Picture two families, each with a daughter. Same scores, same grades, same plan to study economics. Each daughter gets two letters: one from one of the most famous private colleges in the country, and one from her strong state university.

One family earns $90,000 a year. The other earns $400,000.

They should not make the same choice. And neither right answer is the one most families assume. Most families treat the famous school as the expensive dream and the state school as the sensible fallback. They expect the dream to pay more, because its graduates do.

For a top student, the pay half of that picture is wrong. Below new aid lines at most of the famous colleges, the price half is wrong too.

This piece is for families with a strong student who is building a college list this fall. It compares only the things that change when that student picks one door over the other: what the family pays, and what the network adds. It follows What a Harder Door Is Worth, which tested what the acceptance rate tells a family. This one is about the student who can open either door.

The student carries the pay

Start with the number that misleads families most. Among twelve of the most selective private colleges, the ones researchers call the Ivy-Plus (the eight Ivy League schools, plus Stanford, MIT, Duke, and the University of Chicago), the typical student who enrolled with federal aid earns $102,154 a year ten years after starting college. At public universities whose students average 1350 or more on the SAT, the typical figure is $73,641.[1]

That looks like the famous school adds about $28,500 a year. Most of that gap isn’t the school. Those are different students. The typical Ivy-Plus school’s students average 1553 on the SAT. At the public universities, the typical average is 1407.[1] A comparison of school averages credits the school with what its students would have earned anyway.

The research that holds the student fixed tells a different story.

• Stacy Dale and Alan Krueger, economists who compared students accepted by similar colleges who then chose differently, found that students who attended more selective colleges “earned about the same as students of seemingly comparable ability who attended less selective schools.” They also found that “children from low-income families, however, earned more if they attended selective colleges.”[2]
• Jack Mountjoy and Brent Hickman, working with state records on Texas public universities, compared students who applied to and got into the same sets of schools. They found “only a fleeting selectivity earnings premium fading to zero after a few years in the labor market.” They found little sign of a bigger payoff for high scorers. The return to selectivity looked about the same for students with high and low test scores in high school.[3]
• Raj Chetty, David Deming, and John Friedman studied students near the admissions line at the Ivy-Plus colleges, comparing those admitted off the waitlist with similar students who were not. Attending one of these colleges instead of a flagship public university left the chance of reaching the top quarter of earners unchanged, at about 73 percent. In their words, “Admission to an Ivy-Plus college has no significant impact on predicted mean income rank.”[4]

Think of a gifted player traded between two teams. She scores about the same on either roster. What the famous team adds is more scouts in the stands. I described that difference in Yes, the Ivies Produce the CEOs, and it is the right way to think about a top student’s choice. Her typical paycheck travels with her.

So I held the student’s pay fixed in my College ROI model and changed only the price. My model measures lifetime value: how far ahead of the high school path a graduate ends up, after taxes, after the cost of college and four years of wages given up, with later dollars counted for less. When the pay is the same at both schools, it cancels out of the comparison, and what’s left is the difference in price over four years, in today’s dollars.[5] A family does not have to guess what their child will earn. The difference between two letters is the price, plus whatever the network adds.

Price is usually the weakest lever in college. Even at zero cost, about 30 percent of bachelor’s institutions leave their median graduate behind the high school path.[6]

A top student is the exception. The other two levers, finishing and the field, work in her favor either way. Top students are very likely to finish: the typical Ivy-Plus school graduates 96 percent of its students within six years, and the typical public university in this group graduates 86.4 percent.[1] And they can choose the same field at either school. With those set, the school changes two things: the bill and the network.

Prestige for free

Here is what has changed. Most of the twelve Ivy-Plus colleges now cover full tuition for families earning far more than most families assume.[7]

• Harvard, Yale, MIT, and Penn cover at least full tuition for families earning up to about $200,000.
• Princeton and the University of Chicago go to $250,000. Chicago’s line starts with the students who enter in the fall of 2027, the class applying now.
• Dartmouth’s line is $175,000, Stanford’s and Columbia’s $150,000, and Brown’s and Cornell’s $125,000. Duke’s guarantee covers only students from North and South Carolina.
• Below a lower line, between $60,000 and $150,000 depending on the school, parents are expected to pay nothing at all.

