The headline surplus disappears when you strip the windfalls
The sector's reported 4.7% margin in 2024 includes $972M in fair-value investment windfalls. Strip those out and the core teaching and research position is a structural deficit.
Government funding is eroding in real terms while staff and capital costs rise
Real per-student funding has fallen 6% since 2015. Staff costs outpace CGS indexation every year. There is no federal capital program.
Real vs nominal government funding per domestic student
Indexed to 2015 = 100. Real value fell to ~94; nominal grew to ~112. Universities absorb the gap.
CPI vs Wage Price Index vs CGS indexation rate
Staff are 55% of costs. WPI averaged 3.9%/yr (2022–24) against CGS indexation of ~2.7%/yr — the gap compounds annually.
Sector capital expenditure ($B) — universities only
Peaked at $4.5B in 2019 and has not recovered. The EIF was abolished in 2019; infrastructure now competes directly with operating budgets.
Domestic student: revenue vs cost (2024, per EFTSL)
$25,213 revenue vs $37,868 cost per domestic EFTSL. The $12,655 gap — 33% of the cost of teaching — is the designed model, covered by international cross-subsidy until it isn't.
35 of 36 universities lose money on every domestic student — ANU is the only exception
Calculated from Commonwealth Grants Scheme allocations and HECS-HELP revenue divided by total domestic EFTSL (total EFTSL minus overseas EFTSL — the correct denominator, which includes HDR students and domestic full-fee students not counted in CSP places alone). Benchmarked against the NSW Audit Office operating cost of $37,868 per EFTSL. Deficits range from $5,943 (Federation) to $20,124 (UTS). ANU's apparent surplus ($2,778) is an artefact: ANU receives direct Commonwealth appropriations under the Australian National University Act 1991 counted in the DoE CGS line but unavailable to any other university. Its actual cost per domestic EFTSL is ~$145,000 — covered by $441M in research income and $296M in international fees, not domestic student revenue.
Revenue per domestic EFTSL vs $37,868 operating cost benchmark (2024)
Bars show total domestic student revenue (CGS + HECS-HELP). The shaded gap portion represents the structural shortfall against the $37,868 cost benchmark. Sorted largest deficit first.
Australia funds universities less generously than every comparable system — and provides the least research overhead support
Australia ranks 38 of 40 OECD countries on government spend per student. The Research Support Program (RSP) — separate from the Research Training Program (RTP), which funds PhD stipends — covers ~19% of competitive research base costs, roughly one-third of the US negotiated F&A rate of 58% and less than a quarter of the UK fEC model. There is no active federal capital grants program.
AustraliaAustralia
United KingdomUnited Kingdom
United StatesUnited States
New ZealandNew Zealand
Research overhead recovery: Australia, UK, and USA
Australia's Research Support Program (RSP) is a block grant covering ~19% of eligible research base costs, calculated on ARC, NHMRC, MRFF and other competitive grant expenditure — separate from the Research Training Program (RTP), which funds PhD stipends. In the UK, QR (Quality-Related) funding and fEC are two distinct mechanisms: QR is a block grant allocated by Research England based on REF quality scores, spent freely by institutions; fEC (Full Economic Costing, TRAC methodology) is the pricing model for individual grants, under which UKRI Research Councils pay 80% and institutions must fund the remaining 20% — the QR block grant partially covers that gap. US universities negotiate institution-specific Facilities and Administrative (F&A) rates applied to a Modified Total Direct Costs (MTDC) base for all federally funded research, not only NIH — MTDC excludes equipment above certain thresholds, patient care costs, and large subcontracts. The negotiated F&A average is 58%; animal care and biomedical research facilities run ~80% due to facility intensity, which is directly relevant for Go8 universities with major medical and veterinary schools.
Negotiated F&A rates average 58% across 350+ US institutions; animal care and biomedical research facilities run ~80% due to facility intensity — directly relevant for Go8 universities with major medical and veterinary schools (SRAI, 2021). Effective rates after MTDC exclusions and the 26% administration cap average ~42%. In the UK, QR and fEC are separate: QR is a Research England block grant (~£1.56B/yr) distributed on REF quality scores and freely spendable; fEC is the grant-pricing methodology. The QR grant helps cover the unfunded 20% fEC gap. Source: Azoulay et al. 2025; UKRI EPSRC Guidance 2024; Research England QR Funding 2023/24.
