Opinion
Craig Fowler presents his imaginings of a fictional debate in response to a Jobs and Skills Australia paper on a Skills-First tertiary system, making valuable observations on some potential issues with achieving an agreed way forward for Australia’s tertiary sector.
Moderator: “Jobs and Skills Australia (JSA) in a recent paper propose a Skills-First tertiary system built on a national skills taxonomy, a modernised AQF, multidirectional credit transfer and an expanded RPL. Tonight, this debate hears four diverse views. A Vice-Chancellor who believes JSA’s paper misunderstands what a university is for, a labour market economist who believes the paper is half right, a Chief People Officer who says as a customer they were never interviewed, and a Futurist who believes it’s just rearranging deckchairs. So, VC you have the floor first!”
Vice-Chancellor: “A Degree Is Not a Skill Box”
“At the risk of being cast as a defender of privilege, a sandstone reactionary guarding the gates against carers, migrants and mid-career workers, I stand here not to defend privilege but to defend precision; and – in our world – this my peer review.
Consider the sleight of hand at its heart. We are assured, repeatedly and soothingly, that skills will sit “alongside, not instead of qualifications (pp2)”. Yet the paradigm is ‘skills-first’. You cannot put something first without putting something else second. Language is not decoration, it sets policy. When the organising principle of all tertiary architecture becomes the skill, the qualification is quietly demoted from an educational achievement to a box – packaging around a bundle of tradeable units. ‘Alongside’ is the anaesthetic administered before the operation.
And what a definition of skill to swallow! A skill, the paper says, is “a valued and purpose-driven human ability…that is acquired…through learning and practice…. a dynamic function of an individual’s knowledge, experience, and personal attributes (pp6).” Knowledge. Experience. Personal attributes. That is not a definition of skill – that is a definition of a person. Once “skill” has swallowed all three whole, what exclusive territory remains for the qualification to occupy? None! The paper then insists degrees retain some residual “coherence and trust (pp5)” function – having definitionally sliced out all vital organs upon which such functions depend, structured knowledge acquisition, sequenced curriculum, assessed depth.
Now to evidentiary standard. This rests on a single load-bearing claim: that our learning outcomes constitute “comparatively opaque labour market signalling” and “would benefit from disciplined application of a skills-first lens” (pp8)”. Where is the direct evidence – who sat with our Academic committees? Where is Australian data on employer confusion, or comparison against VET signalling. And is ‘skills-first’ to become an ‘if not, why no’ Compact condition for ‘self-accrediting’ institutions?
And whose agenda is this, really? Nothing but OECD and World Economic Forum citations that support skills-first rhetoric in general and this claim about Australian degrees not at all. It’s Global consultancies, Ed-tech vendors with taxonomy products to sell. The push to rebuild university curricula around granular skills taxonomies was devised in Davos and Silicon Valley, now imported into Canberra wearing the borrowed clothing of productivity reform.
Look to the test case: America’s O*NET, the most sophisticated occupational taxonomy on earth, three decades old. Has Harvard rebuilt its curriculum around it? Has Berkeley? Has any mainstream American university used it for wholesale curriculum blueprints? No. Curriculum remains disciplinary and accreditor-driven, as it should. Australia is being persuaded to adopt something that does not exist anywhere in the world. We would not be fast followers. We would be alone.
And how is ‘skills-first’ to be applied across diverse fields of education – the paper is stone silent. A university education exists to do the things no taxonomy can capture: to teach a young person to synthesise contested ideas, to challenge paradigms, to wrestle with material that resists tidy mapping. A pure mathematics doctorate, a fine arts program, a medical degree are not structurally comparable cases for a single reform logic. Strip-mine the curriculum into a checklist of mapped, machine-readable skill competencies and you will produce graduates who are legible to algorithms and useless to civilisation.
Arguing ‘If we understand the skills base – the taxonomy – we have a better chance of aligning our education and training system with the jobs of the future’ is not a neutral framing. It implies that the traditional knowledge orientation of universities is a systemic limitation to be corrected – that scholarship is a problem the taxonomy solves. This rhetoric drives a fundamental re-conception of a university's purpose: from the learning, creation and advancement of knowledge to utilitarian skilling.
If this is the case, JSA needs to take this argument to Parliament. The HESA 2003 Objects, alongside our Threshold Standards, make explicit what a university is for. And it’s not to act as a compliance conveyor belt for modular government-designated job training. 'Skills-first' blithely ignores these legal anchors, rewriting HESA 2-1(b) university purpose without amending a single word of the Act. The risk is that deep, transformative scholarship is shunned to slide away via the pejorative exit.
Worse, ‘harmonisation’ via a credit recognition framework will become JSA’s constructed skills taxonomy translated into mandatory curriculum obligations (credit entitlements) and enforced compact allocations. This three-layer architecture risks transfer of pedagogical authority from autonomous institutions to a centrally directed, automated system to chase CSP efficiencies.
Finally, who governs all this? Who arbitrates “outcomes-based equivalence (pp11)” when my institution and a private provider disagree? Every answer will run via ATEC. Universities Australia has consistently argued for respect of institutional autonomy. This Skills-First thesis does little to allay concerns that ATEC’s role is stewardship proxy control of both student numbers and course content.
