents were arriving at four-fifteen, four-thirty, sometimes later. The teachers had to stay until the last child was collected, which meant they were working unpaid overtime several days a week and resenting the parents who were causing it.

The daycare centers tried a fix. They imposed a small fine on parents who picked up their children after the contracted hour. The fine was modest — a few shekels — but it was real, and it appeared on the parent’s monthly bill. The economic logic was clean. Parents had been free-riding on the teachers’ willingness to absorb the cost of late pickups. Putting a price on the externality should internalize it. Late pickups should fall.

Late pickups rose. Not by a small amount. By a large amount. They roughly doubled.

The researchers had built a control group of daycare centers that did not impose the fine, and the comparison made the result clean. The fine did not supplement the existing norm. It replaced it. Before the fine, parents had been late occasionally and felt guilty about it. After the fine, parents were late more often and did not feel guilty about it, because they had paid for the time. The transaction had converted a moral relationship into a commercial one. The parents had not become worse people. They had become customers. The teachers were no longer doing the parents a favor by staying late. They were providing a service that the parents had paid for. The price was the new norm, and the new norm produced more of the behavior the old norm had been suppressing.

Then the researchers did the second part of the experiment, which is the part that matters most for the series. After several weeks, the daycare centers removed the fine. The economic logic predicted that late pickups should return to baseline, because the price had been removed and the original incentive structure restored. They did not return to baseline. They stayed elevated. The norm that had been crowded out by the fine did not come back when the fine went away. The moral relationship the fine had replaced was gone, and the absence of the fine did not regenerate it. What was left was the absence of the price, and the absence of the price did not recover the norm that the price had displaced.

This is the empirical demonstration of the corruption objection. Market norms do not supplement moral norms. They replace them. And the replacement is not symmetric. Once a moral relationship has been converted into a commercial one, the conversion does not run backward when the commerce is withdrawn. The norm has been destroyed, not suspended. The fine did not just fail to fix the late-pickup problem. It changed what late pickup meant in the minds of the people doing it. And the change was permanent.

Apply this to American education. The argument the series has been building is that the imposition of market logic on the credential-distribution function of elite higher education has done the same thing the daycare fine did to late pickup. It has replaced the moral relationship between an institution and the students it serves with a commercial relationship between a sorting mechanism and the families that buy access to it. The Latin examination at colonial Yale was the original credential, and it was a credential that depended on an actual competence the student had to acquire. The 1922 Lowell apparatus was the moment when the credential became uncoupled from competence and re-coupled to the photograph and the recommendation letter and the geographic distribution. The development cases and the IEC industry and the Common App and the ALDC categories are the modern apparatus for converting family wealth into credential. Each of these conversions is the daycare fine being imposed on a different layer of the system. And each of them has produced the same result: the moral relationship the conversion replaced does not return when the conversion is exposed. The Arcidiacono numbers, when they were finally made public in federal litigation, did not cause anyone to stop applying to Harvard. They caused the families who could already pay to pay more carefully. The norm that the credential was supposed to embody — that the institution selected on demonstrated promise rather than purchased pedigree — was already gone, and the disclosure of the numbers did not bring it back.

The Sandel reading of this is that the corruption is structural, and structural corruption is not undone by transparency. Transparency is necessary — the series has been arguing for six articles that the receipts have to be put back in the same room with the mythology — but transparency by itself is not sufficient. The norm has to be rebuilt, not merely the price disclosed. And the norm cannot be rebuilt by the same institutions that converted it into a price in the first place, because those institutions have an existential interest in the new arrangement persisting. The 1922 letters had to be hidden for the apparatus to function. The Arcidiacono numbers had to be obtained through litigation against the institution’s active resistance. The Harvard general counsel did not file those motions because Harvard had nothing to hide. Harvard filed those motions because the disclosure of the numbers was an attack on the credential’s value, and the credential’s value is the institution’s product, and protecting the product is what general counsel exists to do.