Dollar gap — RSP vs international rates on Australia’s 2024 Cat 1 grants (ARC $916M + NHMRC $882M + MRFF $605M = $2.40B)
Hypothetical: USA rate (58%) and UK rate (80%) applied to 2024 Cat 1 grant totals (ARC GrantConnect 2024; NHMRC grants data 2024; MRFF grants data 2024). RSP uses a different base and also covers Cat 2–4 engagement grants — not a like-for-like comparison. Gap illustrates structural underfunding relative to comparable systems.
Eight forces are pushing and pulling on university margins simultaneously
No single factor explains the sector's position. These pressures compound — and the interactions matter as much as the individual forces.
International student revenue
The primary cross-subsidy for domestic teaching deficits, now subject to visa caps (145,000 new places for public universities in 2025) and concentration risk — 43.3% of NSW fee income from three countries.
Research: income vs overhead
Competitive grants bring income but also indirect infrastructure costs the project budget doesn't cover. RSP underfunding means research-intensive institutions cross-subsidise overhead from teaching revenue.
Domestic student margin (NSW Audit Office)
Operating costs per domestic student ($37,868) exceeded average revenue ($25,213) by $12,655 — a 33% deficit margin. The 2021 Job-Ready Graduates Scheme repriced disciplines without closing this gap; it redistributed the subsidy across disciplines but left the structural deficit intact.
HELP debt and student demand
In June 2025, the Albanese Government wiped 20% off all student loan debt — $16B across 3 million Australians, with the minimum repayment threshold lifted from $54,435 to $67,000. The relief reduces individual burden but does not change the structural dynamic: new HECS-HELP lending runs at ~$5.9B/yr, and the political ceiling on fee levels continues to constrain domestic revenue. Outstanding debt still creates enrolment hesitancy among price-sensitive cohorts — regional, low-SES, mature-age.
CPI and Wage Price Index
Staff are 55% of costs. WPI averaged 3.9%/yr (2022–24) against CGS indexation of ~2.7%/yr — enterprise bargaining in 2023–24 widened the gap further.
Low SES, First Nations, and equity obligations
High equity-cohort load means higher per-student support costs and lower per-EFTSL revenue. These students concentrate in regional and outer-suburban institutions already on the thinnest margins.
Investment portfolios and endowments
The 2024 headline surplus was driven by investment windfalls concentrated in a small number of Go8 institutions. Regional and teaching-focused universities hold few investable assets — sector-wide benchmarks that ignore this divide mislead.
Infrastructure self-funding requirement
No federal capital program since the EIF was abolished in 2019. Capex fell from $4.5B (2019) to $3.86B (2024). Every dollar of capital investment must come from operating surplus, debt, or state grants — a 1–2% margin generates almost nothing for this purpose.
The 36 public universities: 2024 financial health
Australia has no published minimum operating margin. These ranges derive from international comparators — NZ TEC 3% explicit floor, UK OfS 3% implicit threshold, Moody's EBIDA watch below 3% — adjusted upward for Australia's lack of a federal capital program. Go8 institutions must generate more to cover research overhead and self-fund capital investment; the same headline margin means very different things at different tiers.
Representative estimates from sector audit data. Verify individual figures against each university's published annual report. Adelaide University merger effective 1 January 2026 — University of Adelaide and UniSA shown at 2024 pre-merger position.
OP score (0–10): 3yr adjusted margin vs tier floor (3pts) · revenue CAGR (2pts) · salary real growth above CPI (2pts) · international revenue concentration (2pts) · capex vs 2%-of-assets maintenance benchmark (1pt). LT score (0–10): financial portfolio / revenue (3pts) · debt / revenue (3pts) · 3yr net asset change as % of revenue (2pts) · cash months (2pts). All ten inputs are ratios or percentages — scale-neutral across institution sizes. ≥7 Strong · ≥5 Stable · ≥3 Watch · <3 Stress. Pulse: Green = both ≥ Stable; Red = both ≤ Watch; Orange = structural split (strong LT, stressed OP or vice versa); Blue = mixed. Margin target per tier — Go8 5–7%, ATN/IRU 4–6%, all others 3–5%. Not a credit rating. Additional indicators shown per card (not scored): domestic enrolment trend is an estimated 3yr EFTSL CAGR — verify against each institution's published enrolment data. Debt service coverage ratio = EBITDA / (debt × 12%) — ≥2.0× serviceable, ≥1.25× watch, <1.25× pressure. Government funding dependency is not scored: CGS and HECS-HELP are the most reliable revenue in the sector and partially de-risk institutions with lower market diversification. See Method note. All figures representative estimates; verify against each university's audited annual report.
