Education: Schooling, Admissions, and Advantage
The Million Dollar Question: What share of Harvard’s white undergraduate admits are athletes, legacies, on the Dean’s Interest List, or children of faculty and staff?
A) 12% B) 23% C) 43% D) 61%
Read on for the answer.
A working explainer of how wealth actually moves through schooling — pre-K through college — and what the dollars buy at each gate. The short version is queue management, sustained for fifteen years. The longer version is what this article is about.
What it is
The wealthy-family education stack is a sequence of roughly ten schooling decisions, taken between the time a child is three and the time the child is eighteen, each one a small probability bet on the next. Preschool. Kindergarten admit. Lower-school continuation. Middle-school transition, often to a feeder. High school — day or boarding. Summer programs and enrichment. Test preparation. Outside college counselor. The college admit itself. And, for the most engineered version of the stack, the postgraduate continuation. Most readers conceive of this as a series of individual decisions. Inside a wealthy household it functions as a single fifteen-year project with a single deliverable — the college admit at eighteen, and the network behind it.
It is useful to separate four families of spending. Tuition is the published number: the school’s listed annual price. Preparation is the layer of tutoring, enrichment, summer programs, and music or athletic intensives that overlays the school day. Positioning is the part most readers misunderstand — outside admissions consultants, recruited-athletic pipelines, the maintenance of legacy and faculty relationships, and the long management of the college-counselor relationship. Donation is the smallest line by participant count but the largest by dollar — the development-office relationship, where seven-figure pledged giving sits on the institutional side of the wall from the admissions office, but where the wall is, as a matter of public record, thinner than the institutions like to say.
None of this is primarily a project to purchase intelligence. The dominant value is sustained exposure to small probability advantages at many independent gates, compounded over fifteen years. Each gate’s admit makes the next gate easier. That is the asset the household is buying. It is real, it is expensive, and the most-misunderstood part of it is that the highest-spending households are not paying for certainty — they are paying for a slightly improved base rate at every step.
The site uses the wealth-bracket shorthand from Wealth Levels — $1M–$5M, $5M–$30M, $30M–$100M, $100M+, $1B+. The dense-cost version of the education stack starts to make economic sense at the upper end of the $1M–$5M band and gets fully built out at $30M–$100M, where the household has the cash flow and the network capital to play the legitimate version of the long game.
Who uses it
Schooling spend scales with wealth band, but not linearly. The shape of what the household buys changes at each bracket more than the absolute size of the bill does.
$1M–$5M band. A coastal-metro household stretches for one or two private placements, frequently funded in part by grandparents through annual gifts at the 2026 gift-tax exclusion of $19,000 per donor per recipient. Test prep is standard; outside consulting is a $4K–$10K national-average engagement. Boarding school is on the menu but not the default.
$5M–$30M band. The full private-school-plus-tutoring-plus-consultant trajectory. Two children in NYC private elementary at Trinity, Dalton, or Spence. Summer programs that lean toward selective. A college consultant in the $25K–$45K Crimson range or the $50K–$100K boutique range. This is the modal wealthy-family stack and where most of the headline tuition numbers describe.
$30M–$100M band. Boarding school becomes common — Andover, Exeter, Lawrenceville, Choate, Hotchkiss, Deerfield, St. Paul’s, Groton, Milton, Middlesex. The college-counselor engagement moves toward the multi-year, $100K–$250K boutique tier. Recruited-athletic pipelines — squash, rowing, fencing, sailing, golf, lacrosse — become viable investments because the school’s coaching infrastructure can produce a recruit by junior year of high school. A modest donor relationship begins.
$100M+ band. Active engineering. Children with recruited-athlete pipelines from middle school. Legacy maintenance — board service, regular giving, named-room or named-building gifts at the prep school. Development-office tracking at the college level. A senior consultant in the $250K–$750K range, often working since the child was in seventh grade.
$1B+ band. A subset of the above, plus the part this article cannot fully chronicle: institutional relationships that long predate the child’s application. The named building. The endowed chair. The pledged giving. These work, but they work slowly and are visible from the inside — the 2018 SFFA litigation against Harvard surfaced internal admissions emails in which the dean of admissions tracked exceptionally wealthy applicants on what the litigation called the Dean’s Interest List. That mechanism, more than any other, is what the very top of the wealth distribution is purchasing.
