|
Harvey Mudd College will cost $104,512 for the 2026-27 school year. One year. Tuition, fees, housing, food, books, and the ride home at Thanksgiving.
Fourteen more schools cleared six figures on the same list: Duke at $103,975, Chicago at $103,821, USC at $103,162, on down to Haverford, which squeaked in at $100,026. The Princeton Review compiled the numbers and CNBC ran the table. Columbia got left out of it, which tells you roughly where Columbia's number landed.
My entire degree cost less than one year at Harvey Mudd. All four years of it, finished less than a decade ago.
Where I think this is going
Before anyone panics about the $104,512, know that almost nobody pays it. Private nonprofit colleges discounted first-year tuition by an average of 57.1% this past academic year, according to NACUBO, and nine out of ten first-year students got institutional grant money. The list price works more like a negotiating position than a bill.
The discounting is the tell. Schools raise the sticker to a number almost no family pays, then discount off it to fill seats, because the seats keep getting harder to fill.
The number of American high school graduates peaked in 2025. WICHE projects it falling roughly 13% through 2041, pulling about 576,000 students out of the pipeline. That decline is structural. The kids who weren't born in 2008 can't enroll in 2026, and no marketing budget fixes a birth rate.
So you've got a shrinking pool of buyers, a rising list price, and deeper discounts every year to fill the class. That math ends somewhere. At least 49 colleges have closed or announced closures since March 2020, most of them small private nonprofits running on tuition revenue with thin endowments behind them. A Philadelphia Fed model put the worst-case scenario at up to 80 closures a year.
My read is that the schools on that six-figure list are mostly safe. Duke isn't going anywhere, and neither is Georgetown. The risk sits with the small private colleges in shrinking regions, the ones with modest endowments and near-total tuition dependence. There are several hundred of those. If your kid is four today, some of the schools that land on their list in 2040 won't exist by the time they apply.
Which changes how I'd fund this. When the target itself might move, flexibility is worth more than precision.
If you're already putting money away for a kid and you're not sure it's sitting in the right place, you can grab 20 minutes on my calendar and we'll look at what you've got and where it actually belongs.
What I actually do with clients
Two questions come up every time. Does a 529 make sense, and should we open a UTMA. The order matters more than people expect.
The 529 gets a lump sum, early
I like putting a meaningful chunk into a 529 early and then letting it compound. The money grows tax-free and comes out tax-free for qualified education expenses. Most states hand you a deduction on the way in, and Pennsylvania lets you deduct contributions to any state's plan, which is unusually generous.
The old objection to 529s had teeth. If the kid skipped college or landed a full ride, you paid income tax plus a 10% penalty on the growth to get your own money back out. SECURE 2.0 softened that. Leftover 529 money can now roll into the beneficiary's Roth IRA, up to $35,000 across their lifetime.
The conditions are strict:
- The account has to have been open at least 15 years
- Anything contributed in the last 5 years is off limits
- Each year's rollover counts against the annual Roth limit, which is $7,500 in 2026
- The beneficiary needs earned income at least equal to whatever gets rolled
That 15-year clock is why I'd rather open the account when a kid is two with a smaller deposit than open it when they're twelve with a bigger one. The clock is written into the statute, and it doesn't care how much is sitting in there.
The UTMA is where people get surprised
A UTMA is a custodial account. You fund it and you manage it, and at a set age the assets become legally the kid's to do whatever they want with. Boat, business, Vegas, down payment. Your opinion stops counting on that birthday.
Most people assume that age is 18. In most states it's 21. In Pennsylvania the default is 21, and the custodian can elect up to 25 when the account gets opened, which almost nobody does because almost nobody knows it's on the table. That one election is the difference between handing a 21-year-old $180,000 and handing a 25-year-old the same $180,000.
I know exactly what 21-year-old me would have done with a six-figure account, and it would not have been stewarded well. Most of the reps I work with say the same thing about themselves, usually with a specific story attached and a little wince.
There's a second cost that's easier to miss. UTMA money counts as the student's own asset on the FAFSA, assessed at up to 20% a year. A parent-owned 529 gets assessed at a maximum of 5.64%. Same $50,000 either way: the custodial version costs you roughly $10,000 of aid eligibility, the 529 version about $2,800.
Where the rest of it goes
Usually a joint brokerage account in the parents' names.
You can put in as much as you want, pull it out for anything, and nobody's birthday transfers control away from you. You pay capital gains when you sell, and that tax is the price of keeping every option open. For most families I work with, it's worth paying.
If the kid goes to a $104,512 school, you sell some and write the check. If they skip college and start something at 20, you sell some and fund that instead. And if neither happens, the money is still yours, which it was the whole time.
What we're planning to do
We don't have kids yet. When we do, the plan is a 529 funded with a lump sum early, mostly to start that 15-year clock, then consistent monthly money into a joint brokerage account in both our names.
Probably no UTMA. I want to keep the ability to say no at 21.
Every family lands somewhere different on this, and it usually comes down to one question. If your kid told you at 17 that they weren't going to college, would you want that money back?
| 