International students concentrate in ranked city institutions — visa policy hits unevenly
Melbourne and Sydney institutions with 38–79% international dependency carry the largest cap exposure. Regional universities draw <15% international revenue regardless of visa policy.
Melbourne and Sydney bear most of the cap risk
Go8 and inner-city ATN institutions hold 38–79% international teaching revenue dependency — the highest cap exposure in the sector.
India visa approvals dropped sharply in early 2026
Student visa grants for Indian nationals fell sharply in Q1 2026. Institutions with large Indian cohort pipelines face a multi-year revenue effect.
Three countries = 43% of NSW university fee income
43.3% of all NSW university fee income came from just three countries in 2024. Textbook revenue concentration risk — a single stream from a small number of origins.
2025 cap: 270,000 new enrolments, 145,000 to public universities
145,000 new places for public universities in 2025, down from uncapped prior years. Sector-wide revenue impact: ~$600M per year.
Rankings drive international demand — regional universities see little
Rankings drive international demand. Regional and outer-suburban institutions attract very little international enrolment regardless of visa policy — which is why a single sector-wide margin target is structurally inadequate.
Source-country concentration: no single replacement exists
No single country is large enough to replace China, India, or Vietnam if any major source declines sharply.
The 2027 cap holds every university at its 2026 number, so the one growth year the sector was given is now behind it
New Overseas Student Commencements (NOSC) are the Department of Education's per-provider allocation of new international enrolments. The 2026 factsheet lifted allocations sharply on a 2025 base; the 2027 factsheet repeats 2026 exactly for every public university. Planning that reads the 2025 → 2026 step as a trend will over-forecast. This section covers the 39 universities carried in this tool — 36 Table A plus Bond, Notre Dame and Torrens — and is the only view here that runs beyond the 36 public universities the rest of the page reports.
Combined allocation, 2025 to 2027
The 35 Table A universities carried with a figure in all three years. Adelaide University is excluded because it has no single 2025 row — its predecessors held 3,800 (Adelaide) + 2,950 (UniSA) = 6,750, which rolls to 7,350 in 2026, +8.9%.
Share of the 2027 allocation by mission group
Share of the 165,380 allocated across the 39 universities carried here. This is the tool's own denominator, not the Department's Table 1 total.
Change in allocation, 2025 to 2026
The one year in which allocations moved. Sorted by percentage change. Every 2026 figure carries through to 2027 unchanged, so there is no second year to plot. The five universities held at their 2025 number draw a grey stub rather than nothing: the value is a measured zero, not a missing figure.
2027 allocation by university
All 39 universities carried in this tool, ranked. The dashed line is the median. Table B institutions are drawn in gold: the Department allocates them from the other-provider pool, not the Table 1 pool, so their bar is comparable in size but not in denominator.
2027 allocation against 2024 commencing overseas load
Sorted tightest first. A ratio below 1.0 means the allocation is smaller than the commencing overseas load the university actually carried in 2024 — a binding constraint, not headroom. The two series are not on the same unit: the allocation counts students and the comparator counts EFTSL, and a commencing student contributes less than one EFTSL in their starting year. The ratio therefore flatters the allocation, so a university below 1.0 here is certainly constrained, while one a little above 1.0 may not have the room the bar suggests. Bond, Notre Dame and Torrens are absent: the bundle carries no commencing overseas EFTSL for them.
2027 allocation against the standing onshore international book
The chart above asks whether the allocation covers a single year's intake. This one asks how it compares with the whole onshore international operation a university already runs, which is the load its international revenue actually rests on. Each point is one university; both axes are log scaled so the smallest institutions stay legible beside the largest. The dashed line is the sector median, not a break-even: the axes carry different units and no true parity exists between a headcount ceiling and an EFTSL load, so the line is drawn from the data rather than asserted. Points above it hold an allocation that is a larger share of their standing book than the sector median, and have room to refresh their cohort; points below it are running a book the 2027 cap will not replace at the same rate. Australian Catholic University is not plotted, because the bundle suppresses its onshore split.
