Part 1 of this pair of posts addressed the principles that underpin the funding system – important principles such as:
- sharing the cost of education between students and government
- how funding for capital should be handled
- the need for transparency in the funding formula.
It discussed the need to ensure that funding is well-linked to the other levers that drive the tertiary education system. And it argued that funding should be demand-led, reflecting student preferences; attempts to link funding to experts’ and politicians’ judgements about what is strategic often end in tears.
Finally. It canvassed the difference between funding and investing, proposing that the review commission a trial of an investment approach to funding that would see multi-year funding to a trial group of high-performing TEOs.
In this second part, we get onto some of the bread and butter questions:
- Funding for sustainability – making sure the funding system deals with costs by getting a better understanding of what drives the cost of delivery
- Funding of learning – what society, students, families and employers want from the system is learning. Yet, mostly, we fund teaching as the means of generating learning. Why? What is the history of funding learning?
- Funding learning at work – workplace-based tertiary education and work-integrated learning.
Funding for sustainability. Funding and cost
We need a funding system that ensures the sustainability of the system and of well-managed TEOs. Sustainability means that TEOs have the ability to generate financial surpluses at a level that enables them to maintain and invest in their assets, to improve their systems, services and facilities – to enhance their quality. Breakeven leads to erosion of quality. Losses mean a choice between closure and bailouts.
But sustainability means that institutions shouldn’t be able to make excessive surpluses – if institutions can readily generate very high surpluses, then that means that funders (government and students) are paying more than they should. Resourcing needs to be at a level that obliges institutional managers to manage well and prudently, but that allows the sort of surpluses that are needed for maintenance and enhancement of assets.
Redesigning funding relies on understanding costs
If funding is to underpin sustainability and if TEOs are to be able to respond to student preferences in a demand-led funding system, then tuition revenue (and hence, funding) needs to bear a relationship to cost of delivery. Funding for sustainability implies an understanding of the drivers of costs.
Most tuition funding at present is volume driven with differentials by field of education. But it’s not clear that field of study, by itself, is the sole (or even the most important) driver of costs.
Recent Australian research (by Croucher and Rahimi)[1] set out to identify factors associated with cost of delivery. The researchers analysed cost and delivery data on nearly 190,000 units of study (what we would call “courses”) in 12 Australian universities that cover more than half the EFTS in the Australian university system. They grouped the courses into eight broad groups according to cost of delivery per EFTS. They found that higher average cost per student was often associated with factors like small class-size and labour-intensive postgraduate teaching while lower costs arise with scale-based online delivery models. Field of education and delivery mode were found to play a contextual role in differentiating the cost classes. In other words, certain fields are more likely to use high-cost delivery approaches than other fields. But it is not the field, itself, that drives the cost; rather, it is the delivery approach. What we think of as a low-cost field may well include very high-cost courses, while a high-cost field will include some low-cost courses.
That research calls into question the notion that field of education is the primary driver of unit cost and hence that differentials in funding rates should be primarily driven by field[2]. It suggests that differentials need to take account of a wider range of factors: scale of delivery, mode of delivery, possibly even institutional context, as well as field of education.
- On scale: There are serious economies of scale in tertiary education – there are very high fixed costs but, in most cases, relatively low marginal costs. Yet the current funding formula takes no account of enrolment scale. The new Australian research suggests there is a case for recognition in the funding system for providers serving small populations, for types of education that command only small numbers of students, for TEOs whose specialties mean they cater for very small market segments. In the late 1990s there was a small base grant component in the NZ tertiary education funding model that was a token recognition of the challenges faced by smaller TEOs – but this was removed from 2000[3].
- On mode of delivery: Some modes of delivery are inherently costly – for instance, those involving equipment which imposes limits on numbers. Or supervised work-integrated learning. Others – such as online delivery – may have high set-up costs but are highly scalable thus, can drive economies. Taking no account of delivery mode can incentivise providers to select lower cost delivery modes and to avoid higher cost models with less regard for the quality or value.
- On the characteristics of students: Students who are at risk of attrition or failure impose additional costs on teaching. If there is no allowance in the funding model for student characteristics, that suggests the model has implicitly assumed that the additional costs of teaching those at risk are balanced by “savings” on those who are “easier” (ie, less costly) to teach.
The current funding model for non-degree tertiary education (DQ3-7) recognises student characteristics and delivery mode alongside volume and a loading for field of education (but not scale). That gets close to squaring with the factors identified by Croucher and Rahimi in their recent paper. But the higher education model (DQ7+) recognises only volume and field of education.