Geographically, the dense-cost version concentrates in NYC, the Bay Area, Boston, Washington, Los Angeles, and a handful of smaller pockets in Greenwich, Greater Chicago, and Northern Virginia. Lighter versions of the stack — without the $70,000 tuition line — exist everywhere there are private schools. The stack is a coastal-metro phenomenon principally because the day-school infrastructure exists at that price point principally in those cities.
Why they use it
The honest list of motives, in roughly the order they bind.
Probability compounding. Each gate’s admit improves the base rate at the next gate. A child admitted to a feeder kindergarten in Manhattan does not become smarter, but does become statistically more likely — by a small amount, at each step — to be admitted to a feeder middle school, then to a top-tier high school, then to a college that selects partly on the basis of where the high school transcript came from. Multiplied across ten gates and fifteen years, the cumulative-probability improvement is the asset the household is purchasing. Any single school decision is a small bet. The sustained pattern is what produces the outcome.
Network formation. Day-school and boarding-school cohorts produce peer groups whose own children will be the next generation of applicants. That is a long-horizon asset that is partly mercenary and partly genuine. The wealthy household that pays Trinity $69,000 a year for thirteen years is also paying for thirteen years inside a parent community that is, on average, the largest social network capital its child will form before college.
Risk reduction. The stack is a variance-reduction trade, not a mean-shift trade. A child with academic credentials at the 60th percentile of a Trinity or Andover cohort still emerges into a compressed range of college and post-college outcomes. The household is buying down the worst tails of its child’s trajectory, not maximizing the best ones.
Status, plainly. Acknowledged but secondary in the manual’s voice. Trinity is a recognizable name; Andover is a recognizable name; the credential signals to other parents and to the household’s peer group. The signal is real and it is part of what the dollars buy, but it does not survive economic scrutiny on its own — most of the actual return on the stack comes from the first three motives.
What the stack does not buy is the child’s own engagement. Several of the researchers cited later in this piece — Raj Chetty, David Deming — make this point directly: high-credential post-graduate cohorts whose engagement was extrinsically driven through the schooling years tend to underperform their credentials by their thirties. The schooling stack is necessary if the household is targeting Ivy-Plus outcomes. It is not sufficient.
How it works
A chronological walk through the stack, with the dollar lines and the institutional mechanics at each gate.
Pre-K and kindergarten. The Manhattan and Bay Area application gauntlets begin when the child is three. The household submits applications to between six and twelve schools, each with its own ERB or developmental-assessment requirement, a parent interview, and a child visit. The preschool director’s phone call to the kindergarten admissions office matters considerably; the strongest single predictor of a kindergarten admit at a top NYC school is the preschool the child currently attends. Per Education Next’s long-run enrollment analysis, the share of middle-income students at private schools has fallen by roughly half since 1968, while the share of high-income students has held steady — private schooling is, increasingly, a wealthy-family institution.
Lower and middle school. Continuation is the default for most children admitted to a K-12 day school, but it is not automatic. Teachers’ assessments matter. Departmentalized tracking begins in fifth or sixth grade in many schools, and a child not assigned to the higher math or English track in middle school will have meaningfully fewer options at the next gate. Outside tutoring during this period is principally remedial or accelerative, not test-driven.
High school. The household chooses day school (Manhattan, the Bay Area) or boarding (Andover, Exeter, Lawrenceville, Choate, Hotchkiss, Deerfield, St. Paul’s, Groton, Milton, Middlesex). Boarding has its own ecology — three-year cohorts, college-counseling departments that work the application machinery for a living, recruited-athletic pipelines that operate at scale — and is properly its own subject. The short version is that for a $30M–$100M household with an academically-engaged child, a strong boarding school is the most efficient single point of leverage in the stack.
Test preparation. SAT and ACT tutoring runs $100–$200 per hour at the normal-pricing tier and $500 to $1,000+ per hour for the highest-billing specialists per Tutors.com’s 2026 SAT tutor cost data. Most students do 20 to 30 hours of private tutoring across the preparation arc, with total spend running roughly $1,600 at the low end and $10,000+ at the higher end. The wealthy-family version typically lands at $5,000–$30,000 per child across two years, per Private Prep’s 2026 pricing data. At the very top, boutique tutors in NYC and the Bay Area charge above $1,000 per hour and bill multi-year retainers.
Summer programs and enrichment. The credentialing layer outside the school day. Selective summer programs — Telluride, MITES, RSI, Iowa Young Writers’ Studio — function partly as low-cost college-application credentials. Music intensives at Tanglewood or Aspen, debate camps, math camps. The household spend during the active years (roughly grades 6 through 12) lands in the $10,000–$40,000 per child per year range, before travel.