The daycare fine is the empirical anchor for the corruption objection because it demonstrates, in a small, well-controlled, replicable case, that a moral relationship can be destroyed by the imposition of a price even when the price is small, even when the parties consent, even when the bargaining is between rough equals. And it demonstrates that the destruction is not reversed by removing the price. The norm is gone. Apply this to four hundred years of American higher education and the implications are large. The corruption is not in the side door. The corruption is in the conversion of credentialing from a moral relationship into a commercial one, a conversion that began in 1922, accelerated through the IEC industry from the 1980s on, and reached its current form in an apparatus where 30 percent of the admitted class at Harvard comes from the 5 percent of applicants whose families have learned how to convert capital into the appearance of merit. The fine is on the bill. The norm has been gone for a long time.

IV. Spheres

Sandel did not invent the framework he uses for the corruption objection. He inherited it, and he is generous about saying so. The framework is Michael Walzer’s, from a book published in 1983 called Spheres of Justice: A Defense of Pluralism and Equality. Walzer’s book has not aged into the canon the way Rawls’s Theory of Justice has, partly because Walzer was writing against the dominant Rawlsian frame and partly because his argument is harder to operationalize as a single distributive principle. But it is the philosophical apparatus underneath everything Sandel has done since, and it is the apparatus the series needs in order to develop the carpenter/house distinction in its full form.

Walzer’s claim is that justice is not a single thing. Different goods belong to different spheres, and each sphere has its own internal logic and its own rules of distribution. Money governs the sphere of material goods — food, shelter, transportation, the things you can fairly buy with money because they are the kinds of things money was invented to allocate. But other goods belong to other spheres. Love belongs to a sphere with its own logic; it is given and reciprocated and earned over time, not purchased. Political power belongs to a sphere with its own logic; it is conferred by election or appointment or trust, not bought. Honor belongs to a sphere with its own logic; it is awarded by the recognition of peers, not by the highest bidder. Divine grace, on the religious view, belongs to a sphere with its own logic; it is given freely or earned through devotion or both, but it is not for sale. Education belongs to a sphere with its own logic; it is the result of attention, instruction, struggle, and the accumulation of competence, and it cannot be transferred from one person to another by transferring money.

Injustice, on Walzer’s view, occurs when one sphere invades and dominates another — when money buys political power, when wealth purchases admission to elite education, when grades become purchasable, when a campaign donation becomes the price of access to a senator. The injustice is not in the unequal distribution of money; some unequal distribution of money is compatible with justice in any framework that recognizes any role for economic activity at all. The injustice is in the conversion of money into a good that does not belong to the sphere money governs. The conversion is what corrupts both spheres at once. The sphere that money invades is corrupted because the good in question is now allocated by a logic foreign to its nature. And the sphere of money is corrupted in its own way, because the things money has been converted into now teach it to behave as though everything were for sale. The corruption is mutual. The remedy is to restore the boundaries between spheres and to insist that the good of each sphere be distributed by the logic native to it.

Walzer’s framework gives Sandel the vocabulary for what has happened to American education. The sphere of learning has been invaded and dominated by the sphere of markets, and the internal logic of learning has been deformed by the invasion. The deformation is not a failure of fairness inside the sphere of learning. It is the substitution of the logic of one sphere for the logic of another. A child who learns Latin from a tutor and a child who has a tutor write a personal essay that the admissions office will read as evidence of the child’s mind are doing different things in different spheres. The first is learning. The second is purchasing the appearance of learning. The first is the sphere of education operating by its native logic. The second is the sphere of markets operating in a sphere it does not belong in. The two activities look similar from outside, and the apparatus of contemporary American higher education has spent forty years training families and students and admissions officers to treat them as the same activity. They are not.

