Since the last update in March 2025 the changes include the 2024 enrolment data, the 2024 university financial information, and graduate employment outcomes.
Research income has been the most recent significant data release, including research block grant funding for 2026 and the HERDC data used to calculate it, which goes up to 2024. The chart below shows the research-specific income sources 2017-2024.
I will next update the spreadsheet when we have projected 2026 spending on Commonwealth supported places in the first quarter of 2026. I hope by then the 2025 staff data will also be available.
Since it came to office, Labor has deferred dealing with Job-ready Graduates student contributions. First it added student contributions to the Universities Accord list of issues. In February 2024 the Accord Final Report suggested basing student contributions on lifetime earnings. Subsequently the minister said the new Australian Tertiary Education Commission would provide advice. Now we have ATEC’s legislation, but how student contributions will be handled is less clear than I expected.
All legislative references, unless otherwise specified, are to the Universities Accord (Australian Tertiary Education Commission) Bill 2025.
The ATEC bill and per student funding
The ATEC bill does not mention student contributions at all. One of ATEC’s functions, however, will be to advise the minister on:
“The efficient cost of higher education across disciplines and student cohorts and in relation to the Commonwealth contribution amounts for places in funding clusters.”: section 11(d)(ii), section labelled “Functions of the ATEC”.
In different words, in a later section on “Advice and recommendations”, a topic of advice is:
“The costs of teaching and learning in higher education and overall higher education funding amounts, including on a per student basis.”: section 41(1)(b).
The main new content in the submission is description of existing legislative powers that can achieve the same claimed policy goals as the course cancellation proposal.
The practical effect of the bill, if it passes, would be to enable the suspension of the rule of law. It would allow the minister to make decisions according to vague criteria, without consulting anyone or considering other relevant laws. Due process would be abolished; providers could be penalised with course cancellation even if they have followed the law and acted ethically at all times.
It shocks me that this Trump-style bid to rule by executive order has even been introduced into Parliament. It’s staggering that, given nearly a year to think again since its original defeat last year, the government has brought back a bill that is, in some places, even more defective than their first attempt. I am referring here to removing the requirement to consult TEQSA or ASQA before cancelling a course on ‘standard of delivery’ grounds.
Update 28/11/2025: Last night the Senate accepted Coalition amendments that exempt higher education providers and TAFEs from the requirement to offer courses to domestic students for two years before being eligible to offer courses to international students. So effectively the provision discussed in this post applies only to non-TAFE registered training organisations. As I noted in the original post, offering courses to domestic students for two years is much easier for RTOs than higher education providers. Large numbers of RTOs have already met the requirement and could move into international education.
While this is good news, enrolment caps the government will try again to legislate next year could prove another insurmountable obstacle to education providers of any kind entering the international market.
————————————————————————————————-
Last week Claire Field published an interesting overview of 15 new higher education providers since January 2024. But growth of this kind would become very difficult if the government’s ESOS amendment bill passes unamended. It would limit registration of new providers offering courses to international students. This post examines whether the proposed restriction would, in practice, be a de facto ban on new higher education providers.
Under the ESOS amendment bill providers could not offer courses to international students without first delivering courses to domestic students, but providers are generally not competitive in the domestic market without offering FEE-HELP loans. But to get access to FEE-HELP, providers must demonstrate experience in delivering higher education – in practice usually by teaching the international students the ESOS bill would stop them recruiting.
Legislative references are to ESOS Act 2000 section numbers, as they are or would be if the amendment bill passes unchanged.
The proposed changes
The ESOS amendment bill would give the minister the power to suspend, for up to 12 months, applications and processing of applications for course and provider registration: sections 14C to 14F.
To be registered on CRICOS to offer courses to international students the provider must have delivered courses for consecutive study periods over at least two years to domestic students in Australia: section 11(2).
This post focuses on the section 11(2) change by looking at how providers have entered the international and domestic markets in recent years.
While well-intentioned, this policy is unlikely to make any significant difference to Indigenous medical student numbers and could accidentally reduce the number of non-Indigenous medical students.
Is there a problem that demand driven funding can solve?
In his second reading speech, the minister noted the current low number of Indigenous doctors and the benefits for Indigenous patients of Indigenous health care workers.
The main obstacle to further enrolment increases is unlikely to be funding rather than the difficulties in finding potential students who meet the entry requirements and are not being set up to fail.
As announced last year, the government plans to crack down on so-called ‘over-enrolments’ – enrolling additional students on a student-contribution only basis once all a university’s Commonwealth Grant Scheme allocation has been used.
When a proposed new funding system is in place, from 2027, student contribution-only places will only be possible in a buffer zone above a university’s Australian Tertiary Education Commission allocation. 2% and 5% buffers have both been suggested. Currently over-enrolled universities will receive some additional funding to bring over-enrolments within their official allocation of places. However, this will not in all cases reduce over-enrolments to the permitted range. Significantly over-enrolled universities need to moderate student intakes in 2026 to bring their medium-term enrolments down.
Not many current Department of Education staff were there the last time a minister thought reducing over-enrolments might be a good idea. The story is worth telling.
Brendan Nelson and over-enrolment
From November 2001 to January 2006 the education minister was Brendan Nelson, a Liberal. Nelson was worried about the quality implications of significant over-enrolments. The first reference I can find to Nelson’s concern is in a media release from December 2001, a month into his term.