The new review of the funding model needs to take better account of cost drivers – possibly using the current DQ3-7 model as a starting point. But the redesign needs to ensure that the model takes due account of the overlap between factors that are correlated (for instance, mode of delivery and field of education) lest it create double jeopardy for some TEOs, double rewards for others[4].
And it needs to recognise that creating a new formula creates new incentives. It needs to balance simplicity and complexity in the allocation formula. It Complexity risks TEOs targeting measures that maximise funding, rather than lifting performance.
Funding of teaching, funding of learning
Typically, the teaching and learning funding system funds delivery of education. In practice, that is funding for the primary activity – teaching. Learning – students’ acquisition of skills and knowledge – is a consequential activity. We fund learning mostly on the assumption that when teaching occurs, learning follows. But that isn’t a perfect assumption – some TEOs add greater value than others, but that difference is usually invisible.
We have flirted with recognising learning explicitly in the funding formula in the past. Between 2012 and 2018, 5 percent of teaching and learning funding was “at risk”, performance linked, subject to measures like course completion and qualification completion etc. In effect, this was an attempt to moderate the funding for teaching by an explicit measurement of how much learning had occurred. This approach was abandoned by the government in 2018.
The main flaw in the performance-linked funding approach used in the past was that there were no controls for the type of student; TEOs that attracted older students or students from educationally disadvantaged backgrounds were likely to score less well on the measures that drove the performance-linked funding. Because performance-linked funding worked as a “fine” for low performance (rather than a bonus for good performance) and because the threshold for fining was very high, few TEOs were fined – though it was painful and humiliating for those that were. (One institution publicly declared an intention to reduce part-time and distance delivery in response to having been fined – an utterly perverse outcome, given the likely effects on access).
While it may have been possible to apply statistical controls for some younger learners (those who had taken NCEA in the three or four years before entering tertiary study[5]), there are no variables that would allow robust controls for learners who are older or who have non-standard backgrounds. The decision to abandon performance-linked funding was a good one and its demise appears to have been quietly welcomed.
However, interestingly, there has recently been a proposal (in a white paper from the Skills Group[6] and also in the Green Party’s current election policy) to reintroduce performance-linked funding for apprenticeships, paid as a bonus to the Apprenticeship Boost funding.
I am not convinced – the experience of performance-linked funding makes me cautious…. But the review of the funding system will need to consider this proposal, as well as the broader question of funding learning.
Learning at work …
Workplace-based training
Much of the discussion above relates implicitly to provider-based teaching and learning. But it is also applicable to workplace-based learning (WBL). WBL, however, has different dynamics, in that the primary instructor is not an educational professional working in or for a TEO, but, rather, a skilled and experienced staffer in the firm that employs the trainee. The instructor and the learner are supported by itinerant advisors employed by the TEO that signs the training agreement, and that arranges and manages the training.
The WBL system has a number of distinctive features relating to the funding system:
Employer costs
First, research[7] in both New Zealand and Australia shows that firms bear two types of costs in taking on trainees:
- Paying wages to trainees (who are relatively non-productive as employees in the early years of their training)[8]
- Reduced productivity due to giving some of the most skilled workers training, mentoring and supervisory roles, rather than from pure production roles.
One effect of that is to disincentivise smaller employers from taking on trainees, encouraging them instead to recruit qualified staff who have gained their certification through larger employers – the problem of “poaching”[9].
The funding system currently recognises those employer costs only through the Apprenticeship Boost system, which has succeeded in lifting the take-up of apprentices by employers but which targets only some types of trainees and only some industries. In Australia, there are similar government subsidies for employers to take on apprentices, recognising the costs of training to businesses.
Analysis by the Apprenticeship Employer Network Victoria suggests that the return on investment to employers from taking on an apprentice is positive from the third year of a four-year apprenticeship – even discounting the government subsidy[10]. However, some NZ firms, even in industries favoured by the Apprenticeship Boost choose to hire qualified staff, in preference to hiring apprentices or trainees. In effect, some small firms rely on the large leading firms supply most or all of the training in the industry and the Apprenticeship Boost may have only partially addressed this issue.