The college counselor. The household engages an outside college counselor in addition to the in-school college office. The pricing taxonomy is wider than most readers expect. Per Tutors.com 2026 pricing data, the national average for full-service college counseling lands $4,000–$10,000. Crimson Education’s typical two-year package is $25,000 to $45,000; per Fortune’s 2024 profile of Crimson, working directly with co-founder Jamie Beaton runs upwards of $200,000. Boutique multi-year programs from firms like Top Tier Admissions, Command Education, and Lakhani Coaching range from $50,000 to $250,000. At the very top, Christopher Rim’s Command Education and a small number of comparable shops charge as much as $750,000 for application packages that start in seventh grade. Per CNBC’s reporting on Lakhani Coaching, average client spend is around $58,000 but the firm has billed $800,000 across multi-year engagements.
Recruited athletics. The most underrated lever in the stack. Per the Arcidiacono, Kinsler, and Ransom NBER working paper based on Harvard’s court-disclosed admissions data, Harvard admits 86% of recruited athletes — the highest single-category admit rate in the data. The sports where private-school day cohorts can produce recruits at competitive scale are the ones in which most Americans cannot — squash, rowing, fencing, sailing, golf, lacrosse, and to a lesser extent water polo, field hockey, and skiing. The economic return on five years of recruited-athletic pipeline funding, for a child whose academics already put them in the conversation, is one of the cleanest single investments in the legitimate stack.
Donor and legacy paths. The smallest by participant count, the largest by dollar. Per the Harvard Crimson’s 2018 coverage of internal admissions emails surfaced by the SFFA litigation, the Dean’s Interest List tracks applicants with significant family giving or pledged giving, and admissions decisions for these applicants are made with explicit input from the development office. Per the 2019 Arcidiacono paper, the Dean’s Interest List admit rate was 42% — roughly nine times the overall rate. Legacy admit rate was 33%. Faculty and staff children, 47%. The Harvard Crimson noted in a 2023 follow-up that the SFFA decision did not change the legacy mechanism.
What it costs
A worked example for a Manhattan family with two children, ages 4 and 7, on the dense-cost path. All numbers in USD, 2025–2026 base year, mid-single-digit annual inflation assumed for forward extrapolation.
K-12 tuition. Trinity charges $69,000 for 2025–2026, Dalton $67,480, Spence $68,480. Avenues, Brearley, and Collegiate cluster in the same range. Across thirteen years per child with mid-single-digit annual escalators, the household commits roughly $1.7M–$2.0M per child in nominal pre-tax tuition, or $3.4M–$4.0M for two children — before fees, supplemental travel, and incidental capital contributions to the schools themselves.
Enrichment, tutoring, and summer programs. $10,000–$40,000 per child per year during the active years (roughly grades 6 through 12), or $70,000–$280,000 per child across the active period. For two children, $140,000–$560,000.
Test preparation. SAT/ACT preparation at the higher-end tier is $3,000–$30,000 per child per Private Prep’s 2026 pricing; the boutique-tutor end is meaningfully higher. Call it $10,000–$50,000 per child at the wealthy-family standard. For two, $20,000–$100,000.
College admissions consultant. $25,000–$45,000 for a Crimson Education two-year package; $50,000–$200,000 for a typical boutique multi-year engagement; $500,000–$750,000+ at the very top. For two children, the modal $30M–$100M household spend lands in the $200,000–$400,000 range across both children’s application arcs.
Undergraduate college sticker. Harvard’s published 2025–2026 cost of attendance is $86,926. Stanford, Yale, Princeton, Columbia, Penn, and MIT cluster in the same band. Across four years, roughly $370,000 per child, or $740,000 for two — sticker, before any financial aid. Note that Harvard’s financial aid is now free tuition for families earning under $200,000, so the sticker is paid principally by the wealthy households this article describes.
Boarding-school alternative. Phillips Andover’s 2025–2026 boarding fee is $76,800; Phillips Exeter is in the same range. Lawrenceville, Choate, Hotchkiss, Deerfield, St. Paul’s cluster $70K–$80K per boarding year. International comparison: Eton College’s 2025–2026 boarding fee is about $80,000 per year at recent exchange rates. A four-year boarding stint runs $300,000–$320,000 per child, plus the lost partial-year of day-school tuition in transition.
