Does free university pay off? Countries that charge fees come out ahead on four of five measures
Two claims, each in its maker's own words, against the same five
measures from the same OECD tables for every country. The Australian Greens say education
“should be accessible to everyone, regardless of their background or financial
situation”
, and want university free. The Commonwealth's 2014-15 Budget said the cost should be shared
“between students — who benefit from having tertiary qualifications — and
taxpayers”
— a sentence from a budget whose fee deregulation did not pass the Senate, but whose
cost-sharing is still how HELP runs. We counted 7 countries where public university is free or
near-free and 10 that charge, on participation, completion, graduate outcomes, public cost and equity —
and published the table as it came out, including where each slogan loses.
The rule
Pinned before any data was opened, and committed in the open on 8 October 2026: every OECD member with a published tuition-fee figure for domestic students at public bachelor-level institutions in the OECD's Education at a Glance 2025 (Table C5.3, the chapter's own workbook, 2022/23, USD converted at purchasing-power parity). Free or near-free: average annual tuition under US$600. Charging: over US$4,000. Countries in between are listed below with their fee, not compared — the comparison is between clear cases. Beside every charging country's fee: when a student pays — upfront, by deferred income-contingent loan, or a choice — from the same chapter's loans table, because a HECS debt and an American cheque are not one situation, whatever the sticker says. The rule had casualties and we kept them: Germany, the free side's favourite example, falls out because its bachelor-level fee cell is not published, and Argentina falls out with no fee figure at all. The rule beat our own pitch's examples; that is what it is for.
The two claims
The free side's words, from the Australian Greens' platform, opened 8 October 2026:
"bring back free uni & TAFE", because, in
their tertiary-education page's words,
“Education is a basic human right and a public good that should be accessible to
everyone, regardless of their background or financial situation.”
The same page promises to wipe student
debt; that is beside the quote and not counted here. If free delivers that accessibility, it should show in
measures 1 (who gets in) and 5 (who gets in from homes without degrees).
The fees side's words, from the
Commonwealth
Budget overview of 2014-15, opened 8 October 2026: “sharing fairly the cost between
students — who benefit from having tertiary qualifications — and taxpayers”
. The 2014 deregulation
behind that sentence failed in the Senate; the cost-sharing did not, and runs today as HELP. If students
really capture the benefit, it should show in measure 3 (graduate employment and the earnings premium).
The table
Fees are 2022/23 in USD PPP (Table C5.3). "When you pay" and borrowing shares are the same chapter's loans table (C5.5). Participation is tertiary attainment of 25–34 year-olds, 2025, from the OECD's own data API. Completion is bachelor entrants completing any tertiary level within the theoretical duration plus three years (Table B5.1). Employment is 25–34 tertiary-educated, 2025; the earnings premium indexes upper-secondary pay at 100 for 25–64 year-olds in full-time work, year in the cell (one series for every country). Table C5.3’s own footnote gives differing fee reference years for Australia, England, the United States and Türkiye. Public cost is government direct spending on tertiary institutions as a share of GDP, 2020 (the latest the dataflow publishes; New Zealand 2021), with spending per student (all sources, USD PPP, 2023) beside it. Equity is the share of 25–34 year-olds whose parents did not finish upper secondary who themselves attained tertiary, 2023 (Table A1.4, Survey of Adult Skills). England appears where the fee table reports England and the UK where the labour tables report the United Kingdom; the unit is in the row's name.