Walzer’s and Sandel’s point is that both spheres have to remain intact for either to function. The sphere of education cannot do its work if it is being colonized by the logic of markets, because the colonization replaces the native distributive principle (attention, instruction, struggle, competence) with a foreign one (purchase, signaling, optimization, scarcity). And the sphere of markets cannot do its own legitimate work if it is being asked to allocate goods that do not belong to it, because the legitimacy of market allocation depends on the goods being allocated being the kinds of goods markets can fairly handle. Asking the market to distribute Harvard admissions is asking the market to do something that no market can do without corrupting both itself and the credential. The market becomes a vehicle for laundering inheritance. The credential becomes a token whose meaning is the inheritance it laundered. Both spheres are degraded by the arrangement.

The reason the series needs Walzer to make this argument is that the fairness objection, by itself, cannot reach it. The fairness objection assumes the question is whether the existing distribution inside the market sphere is fair. Walzer’s question is prior: whether the good in question belongs in the market sphere at all. If it does not, then no distribution of the good through the market is fair, no matter how equal the bargaining positions. The corruption is in the conversion, not in the price. The series has been resting on this prior question without naming it for six articles. This article names it.

V. Three Cases

What follows is the move the entire series has been preparing for. It is three cases — not three categories, three cases — that walk the reader from honest commerce to corrupted commerce on a slope rather than a binary. The cases come from Sandel by way of Walzer, but the framing as a sequence of three escalating cases belongs to the series. The first case is honest commerce in its purest form. The second case is still honest commerce, sitting closer to the line. The third case is the case the corruption objection is built for. The slope is what matters. The reader who thinks the difference between case one and case three is a difference of degree is missing the argument. The reader who thinks the difference is a difference of kind, on the other hand, has been given the distinction the series exists to draw.

The slope from Case One to Case Three is the philosophical work of the series. Each case adds something to the previous one, and the something it adds is a step toward the corruption. Case One is pure exchange of labor for money in a sphere where money belongs. Case Two is exchange of expert labor for money in a sphere where money is doing work the sphere can in principle accommodate. Case Three is exchange of money for a position whose value depends on the exchange being invisible. The first two cases are the kinds of transactions any society has to make in order to reproduce its own crafts and competences. The third case is the kind of transaction Walzer says corrupts both spheres — the sphere of education by replacing its native distributive logic with a foreign one, and the sphere of money by teaching it to behave as though everything were for sale. The first two cases scale without harming the good. The third case devours itself when it scales, because if everyone bought the admission the admission would be worth nothing.

The crisp formulation, the one the series has been carrying toward this paragraph for six articles: honest commerce pays for the good itself. Corrupted commerce pays for position in a sorting mechanism whose value depends on its purchase being invisible. The load-bearing word is invisible. The reason corrupted commerce corrupts is not that money changed hands. It is that the value of the credential depends on people not knowing it was bought. The moment everyone knows that a Harvard admission was secured by a $2.5 million development donation, the admission stops signaling what it was supposed to signal. The signal value is the secrecy. The sorting requires opacity to function as sorting.

This is why the series has spent so much of its length on disclosure. The Lowell letters had to be hidden for the apparatus to function. They are not hidden anymore. The Arcidiacono numbers had to be obtained through federal litigation against the institution’s active resistance. They are now in the record. The Singer wiretaps had to be played in court before the public could see what its own elite institutions were doing in plain sight. They have been played. The HOLC residential security maps had to be filed in cabinets at federal agencies that did not advertise them. The University of Richmond’s Mapping Inequality archive has put them on the open web. The 86 percent recruited-athlete admit rate had to be a closely held internal number. It is now an open citation in any high school senior’s college search. None of these disclosures has destroyed the apparatus they describe. The apparatus still operates. But each disclosure has chipped away at the opacity, and the opacity is what the credential’s value depends on, and the chipping away is the closest thing to structural intervention that journalism can provide. The series, by putting six articles of receipts back in the same room with the mythology, is participating in the chipping. The chipping is the work.