Cap pressure against financial position
Horizontal axis: 2024 net operating margin, as reported — revenue less expenses over revenue, so investment gains are included and the position is flattered in a strong market year. Vertical axis: the allocation ratio above. Bubble area scales with the size of the 2027 allocation. The lower-left quadrant — below the 3% margin floor this page uses and below 1.0 on the cap — is where a cap constraint lands on a balance sheet with no absorption.
Share of the 2027 allocation by state
Share of the same 165,380 total. Australian Catholic University and Torrens are multi-state and shown as National rather than assigned to a single jurisdiction. The cap is set nationally; where it binds is a question about two capital-city markets.
Every university, every year
Click any column heading to sort. Share of pool follows the Department's own grouping: Table 1 institutions against the 161,725 allocated to Table 1, Bond and Torrens against the 38,500 other-provider pool.
39 rows, 15 columns · scroll down and sideways within the frame; the university name stays pinned · sorted by 2027 allocation, largest first
What NOSC is. The National Planning Level sets a ceiling on new overseas student commencements across all education sectors; the Department then issues each provider an indicative allocation. 2027: National Planning Level 295,000, unchanged from 2026; higher education component 200,225, of which 161,725 goes to the institutions in the Department's Table 1 and 38,500 to all other higher education providers.
Basis of the allocation. Allocations were revised upwards where applicable on 90% of a provider's 2025 higher education NOSC activity, with 2025 offshore activity weighted 1.0 and onshore 0.5; a minimum of 50 NOSCs applies and figures are rounded to the nearest 5 or 10. Figures are before any redistribution between a provider's CRICOS registrations or its embedded pathway college. VET delivered at public universities is exempt from the 2027 planning level. An allocation is a ceiling, not a forecast and not an outturn — no university is obliged to fill it, and this section carries no favourable direction on the measure for that reason.
Two counting traps in the source. First, the 2025 factsheet lists a provider once per CRICOS registration while the 2026 and 2027 factsheets combine registrations onto one row. University of Southern Queensland (00244B + 02225M) and Victoria University (00124K + 02475D) appear twice in 2025 and once thereafter; their 2025 figures here are the sum of both rows. Reading a single 2025 row for either university understates it and manufactures a large false step into 2026. Second, the Department's 2027 factsheet lists The University of Notre Dame Australia in Table 1, its table of Table A universities. That is an error in the source: Notre Dame is a Table B institution under the Higher Education Support Act 2003, and this tool keeps it classified Table B (confirmed by Benjie Norman, 27 July 2026). Only the share-of-pool denominator follows the Department's grouping, because that is the pool the allocation was drawn from.
The Adelaide merger. Adelaide University exists from 1 January 2026. It has no 2025 allocation of its own, so it is excluded from every year-on-year comparison on this page; the University of Adelaide (3,800) and UniSA (2,950) held 6,750 between them in 2025 against Adelaide University's 7,350 in 2026, +8.9%. Its 2024 comparator figures are the sum of both predecessors.
The comparator. Commencing overseas EFTSL is 2024, the latest published year, and student load lags an allocation by construction. Net operating margin is 2024 net result over 2024 total revenue, as reported — not the adjusted margin used elsewhere on this page, which strips fair-value investment gains and is available only at sector level. Treat the ratio and the quadrant as a screen for where to look, not a finding.
The onshore book is derived, not published. The Department publishes overseas load and onshore load separately but not the overlap, so onshore overseas EFTSL here is inferred: 2024 overseas EFTSL less the load implied offshore, which is total load minus the onshore series. That arithmetic assumes every offshore EFTSL is an overseas student, which is true in all but rounding. It is the weakest number in this section and it is the only one on the page that is not read straight from a published cell — read the onshore chart and the two onshore table columns as an order of magnitude, not a measurement. Australian Catholic University is suppressed rather than shown. Its onshore series diverges from its total from 2022, and the arithmetic returns 180 EFTSL onshore against 4,345 overseas — it would have 96% of its international students studying outside Australia, which is not the case. One institution's inputs being wrong is a reason to drop that institution, not to publish the figure with a caveat. Comparing an allocation counted in students with a book counted in EFTSL is a ratio between two units, which is why the reference line on that chart is the sector median drawn from the plotted points and never a break-even.