Labour Market Economist: “Valuable diagnosis – solutions inadequate”
“Dusting myself down from the sandstone blast, I find myself agreeing with roughly half of the VC’s diagnosis. She overreads ‘first’, and under emphasises it is “not about narrowing HE to immediate job outcomes, but … making the skills…embedded in degrees more visible…transferable (pp8)”.
So, let’s do what economists do: separate the functional mechanism from the political rhetoric. What does this paper get right? Its best move – and one most skills-first advocates completely miss – is explicitly identifying matching efficiency as the primary lever through which tertiary reform lifts productivity. Unemployment and vacancies coexist in Australia not simply because we have too few workers or too few jobs, but because of friction in how people and jobs find each other.
That friction is quantifiable, highly costly, and driven by information asymmetry. If sharper skills signalling reduces structural mismatch, aggregate economic output rises without requiring a single extra dollar of capital injection. That isn't a vibe; it’s foundational labour economics, and JSA deserves credit for anchoring its paper there.
The empirical evidence on mobility is telling. Look at JSA’s Occupation Mobility Interactive Tool. It proves the Australian labour market is a complex network, not a linear ladder. Pathways cluster tightly around shared capabilities and overlapping training tracks. Certain roles act as high-velocity gateways connecting workers to a wide range of destinations; others are absolute cul-de-sacs. Workers move fluidly within these clusters but rarely cross their boundaries.
Now ask: what determines the walls of those clusters? Partly genuine capability boundaries – but partly recognition failure. A worker whose skills would transfer perfectly well cannot prove it, so the qualification system prices the move as if they were starting from zero. That is the friction the paper is trying to attack, and the equity dimension is real: the people trapped behind those walls are disproportionately migrants with unrecognised credentials, carers returning to work, and mid-career workers whose learning happened on the job.
But a sound diagnosis is not a viable reform blueprint, and here the paper stalls. First, the definitional flaw the VC exposed is fatal, not pedantic. If 'skill' lazily absorbs knowledge, experience, and character traits, you cannot build an auditable measurement framework. Economics runs on data we can count. Give me a structured, three-tier framework – task-level skills, transferable capabilities, and domain knowledge—and I can model clear credit-transfer rules. You cannot build a stable regulatory architecture on 'a dynamic function of individual attributes.'
Second, a move towards ‘capability’ instead of ‘competency’ in VET assessment is doable in some fields but problematic. Competency-based training, whatever its systemic flaws, yields falsifiable, assessable units. Capability and integrated proficiency demonstrated “over time… rather than… at a single point of assessment (pp10),” may be a truer account of how expertise develops, but at scale, without an assessment methodology, providers will quietly revert to task-based marking or drift into unfalsifiable self-report. Exactly the gameable low-trust RPL regime JSA wants to eradicate.
Third, the framework is entirely field-blind. Skills-first logic maps cleanly to fast-moving digital roles and trades where legacy credentials carry weak signalling value. It strains to the point of breaking in highly regulated, high-stakes professions like nursing, engineering, and law. Treating a plumbing apprenticeship and a medical degree as one reform problem is the paper’s largest unexamined assumption, and it hands the VC her best argument for free.
So, my prescription is practical. Pilot before you build: run pre-registered trials across a regulated profession, a trade, and a generalist degree, and publish the failures alongside the successes. Quantify the taxonomy: tier the definitions so they can be measured. Specify the anti-gaming and quality assurance regime before scaling RPL, not after. And tackle the governance question honestly rather than dressing a transfer of authority as stewardship coordination”.
Chief People Officer: “I Am the Customer Nobody Interviewed”
“Let’s start with a reality check. I run recruitment for one of this country’s largest employers. We hire ten thousand people a year. I’ve just listened to arguments about what signal my organisation will trust. None of them have ever sat in a high-volume hiring huddle, watched a recruiter clear three hundred applications before lunch, or defended selections before the Fair Work Commission. The JSA paper has a fatal flaw: it is a supply-side document pretending to answer a demand-side question. I am the customer this entire architecture is being built for, and nobody interviewed me.
Here’s my corporate reality, disconnected from any ivory tower or frictionless skills market. My company, like others on the ASX, stripped degree requirements from our job descriptions. Our press releases were sincere, but they hid the operational truth: our Applicant Tracking Systems still screen on formal credentials. Our hiring managers still use degrees as a fast tiebreaker under time pressure. When a critical hire goes wrong, no manager ever gets sacked for picking a candidate from a top-tier university. The degree persists not because it measures skill – it does that poorly – but because it serves as a cheap, legally defensible risk-hedge in a decision we must make in minutes and justify for years. It is skills-first in theory but in truth filters first in practice.