VI. The Confusion

Here is where the carpenter/house distinction does its hardest work. The mythology of American education does not deny that money plays a role in the system. Every defender of every elite institution will, when pressed, acknowledge that the apparatus runs on dollars. Tuition. Endowments. Donations. Test prep. Tutoring. Application fees. Room and board. The acknowledgment is uncontroversial. The defenders also do not deny that some of the dollars produce sorting outcomes that are uncomfortable to look at directly. The Arcidiacono numbers exist. The defenders read them and find ways to defend them. None of this is in dispute.

What is in dispute is the word access. The mythology of American education uses the word “access” for both kinds of commerce at once and asks the listener to treat them as the same kind of thing. When a public university tells a low-income student that the Pell Grant will cover tuition, it is talking about access to instruction. When an elite private university tells a donor family that a $2.5 million gift will help a student application, it is talking about access to the credential. Both uses of the word are common. Both are heard in the same conversation. Both are treated as instances of the same general concern about “access to higher education,” and the political coalitions that organize around expanding access run the two together as though there were no difference between them. There is a difference. The difference is the difference between Case Two and Case Three. The instruction is the kind of good that is well distributed by paying for it. The credential is not. Pretending the same word covers both is the move the mythology requires in order to sound fair.

The question that almost never gets asked is the question access to what? Access to what, exactly? If the answer is “access to instruction,” then the case for expanding access is a case for making the sphere of education function according to its native distributive logic — pay for the labor, fund the curriculum, train the teachers, build the classrooms, give the students who want to learn the room to learn. Honest commerce. Case Two at scale. The carpenter analogy holds. If the answer is “access to the credential,” then the case is something else entirely — it is a case for redistributing position inside a sorting mechanism whose value depends on the sort, and the redistribution does not destroy the sort, it just shuffles who is on which side of it. The credential’s value to its holder depends on most people not having it. Expanding access to it — without changing what it is — does nothing for the people who get it that the credential they got was not engineered to produce.

The mythology folds these two questions together because the folding is what the apparatus requires. If the questions were separated, the public would have to choose, and the choice would not be friendly to the elite institutions whose business model is the conflation. The elite institutions sell the credential and then justify the sale by pointing to the instruction, as though selling the credential were the same thing as selling the instruction. The instruction is real. The instruction is, in many cases, excellent. The instruction does not require the sort to deliver itself. Stony Brook teaches chemistry. Harvard teaches chemistry. The chemistry departments at both institutions have well-trained faculty, well-equipped labs, careful syllabi, capable students. The chemistry the student learns at Harvard is not orders of magnitude better than the chemistry the student learns at Stony Brook. What the student gets at Harvard that she does not get at Stony Brook is the credential and the network the credential gives her access to. The credential and the network are Case Three goods. The chemistry is a Case Two good. When the institution charges $90,000 a year and tells the student she is paying for the chemistry, the institution is telling her something true about a small part of what she is buying and something misleading about the rest.

The misleading part is the part the series exists to make legible. A reader who can see the difference between Case Two and Case Three can see, when she looks at any institution that calls itself education and asks for money, what the institution is actually selling. Sometimes it is the chemistry. Sometimes it is the network. Sometimes, almost always at the elite tier, it is both, with the second hidden inside the first. The hiding is not always conscious. The institution’s administrators may believe, sincerely, that what they are selling is the chemistry. The mythology has to be taught to the people who run the apparatus before they can teach it to the people who pay for it, and the apparatus has been teaching itself the mythology for almost four hundred years. This is not a story about bad people. It is a story about an institution whose product has changed without the language describing the product changing along with it. The carpenter/house distinction is the language the series proposes for naming the change.

The Sandel reading is sharper still. The mythology does not just confuse two kinds of commerce. It corrupts the moral relationship between the institution and the student. A student who arrives at Harvard believing she has been selected on talent has been told a story about herself that her own subsequent life will be organized around. She will succeed in part because the network she now has access to will route opportunities to her that would not have been routed to her at Stony Brook. The success will reinforce the original story. By the time she is forty she will believe, with some justification on the level of her own observed experience, that the story was true — that her position in the world is the result of her talent and her work. The hubris Sandel diagnoses is not personal. It is the rational conclusion of someone whose biography has been organized inside an apparatus that produces the conclusion as its output. The conclusion is wrong, but the wrongness is invisible from the inside, and the apparatus is engineered to keep it invisible. The mythology is not external to the credential. The mythology is the credential. The credential’s value depends on the student believing it. The institution’s legitimacy depends on the public believing it. And the public believes it because the mythology has been the medium through which Americans have talked about education for at least a century.