Sources. Source: Australian Government Department of Education, Indicative allocations for higher education new overseas student commencements, 2027. © Commonwealth of Australia, licensed under CC BY 4.0. Sourced 27 July 2026. · © Commonwealth of Australia, Department of Education, Higher Education Student Statistics (enrolments, EFTSL, completions), 2026. Licensed under CC BY 4.0. Sourced 12 June 2026. · © Commonwealth of Australia, Department of Education, Financial Reports of Higher Education Providers, 2026. Licensed under CC BY 4.0. Sourced 12 June 2026.
Method and sources
Margin methodology. Operating margin = (Total Revenue − Total Expenses) / Total Revenue. Adjusted margin strips fair-value investment gains, per VAGO adjusted net result methodology. VAGO scope: Victoria's 8 universities only; national figures from Universities Australia, Critical Challenges 2025.
University card figures are representative estimates from sector audit reports and mission-group averages — not individually verified against each institution's audited statements. Verify against each university's Annual Report or the DoE Finance Publication.
Scoring framework. All ten scoring inputs are expressed as ratios or percentages, so a $3B university and a $400M university are measured on the same scales — consistent with Moody's Higher Education sector analytics approach. The tier-adjusted operating margin targets reflect research overhead differentials documented by ARC and UKRI: Go8 institutions carry a higher fixed-cost base from research infrastructure and must sustain a larger margin to self-fund capital renewal. Net asset change is normalised to percentage of annual revenue (>10% = 2pts, >0% = 1pt, ≤0% = 0pts) so that a $50M balance-sheet gain at a small regional university is weighted comparably to a $200M gain at a Go8. Three factors that Moody's weights heavily are not captured in this model: (1) student demand indicators — application volumes, yield rates and selectivity signal future revenue quality and are the leading indicator for institutions facing structural enrolment decline; (2) debt service coverage ratio — operating surplus plus depreciation relative to annual debt obligations is more meaningful than debt/revenue for institutions with long-duration bonds, where a high debt/revenue ratio may still be serviceable; (3) government funding dependency — Commonwealth Grants Scheme and HECS-HELP together are the most reliable revenue in the sector and partially de-risk institutions with lower market diversification, a factor this model captures only indirectly via the international concentration metric. Treat the OP and LT scores as a financial structure screen, not a credit rating.
Academic authority. Tuckman & Chang (1991), Nonprofit and Voluntary Sector Quarterly 20(4). Ryan & Irvine (2012), Australian Accounting Review 22(2). Yang, Simnett & Carson (2022), Accounting and Finance 62(S1). NZ TEC Financial Monitoring Framework (2022). UK OfS Regulatory Framework (2023). Moody's HE Sector Analytics (2024). ANU Council: ANAO 2024.
Research overhead. AU 19%: ARC Annual Report 2024 + RSP allocations. UK 30%: UKRI FEC guidance 2023. US ~42% effective: OMB 2 CFR 200; individual F&A rates (MIT ~58%, Johns Hopkins ~62%, Stanford ~57%) should be verified against current federal cost rate agreements. OECD figures (~USD 9,400; rank ~38 of 40): verify against OECD Education at a Glance 2025, Table B4.1.
NOSC allocations. The NOSC section is the one part of this page that runs on the tool's own data bundle rather than page-local estimates, and the one that covers 39 universities rather than 36. Its figures are the Department's published indicative allocations for 2025, 2026 and 2027, joined to 2024 commencing overseas EFTSL, to a derived onshore overseas book, and to 2024 net result over total revenue. Allocations are ceilings, not outturns, and carry no favourable direction. The onshore book is the one derived figure anywhere on this page — the Department publishes overseas load and onshore load but not their overlap — and the section states that in full rather than burying it here. The section's own footnote carries the counting traps in the source — dual CRICOS rows in 2025, the Notre Dame classification error in the 2027 factsheet, and the Adelaide merger — and every figure in it, including the intelligence summary, is computed at render rather than written.
Primary sources. DoE Finance Publication 2024 (CC BY 4.0) · DoE Indicative allocations for higher education new overseas student commencements 2025, 2026, 2027 (CC BY 4.0) · DoE Higher Education Student Statistics 2024 (CC BY 4.0) · NSW Audit Office, Universities 2024 · VAGO, Universities and TAFEs 2024–25 · OECD Education at a Glance 2025 · ABS 6401.0 CPI (CC BY 4.0) · ABS 6345.0 WPI (CC BY 4.0) · Home Affairs Student Visa Statistics 2025 · Universities Australia, Critical Challenges 2025 · ARC Annual Report 2024.