Stack subtotal. A fully-loaded two-child Manhattan trajectory through one of the Ivy-Plus colleges, the most expensive version of the stack, lands in the $4.5M–$7M range in pre-tax outlay across twenty-two years. Pre-tax matters: at top combined federal/state/city marginal rates, the household must earn closer to $8M–$12M of pre-tax income to fund that spend after taxes. That is the dollar shape of the stack.
A second-frame comparison, three patterns at the same household income:
| Component | NYC, two children, day-school stack | $30M+, two children, boarding stack | Sun-belt metro, two children, public + consultants |
|---|---|---|---|
| K-12 tuition | $3.4M–$4.0M | $2.4M–$3.0M (day K-8 + 4 years boarding) | $0 |
| Enrichment + tutoring | $140K–$560K | $140K–$560K | $50K–$200K |
| Test prep | $20K–$100K | $20K–$100K | $20K–$60K |
| College consultant | $200K–$400K | $400K–$1.5M (top-tier boutique) | $25K–$90K |
| Undergrad sticker (2x) | $740K | $740K | $200K (in-state flagship × 2) |
| Stack total | $4.5M–$5.8M | $3.7M–$5.9M | $295K–$550K |
The bottom row is the point. The dense-cost version is principally a Manhattan / Bay Area / Boston / Washington phenomenon, and is principally a $5M-and-up wealth-band phenomenon. A genuinely wealthy family in Houston, Charlotte, or Nashville frequently elects the public-school path, supplements with strong tutoring and outside consulting, and arrives at college admissions outcomes meaningfully similar to a coastal household that spent ten times the money. The geographic gradient is real, and one of the under-discussed parts of the schooling-spend conversation.
Hidden costs and tradeoffs
What the cost-stack table does not surface.
The one-way-door problem. Private kindergarten in Manhattan or the Bay Area is functionally an irreversible decision. Switching back to public schooling at any point after first grade is socially and academically expensive even when it would be financially correct. The household has signed up for a $130,000-to-$140,000-per-year line item across two children for thirteen years — roughly $1.8M–$2.0M, before escalators, before college. The same family-level cost-stack logic that turns a HENRY household paycheck-to-paycheck is at work here, at a different income tier.
The geographic prison. Once the kindergarten admit is secured, the family is functionally tethered to the metro. A relocation costs the kindergarten admit at the new metro and frequently forfeits the social investment the household has been making since the child was three. Families do relocate; the cost of doing so is much higher than is generally acknowledged in the relocation conversation itself.
The intensification arms race. Each generation of preparation produces a harder applicant pool for the next. Harvard’s overall admit rate has compressed from roughly 6% in the early 2010s to about 3.6% for the Class of 2028, per the Harvard Crimson’s admissions coverage. The same household dollar spend produces lower probability admits at the top schools over time. The arms-race math is structural — the supply of seats is fixed, and the demand has been growing — and the policy implications for the household are uncomfortable. Spending more does not necessarily lead to a better answer; running the stack faster than peers does.
The parental-time tax. The stack consumes meaningful weekly hours of parent labor — application tours, supplemental essay drafts, recruited-athlete travel, fundraising-committee work, parent-association obligations, donor-event attendance. The labor is uncosted but real, and a significant share of why household staff appear in the wealthy-family budget at all. The labor of running the educational machinery is one of the things wealth genuinely buys help for.
The status-network paradox. Children whose schooling is engineered for credential and access often emerge with measurably weaker independent decision-making than children whose education was less managed. The pattern is documented in longitudinal studies of high-credential cohorts and is the most-discussed unresolved problem inside the modern boarding-school community itself. The schools know about it; they talk about it; they have not solved it.
Donation as fragile asset. Pledged giving for development-office consideration is a relationship asset, not a contract. Pledges have been honored after admit decisions went the household’s way; pledges have been quietly revoked after they went the wrong way; relationships have been broken in both directions. The seven-figure pledge is not a transaction with a guaranteed return, and the institutions are careful to maintain it that way precisely so they can decline applicants whose academic profile is genuinely below the threshold without legal exposure. The very top of the donor path is a probability bet, not a transactional one.
What people get wrong
Five corrections, each one earning the section’s promise.
The system is rarely about purchasing intelligence. Most of the dollar spend buys queue position, not IQ. Tutoring delivers real but bounded improvements; the dominant value of the wealthy-family stack is sustained exposure to small probability advantages, compounded over fifteen years. The household paying $5M for the schooling arc is not buying a smarter child than the household paying $0.5M; it is buying a more compressed range of outcomes — fewer bad tails, a higher base rate at every gate.