Most of these lines assume typical assets or a similar test, and each school defines that differently. MIT notes that its $200,000 line is “about the 80th percentile income for American families.”[8]

The federal data, collected before several of these expansions, already showed the pattern. Families earning $48,000 to $75,000 paid a typical $4,459 a year at an Ivy-Plus college, against $11,875 at the strong public universities. Families earning $75,000 to $110,000 paid $12,170 against $18,607.[1] For the same student, the famous school comes out $23,000 to $27,000 ahead in lifetime value on price alone.[5]

Covering tuition is not covering everything. Near the top of a line, a family can still owe for housing, food, and fees. At Harvard this year those come to about $29,000,[7] close to what a strong public university costs in full, in state.[1] So near the line, the famous school costs about the same, not less. Further below the line, it often costs less. MIT estimates that families earning $100,000 to $200,000 who received its aid paid a median of $19,734 last year.[8]

For a family below the line, the question nearly answers itself. The outcome for the typical student is about the same either way. One school gives you the name and its network at no extra cost, often for less. The other gives you no name for about the same money, or more. Why give up prestige for free to take no prestige for free?

And the network is real. In the Chetty, Deming, and Friedman waitlist study, attending an Ivy-Plus college instead of a flagship raised a student’s chance of working at a prestigious firm at age 25 from about 7 percent to about 25 percent. It nearly doubled the chance of attending an elite graduate school, from about 6 percent to about 12 percent.[4] The authors also found the effects “much larger for students with weaker fallback options,” such as students whose home state’s colleges send fewer graduates to the very top.[4] Dale and Krueger’s one exception, students from low-income families, points the same way.

I know what that door looked like from the other side. When I chose a college, cost made Baruch the only door I could walk through. I went on a full scholarship, worked full time during the day, and took a full load at night.[9] A student from a family like mine, admitted to most of these twelve colleges today, would pay little or nothing to attend.

This is the gift of price I described in What the Prestige Study Left Out, now reaching much further up the income ladder. The gift is what makes the difference. The name comes with it, at no extra charge. For a family below the line, the famous school is not the luxury. Run its calculator before you cross it off.

Above the line, full price buys the network, not the typical paycheck

Now the family earning $400,000.

They will likely pay the full cost of attendance, which for this school year runs about $93,000 to $104,000 a year at the eleven of these colleges that publish a total.[7] At their strong state university, they would pay a fraction of that. The 28 public universities in my comparison sit in 20 states.[1] If your state’s leading public university is not among them, the price side still holds. In most other states the in-state price is lower still. The network side may weigh more, since the waitlist study found larger effects for students with weaker options at home.[4]

Hold the student’s pay fixed again. At full price, the strong public university comes out $199,243 ahead in lifetime value. That is on the federal cost figures, which run below today’s prices, so today’s gap is likely wider.[5] Because the pay cancels, the $199,243 is the same at any pay level. At the Ivy-Plus graduates’ typical pay four years out, full price still leaves the student $161,255 ahead of the high school path. The public university leaves the same student $360,498 ahead.[5]

Compare your own colleges side by side, at your own price, in the Guided ROI Interview on collegeroi.org.

What does that money buy? Not the typical paycheck, which did not move in the waitlist study. What moved were the long shots and the doors.[4] The table follows students who attended an Ivy-Plus college. The second column is how they did. The first is the study’s estimate of how the same students would have done at a flagship, not how flagship students did.