The conflation of access-to-instruction and access-to-sorting is not a sloppy use of language. It is the central move on which the entire apparatus depends. Naming the conflation is the structural intervention. The series’ six articles have been naming it from six different angles. This article puts the name in one place where the reader can hold it.

VII. The Steelman, One Last Time

The series has engaged the strongest case for the apparatus at every step. Article 1 took up Lawrence Cremin and Goldin and Katz on the genuinely democratic dimensions of the common school and the high school movement. Article 3 took up Tooze and Mettler on the postwar settlement as a real, if narrow, achievement. Article 4 took up the case for elite institutions as engines of peer effects, resource concentration, and meaningful instruction quality. Article 5 took up Adrian Wooldridge on the historical alternatives to meritocracy and why they were worse. The series has not pretended these cases do not exist. It has tried, at each step, to make them in their strongest form before proceeding.

The synthesis article owes the steelman one more pass, because the carpenter/house distinction will be taken by some readers as a wholesale rejection of the apparatus the steelman defends, and the rejection is not what the series is making. The steelman has to be heard one last time, and answered one last time, for the distinction to land cleanly.

Wooldridge’s argument, in its sharpest form: meritocracy is the least bad sorting mechanism human beings have invented. The historical alternatives — aristocracy of birth, patronage networks, ecclesiastical inheritance, dynastic succession, party loyalty as the sole credential — were worse, by every measure that can be applied across the comparison, for almost everyone who lived under them. The expansion of meritocratic principles since the French Revolution and the nineteenth-century civil service reforms has been one of modernity’s great progressive achievements. The Northcote-Trevelyan reforms in 1850s Britain, replacing patronage with competitive examination for the civil service, produced a more capable and less corrupt government. The American land-grant universities, however vexed their actual founding, produced the engineering and agricultural workforce that built the modern American economy. The Pell Grant generation was a real generation, whose lives were measurably better than they would have been in a system that distributed educational access by birth or wealth alone. To compare actual meritocracy to an idealized non-meritocratic alternative is to compare the existing imperfect arrangement to a fantasy. To compare actual meritocracy to actual aristocracy — to compare 2026 Boston to 1626 Boston — is to find that meritocracy, for all its corruptions, is the system most of the human population would still choose if asked.

This is the Wooldridge argument. It is a serious argument. It is right on the historical comparison. The aristocrat who inherited his position was not, on average, more humble than the meritocrat who earned his; he was simply entitled in a different grammar. The patronage network that distributed jobs by family connection was not more honest than the examination that distributes them by test score; it was simply opaque in a different way. Wooldridge has the receipts on the historical comparison and the series will not pretend he does not. The historical case for meritocracy is real.

The Sandel move — and it is the move the series inherits and now makes one last time — is to grant Wooldridge everything on the historical comparison and then to ask a different question. Sandel does not deny that meritocracy is better than aristocracy. He insists that the question of what to compare meritocracy to is the question of what we are willing to imagine. If the only options on the table are meritocracy and the various premodern alternatives meritocracy replaced, then meritocracy wins by default and the conversation is over. But Sandel’s argument is that the meritocratic frame, even when it works as advertised, is morally corrosive in ways the historical comparison cannot detect. It produces hubris in the winners and humiliation in the losers. It teaches the credentialed class that their position reflects their worth. It teaches the non-credentialed that their position reflects theirs. It hollows out the dignity of work that does not pass through the sorting. It severs democratic deliberation from the lives of the people deliberation is supposed to be about. The corrosion is not in the gap between meritocracy and its ideals. The corrosion is in the ideals themselves — in the assumption that sorting tracks moral desert. Even a perfect sorting mechanism produces the corrosion, because the corrosion is downstream of the assumption, not of the sorting’s imperfection.