The funding review should investigate if the Apprenticeship Boost or a similar subsidy is actually needed, and if so, at what level. It should look at whether other forms of WBL – not simply apprenticeships – should be included. It should question whether it makes sense to restrict the subsidy to just some favoured industries (as in the current Apprenticeship Boost funding) – rather than, as in Australia, providing the subsidy to all apprenticeship agreements.
Finally, to address the problem of poaching, it should consider whether there is a case for an industry training levy of employers with discounts or credits for employers who are active in training or workforce development[11]
Itinerant training advisors – is the training advisor scheme resourced adequately?
Second, the evidence[12] shows that the itinerant advisor system has often been seriously deficient, with some training arrangements allowing for no more than four visits to trainees on the job per year, where some trainees are left to cope with online distance delivery for the theoretical side of their learning even if they face literacy or numeracy challenges, where some training advisors are not able to provide pastoral support to trainees.
It’s not clear how much the weakness of the itinerant advisor approach is down to TEO performance or to funding. It is possible that this problem may have its origins in the effective monopoly held by industry training organisations on arranging and managing training that applied before the VET reforms that began in 2019 and ran through to 2025 – in which case, the principal driver of the problem was TEO performance. Or it may be that the training advice under-resourced[13]. Has greater contestability in training arranging addressed this problem? That is a matter for investigation on the review. If not, there is a case for greater funding, perhaps funded through learner fees – which could be borrowed through the loan scheme[14].
Living cost support for WBL
The recent Skills Group White paper[15] proposed that apprentices should be eligible for student allowances and the living cost component of student loans. That’s an original and interesting idea
But think about it …. currently, the training wage pays more than $750 for a 40-hour week while a student allowance is abated, dollar for dollar, when a recipient’s weekly wage gets to $285. Meaning if an apprentice earning the minimum wage was entitled to apply for a student allowance, the allowance would be abated to zero. Meaning that, in practice, the Skills Group’s idea is not that apprentices should be entitled to student allowances but, rather, that they should be entitled to borrow living costs through the loan scheme (which would, in effect, lift the training wage to 15% above the minimum wage, but with an obligation to repay an interest-free loan of up to nearly $70k[16]). An original, interesting, but possibly not entirely serious, suggestion. Or is this a subtle way of proposing a shift of support for apprentices from employers to the government?
The Green Party election policy takes the reverse position. Their plan is to abandon the training wage, meaning presumably, that they believe apprentices should be paid the minimum wage or more (and they plan to pay a tools and equipment grant to apprentices (an idea borrowed from the 2013/14 Apprenticeship Reboot scheme).
Neither the Greens and the Skills Group acknowledge the fact that many employers pay apprentices above the training wage, especially because the evidence suggests that, as apprentices progress through training, they contribute to (rather than act as a drag on) the employer’s bottom line[17]. That suggests that a move to shift the living costs of apprentices from employers to government is unlikely to be warmly received, whichever party leads the government in 2027.
Work-integrated learning
Work-integrated learning (WIL) has always underpinned provider-based vocational qualifications (especially in professional areas such as teaching and the medical professions). Evidence[18] on the effectiveness of WIL as a complement to theoretical learning has seen its take-up spread beyond vocational qualifications and into general (ie non-TVET) tertiary qualifications (such as science, IT and business[19]), from certificate level up to postgraduate qualifications.
But effective, high-quality WIL carries extra cost. The placement needs to be built on specific educational input from the provider as well as from the workplace mentor. And assessment of WIL students’ performance is a particular challenge. Because each student is likely to have a different work placement and performs a different set of tasks in a different context, it is all but impossible to achieve the kind of economies of scale that exist for on-campus learners. This means that some providers will avoid WIL because of the fact that the cost is very much higher[20], (perhaps reverting instead to compliance-based work experience placements).
The problem faced by institutions in developing and managing high-quality WIL is that, except in a few areas, the resourcing system fails to recognise this additional cost, thus creating incentives on institutions to opt for less costly approaches to teaching and assessment (potentially compromising quality), despite the acknowledged value of WIL. Alternatively, providers may be forced to reduce the rigour (and hence, the quality) of WIL delivery and (especially) assessment – in effect, replacing WIL with work experience – ie, job placements but without the deliberate structured learning and assessment that characterises WIL.
This suggests that the funding review should treat WIL as a one of the modes of delivery.
Payment for interns
In some professional areas where WIL is mandatory – teaching, nursing – student advocates have argued that learners on internships should receive some form of stipend. Interns are mostly eligible for a living allowance through the student loan scheme (or via student allowances). At present, the only additional government support for interns or other WIL placements is for medical students whose sixth year is a trainee internship year – which involves a number of extended placements, often with an overseas placement.