For the same student	At a flagship	At an Ivy-Plus college
Chance of reaching the top quarter of earners	73.3%	73.3%
Chance of reaching the top 1% of earners	11.8%	16.8%
Chance of attending an elite graduate school at 28	6.1%	11.8%
Chance of working at a prestigious firm at 25	7.2%	24.7%

There is a fair objection here, and I want to meet it directly. The same study reports that Ivy-Plus students earn about $101,000 more a year at age 33. If that were the typical gain, full price would be a bargain. But that figure does not come from the waitlist comparison. It compares Ivy-Plus students with flagship students reweighted to look like them on test scores, family income, race, and gender.

The authors read it as the effect of attending, and it is where they part ways with Dale and Krueger and with Mountjoy and Hickman. They also say plainly that with the waitlist design they “lack adequate precision to obtain an informative estimate” of the effect on mean income in dollars.[4] And they trace the gain to the very top of the income distribution. A family cannot spread one child across many tries. Why that makes the Ivy-Plus seat a long-shot ticket is the subject of Yes, the Ivies Produce the CEOs.

The Ivy-Plus pay gap also differs widely by field. Compare programs in the same major, and the Ivy-Plus graduates’ pay four years out pulls far ahead in some fields and moves far less in others:[10]

• Computer science: $208,824 at the typical Ivy-Plus program, $126,183 at the typical public one
• Economics: $142,936 against $83,880
• Biology: $68,435 against $62,659
• English: $62,215 against $53,860
• Fine arts: $52,860 against $48,684

These are raw comparisons, and the students differ too. They are also fourth-year paychecks, not starting offers. They count everything on the W-2, including bonuses and stock paid as wages, in 2022 and 2023.[10] Some of these graduates may also hold a master’s by then. From this table I can’t tell how much of the computer science gap is the student and how much is the campus. So for a student set on computer science or economics, I am less sure the typical paycheck travels than for one headed to biology or English.

What the table does show is where the network has room to matter. In What a Harder Door Is Worth, I read the same pattern: scores track pay most where employers recruit from a short list of campuses, and least where a license or a pay scale sets the wage. The scouts sit at some games and not others.

So for the full-pay family, the question is what the network is worth to your child’s plans, in that field. For a family that wants those odds and can absorb the cost, it is a legitimate purchase. For a family that would borrow to buy it, it is a mistake. A Parent Did the $400,000 Math in Public shows what that money grows into if you invest it instead.

The next tier is where full price is hardest to justify

The bigger risk sits one step down.

There are 52 other private colleges whose students average 1450 or more on the SAT. They are hard to get into, and many families treat them as nearly the same as the Ivy-Plus.[1]

At full price, they are: $85,913 a year on the federal figures, against $87,813 at the Ivy-Plus. Among aided families earning more than $110,000, the typical net price was $48,622, against $50,307. For families earning $48,000 to $75,000 they were less generous: $13,112 a year, against $4,459 at the Ivy-Plus.[1]

On the long shots, they add less. The waitlist study measured the Ivy-Plus against flagship public universities, but it also estimated what the colleges its turned-down applicants attended add to the chance of reaching the top 1 percent. Twelve other highly selective private colleges it names, including Northwestern, Johns Hopkins, and Rice, added about a third as much as the Ivy-Plus over a flagship.[4] Eleven of those twelve are among the 52. For the other private colleges its turned-down applicants attended, the study’s figure is below a flagship’s. That group is broad, from Boston University to small liberal arts colleges, so its figure is a rough guide for any one of them.

On pay, What a Harder Door Is Worth found little to separate a harder door from an easier one: at full private prices, programs at selective private colleges came out within about $10,000 in lifetime value of matched programs at easier schools.

Hold the student’s pay fixed one more time. At full price, the strong public university comes out $192,406 ahead of the typical college in this group.[5] The price is about the same as the Ivy-Plus, and the gain in long-shot odds, on the one estimate we have, is about a third as large or less.

There is one exception, and it lives below the most famous schools. Among the Ivy-Plus, only Duke and the University of Chicago offer merit scholarships.[7] Among very selective private colleges, a group that includes the Ivy-Plus, 31 percent of dependent full-time students received merit or non-need grants in 2019-20, with a median award of $19,480. At less selective private colleges, 73 percent did.[11]

For a top student, a generous merit offer can make the college offering it the cheapest letter in the pile. But a merit award can also be, in the words of one analysis, “nothing more than a notional discount” off an inflated price.[12] Compare the final price, never the size of the award.