| Country | Fee US$/yr | When you pay | Attain 25–34 | Complete +3y | Employed 25–34 | Earnings idx (yr) | Spend/student US$ | Public cost %GDP | Equity: first-in-family attain |
|---|---|---|---|---|---|---|---|---|---|
| Free or near-free | |||||||||
| Denmark | 0 | nothing to pay at public institutions | 52.4% | 66.8% | 86.0% | 123 (2024) | 27,594 | 1.59% | 48.5% |
| Finland | 0 | nothing to pay at public institutions | 38.4% | 61.4% | 87.9% | 139 (2023) | 21,179 | 1.45% | not published |
| Norway | 0 | nothing to pay at public institutions | 56.8% | 67.5% | 89.8% | 118 (2024) | 20,963 | 1.82% | 29.6% |
| Sweden | 0 | nothing to pay at public institutions | 53.6% | 54.3% | 86.8% | 125 (2024) | 29,167 | 1.34% | 22.9% |
| Türkiye | 0 | nothing to pay at public institutions | 45.5% | 78.5% | 73.3% | 153 (2024) | 9,256 | 1.06% | not published |
| France | 252 | nothing to pay at public institutions | 55.8% | 45.7% | 86.9% | 157 (2022) | 20,823 | 1.17% | 32.1% |
| Luxembourg | 494 | nothing to pay at public institutions | 62.4% | 54.7% | 88.0% | 150 (2024) | 64,008 | 0.43% | not published |
| Charging | |||||||||
| England (UK) | 13,135 | deferred, income-contingent (93% borrow) | 61.2% | 80.4% | 90.9% | 135 (2024) | 30,613 | 0.51% | 37.4% |
| United States | 9,596 | loan optional, fixed or income-contingent (45% borrow) | 52.8% | not published | 85.4% | 170 (2024) | 35,313 | 0.94% | 22.3% |
| Poland | 7,497 | upfront — but full-time public study is free; the fee is for part-time and second degrees | 45.2% | 62.1% | 92.6% | 152 (2024) | 14,971 | 1.04% | 10.9% |
| Canada | 5,590 | loan optional, fixed or income-contingent (39% borrow) | 67.7% | 68.6% | 85.7% | 141 (2023) | 29,822 | 1.21% | 38.1% |
| Japan | 5,645 | loan optional, fixed or income-contingent | 67.5% | not published | 91.2% | not published | 19,181 | 0.49% | not published |
| Lithuania | 5,458 | upfront, or fixed-repayment loan (5% borrow) | 60.8% | 63.5% | 91.6% | 169 (2022) | 15,289 | 0.86% | 4.5% |
| Korea | 5,132 | loan optional, fixed or income-contingent | 71.1% | not published | 79.5% | 130 (2024) | 15,035 | 0.67% | 45.3% |
| Australia | 5,108 | deferred, income-contingent (80% borrow) | 58.2% | 58.4% | 91.4% | 131 (2024) | 23,152 | 0.68% | not published |
| Latvia | 4,824 | upfront, or fixed-repayment loan (4% borrow) | 44.7% | not published | 89.0% | 153 (2024) | 14,070 | 0.80% | not published |
| New Zealand | 4,748 | deferred, income-contingent (69% borrow) | 47.7% | 58.9% | 88.7% | 136 (2024) | 20,629 | 0.96% (2021) | 38.8% |
In between, listed with their fee and not compared: Israel 3,088 · Netherlands 3,041 · Italy 2,570 · Romania 2,163 · Spain 1,708 · Croatia 1,660 · Switzerland 1,427 · Flemish Community (Belgium) 1,410 · Austria 1,043 · French Community (Belgium) 433. The Belgian communities are not OECD members in their own right, so they stay listed and are never compared (the checker's view, taken; the French Community sits under the US$600 band, which is why the sentence is needed). Sensitivity, as promised when the bands were pinned: lifting the free band to US$1,500 would add Austria and Switzerland; dropping the charging band to US$3,000 would add Israel and the Netherlands; no lead below changes under either move. The tenderest cell is Poland: Charging by the pinned fee rule, though its full-time public study is free. Move Poland to the free group and the earnings medians flip to 144.5 against 138.5, free ahead; drop it and they sit level at 139 against 138.5. No other measure's lead moves under either treatment — which is why the earnings comparison is reported as level, not as anyone's lead.
Where free wins, where it doesn't
Counted both ways, as it came out. On the Greens' own measures: the charging group's median participation is higher — 59.5% of 25–34 year-olds hold tertiary against the free group's 53.6% — and on first-in-family attainment the published rows give the charging group 37.4% against the free group's 30.9%, though only four free countries publish the number, so that lead is thin ice, said plainly. Denmark, free, posts the single best equity number in the table (48.5%); Lithuania, charging, the worst (4.5%). On the Commonwealth's measure: tertiary graduates out-earn upper-secondary workers on both sides — median index 141 charging, 139 free, a gap of two points that the sensitivity paragraph shows Poland alone can erase or reverse, so we call it level — "students benefit" is true everywhere, including where they pay nothing, which cuts both ways: the premium exists without fees, and fees exist without killing the premium. Completion and employment medians sit within a few points of each other. The one unambiguous difference is the column no slogan mentions: free-group governments put a median 1.3% of GDP into tertiary institutions; charging-group governments put 0.8% and collect the rest from students — mostly, in England, Australia and New Zealand, as deferred income-contingent debt rather than cash at the door.
The limits
Sources
- Australian Greens, platform: education, opened 8 October 2026
- Australian Greens, tertiary education portfolio, opened 8 October 2026
- Commonwealth Budget 2014-15, overview, opened 8 October 2026
- OECD EAG 2025, chapter C5 tables (fees C5.3, loans C5.5), opened 8 October 2026
- OECD EAG 2025, chapter B5 tables (completion B5.1), opened 8 October 2026
- OECD EAG 2025, chapter A1 tables (equity A1.4), opened 9 October 2026
- OECD data API, EDU.IMEP dataflows (attainment, employment, earnings, finance; query URLs and saved slices in the ledger), opened 8–9 October 2026