Students in other professional qualifications – nursing, for instance – have complained that they too deserve additional income support during placements. Most famously, during the 1987 election campaign, the Prime Minister was confronted by angry teacher trainees complaining that they were obliged to wear smart clothes when undertaking placements and that this was unaffordable within the limits of the student allowances scheme[21]. More recently, as WIL has gained momentum, these complaints have received media attention. Advocates argue that students whose programme involves multiple extended placements have to meet additional costs – potentially travel costs and/or temporary relocation costs, sometimes additional equipment or clothing costs – to undertake their internships. They argue that the level of living cost support from student loans and allowances is insufficient to meet those extra costs and that the student financial support system should recognise those extra costs.
The Greens policy addresses this question, proposing a $380/week payment for those on placement (presumably in addition to the living cost support from the loans and allowances schemes[22]
This issue is at the margins of the scope of this paper (which excludes the loans and allowances schemes). But a funding review that doesn’t acknowledge the issue will be missing an irritant that won’t go away any time soon, especially as WIL gains more prominence as a delivery mode.
Conclusion
A review of the system for funding teaching and learning in tertiary education is a good idea. Even the country’s two cornerstone political parties agree (even if they will likely differ as to their focus and the decisions they make).
The best way forward is for the agencies to take a principles-led approach to the review, ensuring we end up with a system that delivers on the enduring outcomes we expect from tertiary education, with a system that is demand-led, a system that ensures the sustainability of TEOs, that shares the cost of the system between students and government while maintaining access through providing students with liquidity to meet their share of the cost, that manages the tension between promoting collaboration between TEOs while ensuring that we keep a competitive edge, treats learning in the workplace appropriately, …
Read the bibliography here.
Endnotes
[1] Croucher and Rahimi (2026)
[2] Detailed analyses of delivery costs (such as those conducted in New Zealand, Australia and the UK over the last 10-15 years) have reinforced the notion of field of education as the driver of cost, but that is because those studies used field as the unit of analysis; those three studies identify differences between fields but don’t look at what drives those differences.
[3] See Roberts and Peters (1999)
[4] The DQ3-7 model appears to manage this by grouping field of education and mode of delivery in the same component.
[5] Earle (2018) and Engler (2010) both show that NCEA predicts success in tertiary education.
[6] Skills Group (2026) The future of apprenticeships.
[7] See Jones (2023), Scarlatti (2025) and see the analysis in this seminar (view from 22 minutes to 28 minutes).
[8] Noting however, that apprentices can be paid the training wage, which is lower than the minimum wage (although most are paid above that base level)
[9] See Jones (2023), Herriman, Smyth, Mischewski (2025) and Mischewski and Smyth (2025)
[10] See the analysis in this seminar from 22 minutes to 28 minutes. See also Skills Group (2026) for returns to NZ, Australian, US and Canadian employers from taking on apprentices. See also Jones (2023) pp 21-23
[11] Mischewski and Smyth (2025)
[12] See Allen & Clarke (2025) and Herriman, Smyth and Mischewski (2025)
[13] Mischewski and Smyth (2025)
[14] Ibid
[15] Skills Group (2026)
[16] On current data, the training wage is $766 a week. Supplement that by the maximum student loan living cost borrowing ($333 a week) and you get a weekly income of $1,100 a week. Borrow that for 52 weeks a year for four years – you get a debt of nearly $70k – and that’s even without allowing for inflation which might adjust the training wage (or might not, if the ACT party were to get its way) but which would lead to higher borrowing entitlement..
[17] See this set of FAQs. See also Jones (2023) and the Apprenticeship Employer Network Victoria analysis referred to above.
[18] Jackson (2014), Coll et al (2009), Ferns et al (2014), Coll and Kalnins (2009), Orell (2004)
[19] The University of Waikato expects nearly all bachelors students to complete WIL. Victoria University also promotes WIL to students in all fields.
[20] See Jackson (2015), Jackson et al (2017), McNamara (2013), Patrick et al (2009)
[21] Note, there were no student loans in 1987. Mr Lange, unsympathetic, advised the protesting trainees to shop at Farmers, as he did!
[22] Note also that under current allowances settings, a $380/week additional income would see a student’s allowance reduced by ~$95 a week). But the living cost student loan entitlement would be unaffected.