At full price, the college that is hard to get into but sits outside the Ivy-Plus is the one to question hardest.

What to do this fall

1. Run every net price calculator before the list is final, starting with the famous schools. The answer tells you which side of the line you are on. Nearly every school page on collegeroi.org links to its calculator, under Financial Calculator.
2. Build the list on both sides of the line. Include the reach schools, a strong in-state public university, and at least one college likely to compete for your child with merit aid. Many calculators leave merit out, so the real number may arrive only with the April letters.
3. Choose the field first. It still moves pay more than the school does, and it decides how much the network is worth.
4. Never borrow to pay the difference.

The question to bring to April

Go back to the two families. The family earning $90,000 should take the famous school with a clear conscience. It will likely cost them less, and the name and network come with it. The family earning $400,000 should ask hard questions before they sign.

When the letters arrive, ask one question of any school that costs more: what does the extra money buy that my child would not carry anyway? Below the line, the calculator will usually show nothing extra, so take the prestige for free. Above it, the answer is the network and the long shots. Decide what that is worth to your family before the deposit.

Companion Pieces

• What a Harder Door Is Worth: what the acceptance rate tells a family about pay, and what the major does.
• What the Prestige Study Left Out: the Dale and Krueger studies, their exceptions, and the gift of price.
• Yes, the Ivies Produce the CEOs: the long-shot ticket, and who it pays.
• A Parent Did the $400,000 Math in Public: what the money becomes if you invest it instead.
• Run your own colleges’ numbers in the Guided ROI Interview on collegeroi.org.