That is the deeper Sandel move. The series accepts it. But the series adds one move that Sandel himself does not make in quite this form, and the move is the carpenter/house distinction. The corrosion Sandel diagnoses is moral. The corrosion the series has been documenting is also structural: it depends on the opacity of the sorting in a way that ordinary expert labor does not depend on opacity. The carpenter does not need anyone to forget that the house was built and paid for. The Latin tutor does not need anyone to forget that the lessons were purchased. The admissions consultant needs everyone to forget that the application was engineered. The first two transactions can be conducted in the open, openly priced, openly named, without losing any of what makes them the goods they are. The third transaction cannot. The third transaction’s product is the secrecy of the transaction. This is not a moral observation about the dignity of work or the rhetoric of rising. It is a structural observation about the kind of good the transaction produces. The corruption objection has structural teeth that Wooldridge’s historical defense cannot reach, because Wooldridge’s defense compares the sorting’s outcomes to the alternatives without asking whether the sorting’s outcomes can survive being seen.

The reformed meritocracy Wooldridge proposes — legacy preferences eliminated, athletic recruitment opened to genuine cross-class competition, the Dean’s Interest List shut down, the admissions process subject to external audit — is a meritocracy that has been moved closer to its own ideals. It would be better than the present arrangement. It is worth fighting for. The series does not oppose it. But the reformed meritocracy still rests on the assumption that the sort tracks desert, and the sort, however cleaned up, still requires opacity to function as a credential-distribution mechanism, and the credential, however cleanly distributed, still teaches its winners and its losers a story about themselves that Sandel is right to call corrosive. Reform helps with the fairness objection. It does not reach the corruption objection. The two objections have to be answered separately, and the second one is the harder one, and the second one is the one the series has been arguing for six articles without naming it as such. This article names it.

The series’ position on the steelman, then, is the conjunction. Wooldridge is right about the historical comparison and Sandel is right about the moral diagnosis and the series is right about the structural opacity, and all three things are true at once. Reform is worth fighting for. Reform does not finish the job. The harder work is the work the carpenter/house distinction names — the work of distinguishing the kinds of education that survive being commodified from the kinds that do not, and of building, where one can, more of the first kind and less of the second. The series is not optimistic that the second kind will go away. The institutions whose business model depends on it have too much capital and too much continuity for any reform they themselves agree to to touch the structural feature that produces their value. What the series can do, and what it has tried to do, is to make the structural feature legible, so that any future reformer who proposes to keep the apparatus while merely cleaning up its worst excesses has to confront the question of whether the cleanup reaches the corruption or only the fairness.

VIII. What This Series Is Not Arguing

The most important section of this article is also the simplest. The carpenter/house distinction has been doing the work of this section by implication for six articles. This section names the work in plain English so that the easy misreadings are taken off the table before the closing.

The corrupted kind

A parent pays tuition for an elite university. What is the parent actually buying? If the answer is “exceptional instruction from people who know how to teach,” the transaction is ordinary expert labor for ordinary compensation. Honest commerce. Case Two. If the answer is “admission to a sorting mechanism whose value depends on the number of people excluded,” the transaction is buying access to a credential whose worth is defined by scarcity. Sandel’s argument in What Money Can’t Buy is that the second kind is corrupted the moment it is treated as a market good — because the thing being sold is exclusion itself, and exclusion is not something that can be fairly priced. The corrupted kind is the commerce of legacy admissions, of development cases, of the Varsity Blues side door, of the IEC industry, of the test prep capital that began with Stanley Kaplan in 1938 and reached its current scale in the unregulated $5,000-to-$200,000-per-child consultant industry. It is the commerce of every transaction in which the buyer is paying not for the good itself but for position inside a sorting mechanism that requires opacity to retain its value. It is what The Tyranny of Merit describes from the moral side and what the carpenter/house distinction describes from the structural side. It is what the meritocracy mythology protects, because the mythology is the apparatus by which the corrupted kind is presented to the public as though it were the honest kind. The series has been documenting the corrupted kind for six articles. The series argues that it is corrupted. The series does not argue that it can be reformed into honest commerce by adjusting its prices, expanding its access, or cleaning up its worst excesses, because the corruption is not in the prices or the access or the excesses. The corruption is in the kind of good being sold.

