Why two 'elite' schools can sit 170x apart on tuition

The first thing that hit me while loading the data was the Sorbonne's €340 next to Columbia's $66,139. Both are world-recognized elites. Yet convert them roughly and the tuition gap stretches to around 170x. Same word ('top university'), but the numbers live in entirely different orders of magnitude. At first I thought "is the U.S. just gouging?" — but lining the items up one by one, that wasn't it. Tuition isn't a price; it's an institution. It's not a value a market set through supply and demand, but the output of each country's policy on who bears how much of the cost of higher education.

So comparing countries by FX alone is guaranteed to mislead. Germany's TU Munich charges €760 including international students — and even that is essentially just a ~€380-per-semester admin fee. France's Sorbonne is €340. Both are 'nearly free' because national taxes absorb the universities' operating costs. The U.S., by contrast, leans on private schools and endowments, which pushes the sticker upward. A conversion table erases this structural difference wholesale. Put €760 and NYU's $60,438 on the same ruler and read it only as 'a cheap one vs an expensive one,' and the question that actually matters — who pays the cost — evaporates.

Nominal tuition and real burden are different numbers

Everything on the game screen is nominal tuition (the sticker price). But the real burden a student actually pays becomes a completely different number once scholarships and aid run their course. Harvard's sticker is $59,076, yet about 55% of students receive aid, and tuition is zero for families earning under $85,000. KAIST's nominal figure is 3.3M KRW, but it grants full scholarships to high performers, so most students pay effectively nothing. A ranking by sticker and a ranking by real burden become entirely different pictures.

This is the point I was most careful about while building the data. The sentence "Harvard is 60 million won" is only half true. The sticker is right, but fewer students than you'd think pay the full sticker. U.S. private sticker prices commonly fall to 30–70% of the real burden once average aid packages apply. So if you see a big number in this game and stop at "that place is just expensive," you've missed the core. Nominal is the label; real burden is the receipt.

Domestic/international dual pricing — same lecture hall, a 4x tag

Nothing exposes the 'it's an institution' point more nakedly than dual pricing. Oxford charges UK home students up to £9,250 a year, while internationals pay £39,010. Same professors, same lecture hall, same library, but a price tag over 4x apart. International tuition at Australia's University of Melbourne ($45,000) is such a major revenue source for the government that international fees amount to roughly 2% of GDP. Singapore's NUS posts an international sticker of $18,500 but ties it to a grant you repay by working three years in Singapore after graduating.

Why this matters: you can't pin down 'this country's tuition' from a single school's number. Whether you remember Oxford as £9,250 or £39,010 depends entirely on 'whose seat you're in.' FX conversion can't carry that difference of standpoint.

Cheap-looking tuition often hides the real cost elsewhere

ETH Zurich is a global top-10 engineering school, yet tuition is CHF 1,460 — about 1/40 of a U.S. school's. But a fun fact made me pause: the real burden isn't tuition, it's Switzerland's brutal cost of living. China's Peking University (~$3,600) and IIT Bombay (~$2,000) also post dirt-cheap nominal tuition, but the cost didn't vanish — it migrated to admission difficulty. IIT admission runs about 100:1. A low money barrier in exchange for a high exam barrier. There's variation within a single school too: Yonsei's engineering college runs about 2M KRW more than humanities, and its medical school exceeds 10M KRW.

Laid out as a table, it clicks. Look only at the 'nominal tuition' column and you see a 1/40-to-170x spread, but move to 'real burden' and the variable creating the gap differs by school — aid, or nationality, or taxes, or living costs, or the entrance exam. Mash all of that into a single exchange rate and the comparison looks plausible but is hollow.

An institution can also be uniform within one country

Japan shows the power of institutions in the opposite direction. Japanese national university tuition is identical nationwide at ¥535,800. Tokyo, Kyoto, a rural national university — all the same. In a market, a Tokyo University degree should carry a premium, yet Japan fixed a single price. Korea's Seoul National has likewise frozen tuition since 2009. Cases like these show how easily the market intuition that 'more prestigious means pricier' collapses in front of an institution.

What organizes this structural difference one level up is the OECD's annual report, Education at a Glance. It compares, country by country, the ratio in which the public purse (taxes) and households (tuition) split the cost of higher education, and the split cleanly separates countries where the public share dominates (the Nordics, Germany) from those where the household share is large (the U.S.). Our game's numbers are, in the end, points on this 'who-bears-the-cost map.'

✍️ Operator's note — "U.S. tuition is insane, Korea and Europe are the decent ones" — that was the full depth of my own summary before I filled in the items. Afterward, it didn't hold. The U.S. is a system where households pay a lot, Germany a system where taxes pay, the U.K. a system that splits home from international. Not one country had made the cost disappear. Only who pays differed. Now every tuition number gets the same opening questions from me: "Is this sticker or real burden? And who's structured to pay it?"

One honest caveat to close: the numbers here are publish-time snapshots, and each university carries only a single representative value. The spread across majors, class years, and campuses isn't captured. So a score here is just a learning gauge of how calibrated your price sense is — not a basis for an enrollment decision and not investment advice, nor is it user statistics. The right way to use this mode is, when you see a big number, to resist converting it straight by FX and instead ask first: "Is this nominal or real burden, and which system is built to pay for it?"