Reference Sources

1. Author’s analysis of the U.S. Department of Education College Scorecard raw data, March 23, 2026 release (institution file), with the author’s College ROI model. Accessed October 9, 2026. Schools whose predominant award is the bachelor’s degree, in three groups:
   • Ivy-Plus: the eight Ivy League schools, Stanford, MIT, Duke, and the University of Chicago (12).
   • Other private colleges whose average SAT is 1450 or more (52). Webb Institute is excluded because every U.S. student receives a full-tuition scholarship.
   • Public universities whose average SAT is 1350 or more, plus the University of California, Berkeley and UCLA, which report no SAT (28). The service academies are excluded because they are free and carry a service obligation.
   • Figures are medians across schools within each group. Pay ten years after entry is the Scorecard’s median for students who received federal aid, whether or not they finished (MD_EARN_WNE_P10). Students who never received federal aid are not counted.
   • Net prices by family income are the Scorecard’s averages for students receiving federal aid (NPT41 to NPT45), in-state at public universities. The top bracket, $110,000 and up, includes every income above that line and leaves out families who received no federal aid.
   • Completion is the share finishing within six years (C150_4). Average SAT is from students who submitted scores.
   • No adjustment is made for who enrolls.
   • The 28 public universities are in 20 states: California, Colorado, Florida, Georgia, Illinois, Indiana, Maryland, Massachusetts, Michigan, Minnesota, New Jersey, New York, North Carolina, Ohio, Pennsylvania, South Carolina, Texas, Vermont, Virginia, and Wisconsin. In the 28 other states where a public university reports SAT scores, the one with the highest average had a median full cost of $28,319, and 21 of the 28 cost less than this group’s $32,445.
2. Dale, Stacy Berg, and Alan B. Krueger. “Estimating the Payoff to Attending a More Selective College: An Application of Selection on Observables and Unobservables.” Quarterly Journal of Economics 117, no. 4 (2002): 1491–1527. Accessed October 7, 2026. The quotations are from the abstract.
3. Mountjoy, Jack, and Brent Hickman. “The Returns to College(s): Relative Value-Added and Match Effects in Higher Education.” NBER Working Paper 29276, September 2021. Accessed October 7, 2026. The fading-premium sentence is from the abstract. Section 8.2 reports “little evidence of significant heterogeneity in the returns to selectivity by gender, family income as measured by FRPL eligibility, cognitive skills as measured by high school test scores, and non-cognitive skills as measured by high school attendance.” The sample is Texas public universities.
4. Chetty, Raj, David J. Deming, and John N. Friedman. “Diversifying Society’s Leaders? The Determinants and Causal Effects of Admission to Highly Selective Private Colleges.” Published in the Quarterly Journal of Economics, February 2026. Circulated as NBER Working Paper 31492, July 2023, revised August 2025. Accessed October 7, 2026. Quotations and values are from the August 2025 revision. The next-tier estimate is Appendix Figure 24a: observational value-added on the chance of reaching the top 1 percent is 6.5 for the Ivy-Plus, 3.1 for other highly selective private colleges, and 1.4 for flagship publics, and the authors find most such variation reflects causal effects. (3.1 − 1.4) ÷ (6.5 − 1.4) = 0.33. The twelve other highly selective colleges are listed in Appendix Table 1, Panel B. Other private colleges (for example, Boston University) are at 0.4. Caltech, the twelfth college in Panel B, reports no average SAT to the Scorecard, so it is not among the 52.
   • The table values are from Table 4, which compares students who attended an Ivy-Plus college with what the same students would have achieved at a flagship (Column 6 against Column 5, which subtracts the waitlist estimate in Column 1). Income outcomes are predicted income at age 33. Flagship attendees’ own outcomes (Column 4) are lower: 65.8 percent reach the top quarter and 8.1 percent the top 1 percent.
   • The mean-income-rank sentence and the “lack adequate precision” sentence are from Section 4.3.
   • The $101,000 figure is from the introduction and Section 4.5, which compare Ivy-Plus students with flagship students reweighted on test score, race, gender, and parent income (Appendix Figure 28a).
   • The fallback-options sentence is from the introduction.
   • The comparison flagships are Ohio State, UC Berkeley, UCLA, Florida, Georgia, Michigan, North Carolina, Texas at Austin, and Virginia.
5. Author’s College ROI model, same-student comparison.
   • Each group’s median price from note 1 is applied to one student whose pay is held fixed. Lifetime value is after-tax net present value against the high school path ($484,376) at a 7.8 percent discount rate, over four years of cost.
   • Both paths assume the student finishes. The Ivy-Plus colleges graduate a larger share (note 1), mostly because of who enrolls, and any real edge in finishing would narrow the full-cost gap.
   • The gap between two prices is identical at three pay levels: $79,045, $84,260, and $115,621 a year four years after graduation.
   • Public minus Ivy-Plus: −$26,687 at incomes of $48,000 to $75,000, −$23,164 at $75,000 to $110,000, +$80,923 above $110,000, and +$199,243 at full cost of attendance.