The honest kind

A carpenter builds a house that stands for a century. The carpenter is paid for the labor. The house benefits everyone who lives in it, and a few of the people who walk past it, and some of the children who will be born inside it long after the carpenter is dead. The labor is private. The good is at least partly public. The commerce is ordinary and legitimate, and the carpenter’s wage is the medium through which the craft of carpentry sustains itself across generations. Curriculum design is the same. A well-built course — one that sequences ideas carefully, connects disciplines, grounds abstract frameworks in concrete examples, and teaches students to think rather than to repeat — is the product of specialized labor that deserves compensation. The fact that the good it produces — a person capable of judgment — benefits more than the student who paid for it is not an argument against paying for it. It is the normal condition of how expert labor produces public goods. Doctors get paid. Engineers get paid. Carpenters get paid. Latin tutors get paid. Curriculum designers get paid. None of this is in tension with the good being a public one. The honest kind is the kind of commerce in which the buyer is paying for the labor itself, the instruction itself, the actual thing the labor produces, and in which the value of what was bought does not depend on the buyer not knowing it was bought. The Latin the child learned is the Latin she learned. The chemistry the student learned at Stony Brook is the chemistry she learned. The competence is real. The labor is real. The wage is the medium by which the practice continues. There is no opacity required, and no shame in being paid for the work. The series does not oppose this kind of commerce. The series depends on this kind of commerce. The author of this series depends on it personally and the curriculum from which the series is published depends on it institutionally. The honest commerce is the part the series is arguing for.

The confusion

The mythology of American education treats both kinds of commerce as the same thing and calls them “access.” Access to what? Access to sorting, or access to learning? The question almost never gets asked, because asking it would expose what the payment is actually buying in each case. The prep school is selling access to the sorting mechanism its graduates will enter. The textbook is selling access to the labor that produced the writing. The elite university is selling both at once and presenting them as a single product. The community college is selling almost entirely the second kind. The IEC industry is selling almost entirely the first. The test prep company is selling almost entirely the first. The Latin tutor is selling almost entirely the second. These are not the same transactions. They are not the same goods. Conflating them is the move the mythology requires in order to sound fair, and conflating them is what the apparatus has been training Americans to do for a century. The confusion is the structural feature on which the apparatus depends. Naming the confusion is the structural intervention. The series’ six articles have been naming the confusion from six different angles. This section names it once more, in the simplest possible language, so that any reader who has gotten this far can apply the distinction herself, to her own past purchases, to her own current decisions, to her own children’s future purchases, to any institution that asks her for money in exchange for what it calls education. The question is always the same. What am I actually buying? If the answer is the labor and the learning, the commerce is honest. If the answer is the position in the sort, the commerce is corrupted. If the answer is both, then the buyer has to decide what proportion of the price is covering which, and the institution that will not tell her is the institution whose product is the opacity.