Now, look at the paper’s core promise: make skills visible, portable and trusted, and employers will buy them. Visible, perhaps. But trust has been thoroughly destroyed by the current skills-signal market. From where I sit, micro-credentials are not a signal; they are noise with a certificate attached. We see tens of hundreds of unstandardised badges with no common scale, no verification of who assessed them, and no proof the holder can perform under pressure. When everything is credentialed, nothing is. My recruiters do not read university learning outcome statements and will certainly never read a National Skills Taxonomy either – unless by default its integrated into their existing software workflows and backed by legal liability when it is wrong. A degree comes with an institution’s reputation mortgaged against it. Who mortgages anything against a skills passport? Show me who pays when the signal is fraudulent, then you can build trust.
Yet, the VC shouldn't celebrate. I am not defending her product. The JSA mobility data – elegant desktop detail but entirely unsurprising – describes my daily reality better than any AQF ever has. It well reflects our internal talent network. Our best cyber analysts were former bank tellers; our top operations leaders rose through contact centres. We promote across functional clusters constantly on demonstrated capability, ignoring formal qualifications because inside the firm, capability is fully observable. Outside the firm, capability is invisible, so we fall back on proxies we privately distrust. The paper is right, recognition failure is real, it traps workers in occupational cul-de-sacs, and it prices migrants and career-changers out of the market.
So, here is the demand side’s ultimatum. Give me a skills signal with four properties and I will hard wire it into my hiring systems; asap. Verified: assessed performance, not self-report as the economist’s gameable-RPL warning is not theoretical; I have watched credential fraud from the other side of the desk. Current: time-stamped and decaying, because a skill certified five years ago in anything digital is an antique. Contextual: it tells me not just that the skill exists but at what proficiency level and under what conditions, tiered the way the economist proposed. And warranted: an accountable assessor with credibility to lose. This would be far better than a legacy degree filter.
One closing word to JSA: You built a mobility model from worker data. Build the next one from employer data. Map what recruiters click, what they ignore, and what actually changes an algorithmic shortlisting decision. Until you can prove a skills taxonomy seriously moves the dial in a real hiring outcome in an Australian firm, everything argued from this podium tonight is hypothesis”.
Radical Futurist: “You Are Debating Activities in Soon to be Vacant Space”
“I have listened to three eloquent speeches about how to renovate a system already bust. The VC defends the ivory tower. The economist wants to pilot improvements in the tower. The CPO wants the tower to issue better exit receipts. Hello, the tower is vacating and no longer the ‘go to’ place!
Look at the numbers. A bachelor’s degree takes three to four years to design, accredit and deliver. The half-life of a technical skill shrinks with every generative AI release cycle. By the time a curriculum committee has agreed on its learning outcomes – let alone mandatorily mapped to anyone’s taxonomy – the labour market those outcomes describe has moved on.
The qualification is merely a slow, expensive proxy for capability, invented when direct verification of what a person could do was impossibly costly. That cost assumption has completely collapsed. AI now assesses person capability directly, continuously, and at scale – evaluating live work products, simulating professional judgment, and verifying performance against real-world market data. The economist thinks capability is non-assessable at scale. Wrong. Today this is being built in AI tools. No need for a campus, semesters, or graduation ceremony. When the proxy costs four years plus a HELP debt vs. direct measurement done in a day, the proxy dies. Not by legislation – by desertion.
The exit has already begun. Major enterprise and government employers are systematically stripping degree requirements across the Anglosphere. Vendor certifications in cloud, cyber, and AI outcompete traditional degrees as hiring signals in every high-growth sector. Western Governors University assesses competency directly and graduate people at a fraction of the traditional cost. Frameworks like America's O*NET or Europe's ESCO were never built to redesign university curricula; they were built to bypass them completely. They are the address books of a self-navigating labour market, learning to match people to work without any university’s input.
JSA’s mobility tool inadvertently proves this. It is a potent, indeed potentially subversive public policy. It reveals careers as networks: clusters of shared capability, gateway occupations, lattices of adjacent moves. Nobody’s career is a ladder anymore; it is a graph. And a graph is a data structure. Once careers are data structures, the natural navigator is not a three-year degree consulted once at age eighteen. It is a real-time matching engine consulted continuously across a fifty-year working life, reading a skills wallet that updates with every project you complete.
The VC claims a taxonomy cannot capture critical thinking or synthesis. I half agree. Deep human capabilities will survive AI’s absorption of routine cognition; on that the futurists and the classicists converge. But this does not require a self-accrediting monopoly or a rigid four-year block plus a qualification toll gate. Socratic mentorship does not require a sandstone quadrangle or a DVC for Advancement. If universities survive, they will do so as they began in Bologna in 1088: pure communities of inquiry. Their business model now risks being a suburban streetscape shop vacancy.
So, JSA’s paper is not too radical. It is a polite memorandum written to institutions whose consent will, in any case, be soon unnecessary. Build the taxonomy, build the wallets, build the recognition rails – but build them for a world where learning is continuous, verification is instant, and the university is just one niche provider among many. The question here is not whether skills come first. The market has already answered. The question is whether Australia’s institutions can fast adapt”.
Moderator’s PS: “Read the Exec. Summary of AI Epistemic Risks: Emerging Mechanisms & Evidence, co-authored by over 30 leading global minds-none of them Australian. It offers world-class systemic analysis. If you want to drop out of the global intellectual elite, blindly rush 'skills-first' while ignoring the cognitive offloading risks it warns against.”