   • Public minus the other private colleges: +$4,451, +$4,055, +$74,860, and +$192,406.
   • Full costs are the Scorecard’s (COSTT4_A), $87,813 at the typical Ivy-Plus college, $85,913 at the other private colleges, and $32,445 at the public universities, below the 2026-27 published costs.
   • At $115,621, the Ivy-Plus graduates’ typical pay four years after graduation, lifetime value is +$161,255 at the Ivy-Plus full cost and +$360,498 at the public full cost. At $79,045, the public universities’ typical pay, it is −$117,395 and +$81,848.
   • For contrast, on each college’s own median pay rather than one fixed student: at full cost of attendance, 42 of the 52 other private colleges leave their typical student who enrolled with federal aid behind the high school path, on median pay ten years after entry (start salary = that pay divided by 1.02 to the fifth power). On pay four years after graduation (graduates only), 30 of 51 do.
6. Shivamber, Leon. “College ROI Model and Calculator.” collegeroi.org, 2026. Accessed October 9, 2026. Covers 1,679 bachelor’s-granting institutions. Even at zero cost, about 30 percent of bachelor’s institutions leave their median graduate behind the high school path.
7. Each college’s own financial aid and cost-of-attendance pages, accessed October 9, 2026.
   • Thresholds as stated by each school, for families with typical assets or a similar test where the school sets one, as each school defines them. Costs of attendance are for 2026-27.
     • Harvard: tuition covered up to $200,000, parents pay nothing below $100,000, cost of attendance $95,134–$100,134.
     • Yale: tuition covered up to $200,000, parents pay nothing below $100,000, no total cost of attendance published.
     • Princeton: tuition covered up to $250,000 for most families, parents pay nothing below $150,000, cost of attendance $94,624.
     • Columbia: tuition covered up to $150,000, parents pay nothing below $66,000, cost of attendance $99,774.
     • Penn: tuition covered up to $200,000, parents pay nothing toward billed expenses below $75,000, cost of attendance $99,082.
     • Brown: tuition covered up to $125,000 for most families, parents pay nothing below $60,000 with assets under $100,000, cost of attendance $99,994.
     • Dartmouth: tuition covered up to $175,000, parents pay nothing below $125,000, cost of attendance $98,427.
     • Cornell: tuition covered up to $125,000, parents pay nothing below $75,000, cost of attendance $99,734.
     • Stanford: tuition covered up to $150,000, parents pay nothing below $100,000, cost of attendance $97,545.
     • MIT: tuition typically covered up to $200,000, parents pay nothing below $100,000, cost of attendance $92,760.
     • Duke: for North and South Carolina residents only, tuition covered up to $150,000 and parents pay nothing below $65,000. Cost of attendance $103,180–$103,975.
     • Chicago: from fall 2027, tuition covered up to $250,000 and parents pay nothing below $125,000. Cost of attendance $103,821.
   • Cornell’s line is from the Cornell Chronicle report of the 2026-27 budget, March 23, 2026.
   • MIT’s catalog gives $93,730.
   • Costs differ in whether travel is included.
   • Duke and the University of Chicago award merit scholarships. The other ten describe their aid as need-based, apart from Stanford’s NCAA athletic aid.
8. MIT Student Financial Services. “Making MIT affordable.” Accessed October 9, 2026.
9. Shivamber, Leon. We Need to Talk About Higher Education, “My Path: A Great Outcome” and “Why This Book Exists.” The author chairs the Baruch College Fund Campaign Committee.
10. Author’s analysis of the College Scorecard field-of-study file, March 23, 2026 release. Median earnings four years after graduation (EARN_MDN_4YR) for bachelor’s programs, medians across programs in each group of note 1. 27 fields have at least three programs at both the Ivy-Plus and the public universities. Across those fields, the median gap is $16,190 a year. Program counts at the Ivy-Plus are small (3 to 11 per field), and no adjustment is made for who enrolls. Pay is “the sum of wages and deferred compensation from all non-duplicate W-2 forms” plus positive self-employment earnings, measured in calendar years 2022 and 2023 and adjusted to 2024 dollars, per the Scorecard glossary. The IRS General Instructions for Forms W-2 and W-3 put bonuses and income from exercising stock options in box 1 wages. Scorecard’s Field of Study Data Documentation (September 2025, note 4) says graduates who completed a higher credential were excluded only from the earnings years that combined 2017-2018 and 2018-2019, so these medians may include graduates who have since finished a master’s. One year after graduation, the computer science medians for the same programs are $142,619 and $85,809, close to the same ratio.
11. NACAC (analysis by ASA Research). “The Shifting Role of Institutional Aid: What About Need?” Tables 1e to 1h, dependent full-time students in 2019-20 (NPSAS:20). Private colleges: very selective, 31 percent, median $19,480. Moderately and minimally selective, 73 percent, median $15,500. Very selective public universities: 23 percent, median $3,000.
12. Delisle, Jason, and Cody Christensen. “The Merit Aid Illusion: The Hidden Winners in a Competition for Affluent College Students.” American Enterprise Institute, May 8, 2019. Accessed October 9, 2026.