The Varsity Blues scandal is the wrong mental model. Operation Varsity Blues, the 2019 case against Rick Singer, revealed the failure mode at the bottom of the legitimate system — families with enough money to want the prize but not the network capital to access it through legitimate channels. Singer was sentenced in 2023 to three and a half years in prison and over $10 million in restitution. The real side door that most wealthy families use is the recruited-athlete and donor pathway, which is entirely legal, fully disclosed in court filings, and substantially larger in volume than anything the Varsity Blues defendants paid for. Treating all wealthy-family admit advantages as scandalous obscures the more important point — the legitimate machinery is the substantial machinery, and it is doing most of the work in plain sight.
The Ivy-League income advantage is empirically large. Per Chetty, Deming, and Friedman’s 2023 Diversifying Society’s Leaders?, applicants from the top 1% of the household income distribution are roughly twice as likely to be admitted to Ivy-Plus colleges as middle-class applicants with comparable SAT and ACT scores; applicants from the top 0.1% are more than twice as likely again. Three mechanisms account for nearly all of the gap: legacy preference, recruited athletics, and non-academic ratings that disproportionately reward characteristics correlated with wealth. The authors note in the paper that these three mechanisms are uncorrelated or negatively correlated with post-college success — they raise the admit probability without raising the predicted outcome. Pretending the advantage is small or anecdotal is empirically wrong. The honest version is that the advantage is real and the data is now public.
At the top, the consulting market is mostly Veblen. The $750,000 admissions-consultant package is real, but it is also priced for visibility. The marginal probability gain over a $25,000–$45,000 Crimson package is small once the child is already on the boarding-school + recruited-athlete + dean’s-list-relevant track. Households paying the top tier are buying the consultant’s time, pattern-matching, and continuity over five years, not access to mechanisms unavailable at lower tiers. Confusing scarcity-pricing with probability-buying is one of the most common errors in this market, and it is the error most directly responsible for the recurring “is it worth it?” articles in the financial press. The honest answer is that for most wealthy families with engaged children, the marginal dollar spent above roughly $100,000 in college-counseling fees is largely a peace-of-mind purchase, not a probability one.
Money cannot keep a child engaged. The single largest unforced error in the wealthy-family education stack is treating the credential as the end state. Children who arrive at college via the fifteen-year stack but have not internalized independent motivation along the way tend to underperform their credentials by their thirties. The pattern shows up in longitudinal post-college cohort studies and in the consistent (and quiet) anxiety inside the boarding-school faculty community itself. The schooling stack is necessary to produce the credential at scale. It is not sufficient to produce a person who uses the credential well.
Bottom line
Returning to the opening question: per the Arcidiacono, Kinsler, and Ransom NBER paper based on Harvard’s court-disclosed admissions data, about 43% of Harvard’s white undergraduate admits fall into the four ALDC categories — athletes, legacies, Dean’s Interest List, or children of faculty and staff. The comparable share for African-American, Hispanic, and Asian-American admits is under 16%. Athletes are admitted at 86%, legacies at 33%, Dean’s Interest List at 42%, faculty children at 47% — against a class admit rate around 6%. Per Opportunity Insights’ 2023 paper, applicants from the top 1% are 34% more likely to be admitted to Ivy-Plus colleges than middle-class applicants with comparable test scores. These are the central facts of how the wealthy-family education stack actually delivers its outcomes.
A coastal household on the full path commits roughly $4.5M–$7M in pre-tax outlay across twenty-two years to run the stack — $8M–$12M in pre-tax income, after taxes are added. The dollar buys a higher base rate at every gate, compounded for fifteen years, with no guarantee at the end. It buys credential, network, and variance reduction. It does not buy the child’s own engagement, and the households whose children eventually use the credential well are the ones in which the engagement was already going to be there.
The stack is a real asset, honestly described. Its honest description is that it is queue management, not certainty purchase. The households that run it well treat it that way and finish their child’s eighteen-year arc with a credential and a peer group. The households that confuse the credential for the outcome finish with a credential and a child still searching for what to do with it.
Related reading: Wealth Levels: Life at $1M, $10M, $100M, and $1B · HENRY: $500K and Still Paycheck-to-Paycheck · Private Clubs: Membership, Status, and Access · Paths to Millions: How First-Generation Wealth Is Actually Built · ZIP Codes: Where the Wealthy Live