In recent years voluntary HELP repayments increased significantly, peaking at $2.9 billion in 2022-23, before dropping back to $992 million in 2024-25 (according to data released last week). This post looks at why voluntary repayments spiked and what we can expect for future years.
The spike in repayments– indexation
The 2022-23 and 2023-24 big repayment spikes in the chart above are primarily due to people repaying early to avoid high CPI indexation.
With CPI now back to normal levels this should be much less of a factor in the foreseeable future. That said, to reduce indexation costs HELP debtors considering a voluntary repayment should still make it prior to the 1 June indexation date.
Recently the Department of Education published 2021-2022 data on payments under the Higher Education Continuity Guarantee, a 2021-23 Coalition program to compensate universities for under-enrolments. It has previously released data on a predecessor program, the 2020 Higher Education Relief Program.
It shows that over the 2020 to 2022 period under-enrolments cost the Commonwealth nearly $550 million. On my estimates the sector under-enrolled by approximately 47,000 places. Eight universities were under-enrolled in each of 2020, 2021 and 2022. Only four universities received nothing under the HECG or HERP, showing that enrolment shortfalls were widespread across the sector.
What is under-enrolment?
Under the Higher Education Support Act 2003 universities get paid their maximum basic grant amount (MBGA) – see my funding agreement posts for more detail on this – or the value of their Commonwealth supported places delivered (on a relevant Commonwealth contribution * EFTSL basis), whichever is lower.
During the COVID period the Coalition decided that it would let universities keep their MBGA even if they had not enrolled enough students to justify it. This was called the Higher Education Relief Program in 2020 and the Higher Education Continuity Guarantee 2021-2023. The purpose was to provide stability for universities during COVID and post-COVID enrolment turbulence.
There is a 2024-2025 program called the HECG, but it is a redirect of money to equity programs and has nothing to do with the original purpose of the HECG.
With universities back job shedding, academics and their unions are looking for someone to blame. University leaders and consultants are being attacked for poor decisions. The government also gets criticised. UTS history professor Anna Clark says that over the last twenty years ‘we have seen gradual, steady decline in government investment across the sector’. In his recent lament Broken Universities, Graeme Turner says that there has been a ‘steady decline in the levels of funding per student’.
Five years ago, early in the COVID crisis, I wrote a post about government ‘cuts’. This post is an update.
Funding for Commonwealth supported students
As my earlier post noted, time series data is not straightforward. The chart below focuses on the major student funding programs, in today’s terms the Commonwealth Grant Scheme (CGS), HECS-HELP, and upfront student contributions. These funding sources have always had a link to the number of full-time equivalent Commonwealth supported students, although historically the money they delivered supported research as well as teaching expenditure.
Around these core funding sources other schemes serve the same purpose (e.g. transition funding) or similar purposes (e.g. NPILF). The chart below includes the Job-ready Graduates (JRG) transition funding and but excludes NPILF. It includes money paid from the Higher Education Continuity Guarantee, a COVID measure still in place for universities that ‘under-enrol’ that would normally face a CGS penalty. From 2021-2024 the time series excludes the enabling course loading that was previously in the CGS but moved to IRLSAF. But this funding is back in the CGS in 2025 due to the FEE-FREE Uni Ready places. The regional loading remains out from 2021 as it is still in IRLSAF and will join needs-based funding next year.
Overall my time series goes for simplicity over a full count of expenditure on student-related programs. In the time series, one big structural change should be noted, which is research student funding moving to a separate program from 2001, which caused a significant but artificial year-on-year decline.
Trends in total funding
Focusing on recent times, in nominal dollar terms total CGS funding dipped between 2021 and 2022, which was mostly short-term COVID places coming out of the system. HECS-HELP lending fell between 2020 and 2021, driven by the strange decision to pass on reduced JRG student contribution rates to all current students but to grandfather increased student contribution rates, so that only 2021 and later commencing students pay them. HECS-HELP lending fell again in 2022, with lower student numbers also affecting revenue from a university perspective.
Last week I raised concerns about the new HELP repayment system increasing the number of HELP debtors who face very long repayment times or lifetimes of student debt.
The calculations in that post assumed that people maintained their relative income position through their careers – for example that someone who earned the median income at age 25 would still do so at age 35, 45 etc. We know, however, that relative income fluctuates. Family commitments drive movements in and out of full-time work. Careers go better or worse than expected.
Without solving the problems involved in estimating how these changes affect HELP repayments, this post outlines findings on graduate income mobility and labour force status changes.
Quintile 5, the highest, shows strong stability. More than 80% of graduates in quintile 5 were still there or in quintile 4 a decade later. The high starting point and following stability may be due to people already doing well in their careers acquiring postgraduate qualifications.
The other quintiles all show significant movement in relative income. Upward movement is expected as we know graduate incomes increase in the years after course completion. Almost half of graduates in the lowest quintile in year one are in the top two quintiles a decade later.
Bu there is also some stability at the lower end. In the two lowest quintiles, 1 and 2, over a quarter remain in those quintiles a decade later. In quintile 3 we see a similar share falling back to quintiles 1 and 2. While some of this is career stagnation, ten years out takes into the ages when women start leaving full-time work to meet family responsibilities.
Andrew Norton 