Where this leaves us

No policy prescription. A distinction. The series does not propose a federal law, a state law, a municipal ordinance, an admissions reform, a tax incentive, a regulatory regime, or a constitutional amendment. None of these instruments reach the corruption the series has been documenting, because the corruption is in the sphere of commerce that has invaded the sphere of education, and policy instruments work inside the sphere of commerce by definition. They cannot restore a sphere they were built to operate within. Nascent ideas for distinguishing the two kinds of commerce in practice are presently under implementation — by the people writing this series, by others building curricula with the same stance, and by the handful of institutions that have begun experimenting with honest compensation for honest labor without the credential-sorting overhead. Whether any of it scales is not yet clear. What is clear is that the mythology has to go first. The distinction has to be named before the reform can be designed. The distinction the series proposes is the carpenter/house distinction, and the distinction is offered to the reader not as a program but as a tool for reading any future proposal that calls itself reform. Any proposal that confuses honest commerce with corrupted commerce is going to fail, regardless of how much money it spends, because the confusion is the original feature the proposal was supposed to fix. Any proposal that distinguishes them is at least starting in the right place, regardless of whether it succeeds. The series is offering the distinction. The series is not offering the proposal. The proposals will come, or they will not, from people who are doing the work of building the honest kind of commerce in places where the corrupted kind has too much money and too much continuity to be displaced. The series’ job is to make the distinction clear enough that those people, when they appear, will be recognizable.

IX. The Conjunction

One sentence remains. The series has been carrying it for six articles and the closing of this seventh article is the place to put it down where the reader can hold it.

The conjunction Sandel insists on is and.

Some goods are corrupted by markets and some goods are fairly sold by them. Education is both kinds of good at once. The task is telling them apart.

This is not a slogan. It is the philosophical move that holds the entire series together. The corruption objection is real and the honest commerce is real. The mythology is morally corrosive and the instruction it surrounds is sometimes excellent. The 1922 Lowell apparatus was an act of exclusion and the chemistry department two buildings over was teaching real chemistry to real students. The Pell Grant generation was a partial achievement and the IEC industry was already being built to absorb its successors. Wooldridge is right about the historical comparison and Sandel is right about the moral diagnosis and the series is right about the structural opacity. The carpenter is paid and the house gets built and the labor and the wage and the shelter all coexist in the same transaction without any of them being corrupted by the others.

The reason the conjunction matters is that American discourse runs almost entirely on the disjunction. Either the system is good or it is bad. Either the credentials are earned or they are bought. Either commerce is honest or it is corrupt. Either we defend the apparatus or we tear it down. The disjunction is what makes the conversation possible, and the conversation possible by the disjunction is the conversation that has produced forty years of policy stalemate and an apparatus that has gotten worse on every dimension the conversation tracks. The disjunction is the wrong frame. The right frame is the conjunction, and the conjunction is harder to hold because it asks the listener to accept that two things are true at once, that the truth of one does not cancel the truth of the other, that the work of the next move is the work of telling them apart in each particular case rather than choosing one and ignoring the other.

The and is what the series is asking the reader to carry away. Not a verdict on American education. A way of looking at any institution that calls itself education and asks for money. The carpenter is being paid for the labor and the carpenter’s product is real. The Latin tutor is being paid for the lessons and the lessons produce real knowledge. The admissions consultant is being paid for the engineering and the engineering produces a credential whose value depends on the buyer not knowing it was engineered. Three transactions. Three different relationships between the money and the good. The first two scale without harming the good. The third devours itself when it scales. The reader who can see the difference can apply it to anything.

The series will end here. There is no policy bullet list. There is no call to action. There is no resolution, because the situation the series has been describing does not have a clean resolution available to anyone reading these words. There is the distinction. The distinction is what the series exists to draw. The reader has the distinction now, and what she does with it is up to her, and what she does with it is the only part of this conversation that the series cannot do for her.

The architects wrote it down. The pipeline runs from 1872 to 1974 to 2024 without touching the sides. The abundance was real and brief. The reproduction has been continuous since 1636. The mythology is morally corrosive. The class that won protects the game. And underneath all of it, in the transactions that build the apparatus and the transactions that sustain the alternative to the apparatus, is the difference between the two kinds of commerce, the one that pays for the good itself and the one that pays for position in a sort that requires its purchase to be invisible.

The conjunction Sandel insists on is and. Some goods are corrupted by markets and some goods are fairly sold by them. Education is both kinds of good at once. The task is telling them apart.

That is the end of the series. The work of telling them apart is the work that begins after the series ends. The receipts are below.